Faculty of Bioscience Engineering

Academic year 2014 – 2015

Evaluation of new nitrogen analogues of curcumin as potential anticancer agents

Sander D’hoore

Promotor: Prof. dr. ir. John Van Camp Promotor: Prof. dr. ir. Matthias D’hooghe Tutor: dr. ir. Charlotte Grootaert

Master’s dissertation submitted in partial fulfillment of the requirements for the degree of

Master of Science in Bioscience Engineering: Food science and Nutrition

“De auteur en de promotor geven toelating deze scriptie voor consulatie beschikbaar te stellen en delen ervan te kopiëren voor persoonlijk gebruik. Elk ander gebruik valt onder de beperkingen van het auteursrecht, in het bijzonder met betrekking tot de verplichting uitdrukkelijk de bron te vermelden bij het aanhalen van resultaten uit deze scriptie”! “The author and the promotor give the permission to use this thesis for consultation and to copy parts of it for personal use. Every other use is subject to the copyright laws, more specially the source must be extensively specified when using results from this thesis”

Gent, June 2015

!!

!!!!!!!!!!!!!!!!!!The promotor, The promotor, The author, !!!!!!Prof. dr. ir. John Van Camp Prof. dr. ir. Matthias D’hooghe Sander D’hoore!

I

Acknowledgements This master’s dissertation was performed in order to graduate as bio-engineer in food

science and nutrition. This paper would not have been the way it is now without the help of

various people. Therefore I would like to say some words of thank.

I wish to express my sincere thanks to Prof. dr. ir. John Van Camp and Prof. dr. ir. Matthias

D’hooghe for giving me the opportunity to work on this interesting scientific topic, for

providing me with all the necessary utilities for this research, their expertise insights and

critical reading of the manuscript.

I would like to express the deepest appreciation to dr. ir. Charlotte Grootaert for her help in

the lab, her insights, immense enthusiasm, motivation and critical reading of the manuscript.

Furthermore I would like to say thanks for giving me the freedom to help determine the

direction of this research.

I would also like to thank ir. Rob De Vreese for kindly providing me with the curcumin

analogues, his participation in the monthly meetings and scientific help.

In addition, a thank you to everyone from the cell lab for the great atmosphere. Thanks

Bryan, Isabela, Bea, Senem, Evelien, Matthijs, Hanne, Hannah and Katrijn!

Finally, thanks to family and friends for their unconditional support. Specifically, I want to

thank Rachel Macaluso for her support and patience.

II

Abstract Curcumin, a polyphenolic compound from the Curcuma longa plant, is widely used as a

spice in food and possesses anti-cancer activities [1]. The use of curcumin in clinical

applications is however limited due to its low bioavailability. Major reasons are the low water

solubility, poor gut absorption and rapid degradation and metabolisation of curcumin [2]. The

low bioavailability of curcumin may be improved by removing the unstable β-diketone moiety

[3]. In this research the bioactivity of 15 newly synthesized nitrogen analogues in which the

β-diketone function was altered and 3 non-nitrogen analogues were investigated in vitro

using endothelial (EA.hy926), intestinal (Caco-2 and HT-29), ovarian (CHO) and hepatic

(HepG2) cell lines. The analogues bearing an N-alkyl functional group without a methoxy

group next to the hydroxyl group on the aryl moiety demonstrated highest cytotoxic effects

with SVD99 > SVD95 > SVD90. The isoxazole derivative (SVD130), that did contain this

methoxy group, also showed high cellular cytotoxic activities. The non-nitrogen analogues

curcumin, bisdemethoxycurcumin, SVD61 and SVD62 were in general the compounds with

under average cytotoxic effects. SVD62 however was highly toxic in the endothelial cell line.

The cytotoxic effects in vitro were not associated with increased intracellular reactive oxygen

species. The compounds that showed highest cytotoxic effects on the endothelial cell line

could be detected intracellularly through HPLC analysis. SVD67 however showed toxic

effects on the ovarian cell line while it was not detected intracellularly in this cell line. High

antioxidant activities were observed (by the FRAP and DPPH radical scavenging assay) for

the compounds with an ortho-methoxy and para-hydroxyl substitution pattern. When this

methoxy group was removed, small antioxidant effects were still observed. The removal of

the para-hydroxyl group however resulted in no antioxidant activity.

From these results we can conclude that altering the chemical structure of curcumin indeed

leads towards distinct shifts in cellular viability, behaviour and uptake. Therefore, the

bioavailability and anti-cancer bioactivity of curcumin analogues, especially the nitrogen

analogues with hydroxyl and without methoxy on the aryl group are promising to be studied

in vivo.

III

Samenvatting Curcumine is een polyfenol afkomstig van de wortel van Curcuma longa en wordt wereldwijd

gebruikt als specerij in voeding. Er is aangetoond dat curcumine onder andere kanker

positief kan beïnvloeden [1]. Het gebruik van curcumine is echter sterk beperkt voor

klinische toepassingen, dit door de lage biobeschikbaarheid van curcumin. Deze lage

biobeschikbaarheid is het resultaat van een lage wateroplosbaarheid, lage opname in de

darm en snelle degradatie en metabolisatie [2]. De biobeschikbaarheid kan echter verhoogd

worden door het β-diketon in curcumin te verwijderen of te substitueren [3]. In dit onderzoek

worden 15 nieuw-gesynthetiseerde stikstofanalogen waarbij de β-diketongroep werd

veranderd en 3 analogen zonder stikstof in vitro onderzocht. Hiertoe werden endotheliale

(EA.hy926), intestinale (Caco-2, HT-29), ovarium (CHO) en lever (HepG2) cellijnen

aangewend, dewelke al dan niet van kankeroorsprong zijn. De toxische effecten waren het

meest uitgesproken bij de analogen met een alkylaminogroep waarbij geen methoxygroep

aanwezig was op de aromatische ringen (cytotoxiciteit van SVD99 > SVD95 > SVD90). Het

isoxazool curcumine-analoog (SVD130) die wel deze methoxygroep bevat, toonde

eveneens hoge cytotoxische effecten. De analogen zonder stikstof (curcumin,

bisdemethoxycurcumin, SVD61 en SVD62) waren in het algemeen minder toxisch. SVD62

toonde echter wel sterk toxische effecten op de endotheliale cellijn.

De toxische effecten in vitro konden niet verklaard worden door intracellulaire generatie van

reactieve zuurstofspecies (ROS). De meest toxische componenten konden intracellulair

gedetecteerd worden. Een uitzondering hierop was SVD67, dewelke toxische effecten

vertoonde in de CHO cellijn maar niet kon gedetecteerd worden intracellulair in dezelfde

cellijn. De sterkste antioxidanten (getest via FRAP en DPPH assay) bevatten zowel een

ortho-methoxy als een para-hydroxylgroep. Indien deze methoxygroep niet aanwezig was,

werd nog steeds een klein antioxidatief effect gezien. Het verwijderen van de hydroxylgroep

resulteerde echter in geen antioxidatief effect.

Uit deze resultaten kunnen we besluiten dat het veranderen van de chemische structuur van

curcumine leidt tot een verschuiving van de viabiliteit, het gedrag en de opname in de cellen.

De biobeschikbaarheid en anti-kankeractiviteit van curcumine-analogen, en dan met name

de stikstofanalogen met hydroxyl- maar zonder methoxygroep, zijn veelbelovend voor in vivo

onderzoek.

IV

Table of contents List of abbreviations ......................................................................................................... 1

Introduction ..................................................................................................................... 2

1. Literature study ......................................................................................................... 3 1.1. Structure and chemical properties of curcumin ............................................................... 3 1.2. Metabolism ................................................................................................................... 3 1.3. Molecular targets of curcumin ........................................................................................ 5

1.3.1. Transcription factors ......................................................................................................... 5 1.3.2. Enzymes ............................................................................................................................ 7 1.3.3. Receptors .......................................................................................................................... 7 1.3.4. Growth factors .................................................................................................................. 8 1.3.5. Inflammatory cytokines .................................................................................................... 8

1.4. Biological activity ........................................................................................................... 9 1.4.1. Antioxidation .................................................................................................................... 9 1.4.2. Anti-tumor activity via ROS generation .......................................................................... 11 1.4.3. Anti-tumor activity through p53 ..................................................................................... 12 1.4.4. Anti-angiogenesis and anti-tumor activity via NF-κB ..................................................... 13

1.5. Curcumin analogues ..................................................................................................... 13 1.5.1. Changes in substitution on the aryl ring ......................................................................... 14 1.5.2. Mono-carbonyl analogues .............................................................................................. 15 1.5.3. Substitution on the 1,3-keto-enol moiety ...................................................................... 15 1.5.4. Heterocyclic analogues ................................................................................................... 16

1.6. Cell line ....................................................................................................................... 18 1.6.1. Caco-2 ............................................................................................................................. 18 1.6.2. HT-29 ............................................................................................................................... 18 1.6.3. HepG2 ............................................................................................................................. 19 1.6.4. EA.hy926 ......................................................................................................................... 19 1.6.5. CHO ................................................................................................................................. 20

2. Problem statement and objective ............................................................................ 21

3. Materials and methods ............................................................................................ 23 3.1. Reagents and cell lines ................................................................................................. 23 3.2. Curcumin analogues ..................................................................................................... 23 3.3. Cell viability and proliferation tests .............................................................................. 25 3.4. Solubility in exposure medium ..................................................................................... 25 3.5. Antioxidant capacity .................................................................................................... 26

3.5.1. DPPH radical scavenging activity .................................................................................... 26 3.5.2. Ferric reducing ability of plasma (FRAP) assay ............................................................... 26

3.6. Reactive oxygen species (ROS) generation assay ........................................................... 26 3.7. Cellular uptake ............................................................................................................ 27 3.8. Statistics ...................................................................................................................... 28

4. Results .................................................................................................................... 29 4.1. Cell viability and proliferation tests .............................................................................. 29

4.1.1. Caco-2 ............................................................................................................................. 29 4.1.2. HT-29 ............................................................................................................................... 29 4.1.3. CHO ................................................................................................................................. 30 4.1.4. EA.hy926 ......................................................................................................................... 30 4.1.5. HepG2 ............................................................................................................................. 36

4.2. Solubility ..................................................................................................................... 36

V

4.3. Antioxidant capacity .................................................................................................... 37 4.3.1. DPPH radical scavenging activity .................................................................................... 37 4.3.2. Ferric reducing ability of plasma (FRAP) assay ............................................................... 37

4.4. ROS generation assay ................................................................................................... 38 4.5. Cellular uptake ............................................................................................................ 41 4.6. Principal component analysis ....................................................................................... 44

5. Discussion ............................................................................................................... 51

6. Conclusion ............................................................................................................... 59

7. Future developments ............................................................................................... 61

References ..................................................................................................................... 62

Annex............................................................................................................................. 67 7.1. Annex 1. Chromatograms and DAD spectra................................................................... 67 7.2. Annex 2. Cellular uptake assessed by HPLC ................................................................... 76 7.3. Annex 3. PCA analysis on MTT and SRB data ................................................................. 77 7.4. Annex 4. PCA analysis on ROS and cytotoxicity data on EA.hy926 cells .......................... 78 7.5. Annex 5. PCA analysis on ROS and cytotoxicity data on CHO cells .................................. 79

VI

Table of Figures Figure 1. Curcuminoids, degradation products and metabolites of curcumin. ...................................... 4 Figure 2. Molecular targets of curcumin ................................................................................................. 6 Figure 3. Stable 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical. ............................................................ 10 Figure 4. Apoptosis through the mitochondrial pathway. .................................................................... 12 Figure 5. Expression of numerous genes regulated by the activation of NF-�B ................................... 13 Figure 6. Molecular structure of different curcumin analogues ........................................................... 17 Figure 7. Curcumin and its analogues used in this research. ................................................................ 24 Figure 8. MTT values on the undifferentiated Caco-2 cell line ............................................................. 31 Figure 9. SRB values on the undifferentiated Caco-2 cell line .............................................................. 31 Figure 10. MTT values on the differentiated Caco-2 cell line. .............................................................. 32 Figure 11. SRB values on the differentiated Caco-2 cell line ................................................................ 32 Figure 12. MTT values on the HT-29 cell line ........................................................................................ 33 Figure 13. SRB values on the HT-29 cell line ......................................................................................... 33 Figure 14. MTT values on the CHO cell line .......................................................................................... 34 Figure 15. SRB values on the CHO cell line ........................................................................................... 34 Figure 16. MTT values on the EA.hy926 cell line .................................................................................. 35 Figure 17. SRB values on the EA.hy926 cell line ................................................................................... 35 Figure 18. MTT values on the HepG2 cell line ...................................................................................... 36 Figure 19. SRB values on the HepG2 cell line ........................................................................................ 36 Figure 20. Percentage inhibition on the DPPH radical scavenging assay ............................................. 39 Figure 21. Trolox equivalent (µM) on the FRAP assay .......................................................................... 39 Figure 22. Percentage fluorescence on the ROS assay on the EA.hy.926 cell line ............................... 40 Figure 23. Percentage fluorescence on the ROS assay on the CHO cell line ........................................ 40 Figure 24. PCA plot based on the cytotoxicity data .............................................................................. 44 Figure 25. Scree plot based on the cytotoxicity data............................................................................ 45 Figure 26. Factor scores based on the cytotoxicity data ...................................................................... 46 Figure 27. Positions of the R1, R2 and R3 functional groups .................................................................. 47 Figure 28. Factor scores including the cytotoxicity and ROS data on the EA.hy926 cell line. .............. 49 Figure 29. Curcumin and analogues without nitrogen ......................................................................... 51 Figure 30. Most toxic curcumin analogues in vitro ............................................................................... 52 Figure 31. Factor scores based on the on the EA.hy926 cell including cellular uptake ........................ 58 Figure 32. Most promising compounds to be tested in vivo. ............................................................... 59

Table of Tables Table 1. Solubility of curcumin and its analogues................................................................................. 37 Table 2. Assessment of cellular uptake of curcuminoids ...................................................................... 42

1

List of abbreviations AP Activator protein AR Androgen receptor ATCC American Type Culture Collection BDC Bisdemethoxycurcumin CHO Chinese hamster ovary cNOS Constitutive nitric oxide synthases COX Cyclo-oxygenase DCFH-DA 2’,7’-dichloro-dihydro-fluorescein diacetate DMEM Dulbecco's Modified Eagle Medium DMSO Dimethylsulfoxide DPPH 1,1-diphenyl-2-picrylhydrazyl radical EGFR Epidermal growth factor receptor FBS Fetal bovine serum FRAP Ferric reducing antioxidant power IC Inhibition concentration IKK IκB kinase IL Interleukin iNOS Induced nitric oxide synthases JNK c-Jun N-terminal kinases LPS Lipopolysachariden MAPK Mitogen activated protein kinases mAU Milli-absorbance units MTT Thiazolyl blue tetrazolium bromide NEAA Non essential amino acids NF-κB Nuclear factor κB NO Nitric oxide NOS Nitric oxide synthases P/S Penicillin-Streptomycin PBS Phosphate-buffered Saline PC Principal component PCA Principal component analysis ROS Reactive oxygen species SRB sulforhodamine B STAT Signal transducer and activator of transcription TE Trolox equivalent TGF-α Transforming growth factor alpha TNF Tumor necrosis factor TNF-α Tumor necrosis factor α TPTZ Tripyridyltriazine VEGF Vascular growth factor

2

Introduction Curcumin is a plant polyphenol extracted from the curcuma plant. It has a strong yellow

colour and therefore it has extensively been used as a spice in food preparation and as a

dye for textile processing. Curcumin is considered as a bioactive molecule with beneficial

health impact. In India it has traditionally been used to treat e.g. biliary disorders, anorexia,

cough, diabetic wounds, hepatic disorders, rheumatism and sinusitis [4]. Nowadays, the

impact of curcumin on western-type diseases such as obesity [5], diabetes [6], metabolic

syndrome [7], Alzheimer’s disease [8] and cancer [9] has been demonstrated.

Although curcumin has a large bioactive potential, its poor bioavailability strongly limits its

health effects. Therefore, some research groups have tried to modify the curcumin molecular

structure in order to promote its potential beneficial effects [10, 11].

In this research, we will investigate the influence of 19 chemically synthetized curcumin

derivatives on human cell viability, oxidative stress markers and cellular uptake. Therefore,

we will first give a literature overview of the structure and properties of curcumin, its mode-

of-action in a cancer context, and the intestinal fate of curcumin-type products.

Subsequently, various types of analogues of curcumin and their effects will be discussed

and a short overview of other ways to improve bioavailability will be given. Finally an

overview of the properties of the cell lines used in this project will be provided. Next, we will

discuss the methodology used in this thesis, and present the outcome of the in vitro tests.

Finally, these results will be interpreted using multivariate statistical techniques.

3

1. Literature study

1.1. Structure and chemical properties of curcumin Curcumin (diferuloylmethane) is a hydrophobic diarylheptanoid consisting of two aryl groups

linked to each other by 7 carbons. The molecular formula of curcumin is C21H20O6 and the

molecular weight is 368 g/mol. Curcumin is unique considering it occurs in two tautomers,

the enol and β-diketone form (Figure 1). The most stable form in solution as well as in solid

state is the enol form [12]. However Manolova et al. [13] showed that the enol form can be

shifted towards the keto tautomer by addition of water. This is due to the stronger specific

interactions of the water molecules with the keto form. The maximal absorbance of curcumin

lays between 415 and 420 nm in acetone. It has a yellow colour at a pH between 2.5 to 7.0

and becomes more red at a pH higher than 7.0 [14]. Curcumin is found in turmeric, which is

derived from the rhizome of the herb Curcuma longa. Other major curcuminoids in turmeric

are demethoxycurcumin and bisdemethoxycurcumin. Commercial curcumin is a mixture of

different curcuminoids with approximately 77% curcumin, 18% demethoxycurcumin and 5%

bisdemethoxycurcumin [15]. A lesser known curcuminoid which appears in turmeric is

cyclocurcumin [16]. Curcumin is poorly soluble in water at room temperature (less than 0.03

µM) and at least 90% decomposes after 30 minutes in an aqueous neutral-basic solution of

37 °C [17]. Main degradation products are ferulic acid, ferulic aldehyde, feruloyl methane,

vanillin and trans-6-(4’-hydroxy-3’-methoxyphenyl)-2,4-dioxo-5-hexenal (Figure 1) [18]. The

instability of curcumin can be reduced by lowering the pH, adding fetal bovine serum (FBS),

glutathione or ascorbic acid. In the presence of 10% FBS only 50% was decomposed after

8h incubation at 37°C. The reduced stability in neutral-basic conditions may be due to the

deprotonation of the phenolic groups inducing the decomposition of curcumin [17].

Curcumin has been consumed as a dietary spice at doses up to 100 mg/day [4] and is safe

to use. Clinical trials have shown that doses up to 8 grams per day did not show any

adverse side effects [19].

1.2. Metabolism Curcumin is not often used in the treatment of different diseases because it has a rather low

bioavailability. This is due to the low water solubility, poor absorption and rapid degradation

and metabolisation of curcumin [2]. Curcumin undergoes O-conjugation to curcumin

glucuronide and curcumin sulphate and bioreduction to dihydrocurcumin,

tetrahydrocurcumin, hexahydrocurcumin and hexahydrocurcuminol in rat and human

intestinal fractions (Figure 1) [20]. These products can also become subject to

glucuronidation [21]. The metabolites generally have a lower bioactivity, however

4

tetrahydrocurcumin, one of the most abundant metabolites of curcumin, also possesses an

anti-inflammatory activity similar to that of curcumin [20, 22].

Figure 1. Curcuminoids, degradation products and metabolites of curcumin.

5

In a study by Ireson et al. [20] curcumin glucuronide, curcumin sulphate, tetrahydrocurcumin

and hexahydrocurcumin were demonstrated in the cytosol and microsomes of the intestinal

cells from humans and rats. Subsequently, the conversion to curcumin sulphate was also

demonstrated in intact rat gut sacs in situ. This establishes that also the intestinal mucosa of

rats is able to metabolize curcumin. Finally Ireson et al. [20] found that the enzymes

sulfotransferase 1A1 and 1A3 are able to sulphate curcumin and that alcoholdehydrogenase

has the potential to metabolize curcumin to hexahydrocurcumin.

Berginc et al. [17] performed an in vitro study with differentiated Caco-2 cells and an ex vivo

experiment using the small intestine of rats to examine the main barriers for curcumin

absorption. Using different specific transporter inhibitors, it was seen that curcumin

glucuronide formed inside the cells was most probably transported to the apical side of the

Caco-2 cells by P-glycoprotein (Pgp) and MRP2 (member of the ABC transporters) and to

the basal side by MRP1 and MRP3 transporters. Curcumin itself was found to be

transported back into the intestinal lumen by BCRP and to a lesser extent by the two ABC

transporters Pgp and MRP2. When comparing the results from their ex vivo and in vitro experiment it was observed that the permeability of curcumin through the rat intestine was

lower than through Caco-2 cells. This indicates the presence of an extra barrier such as the

mucus, which is present in the rat intestine but not at the apical side of the Caco-2 cells [17].

This mucus consists of primary water and glycoproteins and could possibly bind free

curcumin thus preventing its absorption in the gut.

1.3. Molecular targets of curcumin Many papers have described various biological activities of curcumin, including biochemical

and molecular targets such as transcriptional factors, inflammatory cytokines, enzymes,

kinases, growth factors and receptors [15, 23]. The main molecular targets of curcumin are

presented in Figure 2 and the most important ones will be discussed below.

1.3.1. Transcription factors

Nuclear factor-κB (NF-κB) is a transcription factor, which has been shown to regulate the

expression of more than 200 genes. These genes are involved in e.g. proliferation,

apoptosis and angiogenesis [24, 25]. NF-κB is present in the cytoplasm of almost all cell

types, however in an inactive state as it is bound to inhibitory proteins (such as IκΒα).

Stimuli such as IL-1 or tumour necrosis factor (TNF) can mediate the phosphorylation of the

inhibitory proteins, after which NF-κB is translocated to the nucleus as a heterodimer

consisting of p50-p65 subunits and becomes activated [25, 26]. This phosphorylation

happens through the activation of IκB kinase (IKK). Curcumin has the ability to inhibit the

IKK activity in a concentration dependent way. This inhibition of IKK by curcumin is most

6

likely directly but possibly also via an upstream kinase required for IKK activation [26].

Figure 2. Molecular targets of curcumin [15].

AP-1 is a dimeric transcription factor composed of the subunits Jun, Fos and ATF (activating

transcription factor) and can bind to the AP-1 DNA binding site. AP-1 plays a role in cell

proliferation and transformation [27] and in all probability plays a role in apoptosis [28].

Curcumin has been shown to inhibit the activation of AP-1 [29] and thus could limit the

proliferation and transformation of tumor cells.

The STAT (signal transducer and activator of transcription) proteins are molecules that are

activated after phosphorylation. They contribute to cell survival and cell growth by preventing

apoptosis through the increased expression of anti-apoptotic proteins [24]. STAT3 and

STAT5 have been implicated in multiple myelomas, lymphomas, leukemias and other

tumors [24]. It has been shown that curcumin suppresses the phosphorylation of STAT3

through interleukin-6 and consequent STAT3 nuclear translocation. Thus curcumin has the

potential to down-regulate the activation of STAT3 and can prevent the growth of certain

tumors [30].

7

Furthermore important transcription factors such as β-catenin, PPAR-γ (peroxisome

proliferator-activated recptor-γ), Egr-1 (early growth response-1) are all affected by curcumin

[24]. These transcription factors all play an essential role in various diseases.

1.3.2. Enzymes Cyclo-oxygenase-2 (COX-2) is an enzyme that activates the cyclooxygenase cascade. This

cascade reaction produces prostaglandins. COX-2 is overexpressed in different stages of

cancer [23] and the beneficial anti-tumor effects of COX-2 inhibitors are due to inhibition of

proliferation, increased apoptosis within the tumor, inhibition of stromal cell angiogenesis

and increased myeloid cell anti tumor immune surveillance [31]. Curcumin is able to prevent

the COX-2 induction by the tumor promoter TNF and fecapentaene-12 because COX-2 in its

turn is regulated via NF-κB and because curcumin can down-regulate NF-κB [32]. Another

study however showed that curcumin could directly inhibit COX-2 in the HT-29 colon

carcinoma cell line [33].

Nitric oxide (NO) is produced endogenously from L-arginine by a family of nitric oxide synthases (NOS). NO has a wide range of physiological and pathophysiological actions and

plays a role in inflammation and in carcinogenesis [34]. There are two groups of NOS: the

constitutive NOS (cNOS) and the induced NOS (iNOS). The iNOS is a group of synthases

that becomes induced due to inflammatory cytokines and lipopolysaccharides. The mRNA

for the iNOS is produced via NF-κB [35]. Pan et al. [34] showed that curcumin strongly

reduced the induction of iNOS in RAW 264.7 cells (derived from murine macrophages),

which were activated with LPS. The metabolites tetrahydrocurcumin, hexahydrocurcumin

and octahydrocurcumin however had less effect. The effect of curcumin toward iNOS is

most probably due to the blocking of NF-κB activation.

COX-2 and iNOS are also both up-regulated by activated MAPKs (mitogen activated protein

kinases). P38 is one such MAPK that was shown to be reduced by curcumin in an in vivo

study with rats and subsequently also COX-2 and iNOS levels were reduced. However the

activation of MAPK JNK (c-JUN N-terminal kinase) was not affected by curcumin [36].

1.3.3. Receptors The androgen receptor (AR) is a receptor that plays an important role in the development

and progression of prostate cancer because the growth promoting effects of androgen is

mediated mostly through the AR [37]. During prostate cancer, an increase in the cytoplasmic

levels of β-catenin is observed. The excessive free β-catenins overload the system and are

translocated into the nucleus where they are able to interact with the AR and augment AR-

mediated transcription [37]. Choi et al. [38] provided evidence that curcumin has an anti-

androgenic role in LNCaP prostate cancer cells. This might be due to the degradation of β-

catenin through regulation of downstream intermediates in the Wnt/β-catenin pathway.

8

The epidermal growth factor receptor (EGFR), a growth-factor-receptor tyrosine kinase, is

a receptor and ligands such as EGF and TGF-α can bind to the extracellular domain of this

receptor. After binding with a ligand, a cascade of biological signals is initiated from the

cytoplasm to the nucleus [39] which lead to the regulation of the expression of target genes.

The overexpression of EGFR is associated with various human tumors and appear to

promote solid tumor growth [40]. Curcumin was shown to reduce the expression of EGFR in

vitro in human cancerderived cells, including Moser, Caco-2 and HT-29 [41].

1.3.4. Growth factors Vascular endothelial growth factor (VEGF), a signal protein, is produced by cells and

stimulates angiogenesis. Through this angiogenesis oxygen can be supplied to hypoxic

tissues. This is essential during processes such as wound healing, embryonic development

and the menstrual cycle but also during different diseases such as adenocarcinoma

formation and chronic inflammation [42]. COX-2 is also induced by many pro-inflammatory

cytokines such as VEGF and IL-1β and the inhibition of COX-2 by certain drugs could down-

regulate VEGF via the MAPKs p38 and JNK [43]. As stated above, curcumin could down-

regulate COX-2 levels through p38 and Binion et al. [42] demonstrated that angiogenesis

and COX-2 were inhibited by curcumin in microvascular endothelial cells which were

stimulated with VEGF in vitro.

1.3.5. Inflammatory cytokines

Cytokines are small proteins that play a significant role in cell signalling. TNF-α (tumor

necrosis factor alpha) is involved in the regulation of infectious, inflammatory and

autoimmune responses. The main biological activities of TNF-α are its ability to cause tissue

injury through the pro-inflammatory properties of TNF-α, the ability to induce apoptosis in

some cancerous cell lines and finally the ability to alter intermediate substrate and energy

metabolism [44]. TNF-α is produced by e.g. macrophages and T-lymphocyte in response to

various stimuli [45]. Furthermore TNF-α gene expression is induced by e.g. IL-2, IL-1β and

TNF-α itself. A study performed by Zhong et al. [46] showed that curcumin was able to

suppress TNF-α in vascular smooth muscle cells via the NF-κB pathway. Curcumin might

thus help in cardiovascular diseases such as atherosclerosis and restenosis.

The Interleukins (IL) are another group of cytokines, which play a major role in the immune

system. IL’s are produced by e.g. macrophages and lymphocytes when they become

activated by certain stimuli. As mentioned above the inflammatory responses by TNF-α are

mediated through stimulation of certain interleukins such as IL-1, IL-2, IL-6 and IL-10. The

expression of these cytokines is mediated mainly through NF-κB and MAPK [47]. A study by

Cho et al. [47] investigated the effect of curcumin on the expression of TNF-α induced by

9

interleukins and TNF-α itself. They found that curcumin could inhibit the expression of IL-1β,

IL-6 and TNF-α in a human keratinocyte cell line.

1.4. Biological activity 1.4.1. Antioxidation

Antioxidants are molecules that have the ability to reduce radicals and thus the possible

damage caused by these radicals. The first mechanism of antioxidants arises by a chain-

breaking mechanism in which the antioxidant stabilizes a free radical through the donation of

an electron. The second possible mechanism is to prevent the formation of reactive oxygen

species (or reactive nitrogen species) by the removal of certain initiators of these processes.

In various ways these two mechanisms can occur, such as through electron donation, metal

ion chelation and gene expression regulation [48]. Since ROS can induce DNA damage and

subsequently initiation and progression of cancer [49], antioxidants may reduce cancer by

reducing ROS.

There are three main mechanisms by which polyphenols have antioxidant capacities. In the

first two, polyphenols inactivate radicals by transferring a hydrogen atom or a single

electron. The transfer of a hydrogen atom is accomplished by the homolytic cleavage of the

hydroxyl O-H bond of the phenol. The transfer of an electron is determined by the ionisation

potential of the polyphenol. More specific, a low ionisation potential favours the transfer of an

electron. Since polyphenols are aromatics, the newly formed radicals are more energetic

stable radicals since conjugation effects stabilize them [50]. In the third mechanism (which is

considered less common in polyphenols) the polyphenol chelates transition metal ions

forming stable complexes. These metal ions are further not available in reactions that

generate free radicals [50]. Curcumin possesses antioxidative characteristics [51]. Masuda

et al. [52] investigated the actual radical scavenging reaction of curcumin and found that

radical trapping occurred first at the phenol position (just like other phenolic antioxidants).

The radical position can then move to other positions and finally various non-radical stable

compounds are formed such as vanillin, ferulic acid and a dimer of curcumin. The

antioxidative effects are not only mediated through the phenol group, but also the methoxy

group at the ortho-position plays an important role. More specific, the methoxy group can

form an intramolecular hydrogen bond with the hydrogen atom of the phenol, leading

towards an improved removal of the hydrogen atom [16]. Also other electron-donating

groups at the ortho-position (such as an ethoxy group) are able to improve the radical

scavenging effect [53].

10

The DPPH radical scavenging activity test has been described as a simple and sensitive in

vitro antioxidant assay and is commonly used for the quantification of radical scavenging

activities of various compounds [51]. It is based on the neutralization of the stable 1,1-

diphenyl-2-picrylhydrazyl radical (Figure 3) that has a deep purple colour and absorbs

maximally at a wavelength of 517 nm. Upon neutralization the purple changes to a more

yellow colour and the absorption peak at 517 nm disappears. However, caution should be

taken interpreting the results since Xie and Schaich [54] raised questions on the application

of the DPPH test. The DPPH radical site appears to be difficult to access due to steric

hindrance. This implies that the accessibility of the radical position controls the neutralization

reaction more than the chemical properties of the antioxidant itself. Secondly, absorbance of

the DPPH radical is commonly measured after 30 minutes or longer. This neglects the

difference in reaction kinetic. Since radicals in living tissues have lifetimes of milliseconds to

seconds (hydroxyl radical for example around 10-9 s), the antioxidant reactivity rather than

the total antioxidant capacity is important [54]. Nevertheless, the DPPH assay is still widely

used and can give valuable information regarding total antioxidant capacities when the

neutralization reaction is run to completion.

The ferric reducing ability of plasma (commonly referred to as the ferric reducing antioxidant

power, FRAP) is another antioxidant assay commonly used in which the antioxidant reduces

a ferric-tripyridyltriazine complex (Fe3+-TPTZ) to the ferrous form (Fe2+-TPTZ ) through the

donation of an electron. This ferrous TPTZ form has an intense blue colour and shows an

absorbance maximum at a wavelength of 593 nm [55].

Figure 3. Stable 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical.

11

1.4.2. Anti-tumor activity via ROS generation Reactive oxygen species (ROS) is a term that refers to oxygen radicals (such as hydroxyl

(yOH), superoxide (yO2) or nitroxyl (NOy)) and non-radical oxidizing molecules (such as

hydrogen peroxide (H2O2) or organic hydroperoxides (ROOH)) [56]. ROS is involved in

fundamental activities including enzymatic reactions, activation of nuclear transcription

factors, mitochondrial electron transport, signalling pathways [57]. Nevertheless, it can also

cause harmful effects in living cells, however in normal conditions, antioxidants exist in

balance with ROS and can scavenge these species thus preventing free radical-mediated

damage. Depletion of antioxidants or an increased ROS production can lead to oxidative

stress and can damage compounds such as lipids, DNA and proteins [57]. Reactive oxygen

species can be generated endogenous or exogenous. Endogenous ROS are produced by

cellular metabolism such as mitochondrial respiration. Sources of exogenous ROS are UV

light, ionizing radiation and pathogens [49]. Curcumin can exhibit pro-oxidant activities

depending on its concentration and cell type through the generation of reactive oxygen

species [58-60]. ROS is a major cause of cancer through the damage of DNA and is

assumed to be involved in both initiation and progression of cancer [49]. However, ROS is

also involved in apoptosis through the mitochondrial intrinsic pathway [61] and thus could

play an important role in antitumor activities through apoptotic tumor cell death. Therefore,

cancer cells are more vulnerable to be killed by drugs that boost ROS levels since they

naturally exhibit a higher oxidative stress level when compared to normal cells [62]. Shortly

explained the mechanism of mitochondrial apoptosis via ROS is the following (see Figure 4):

the mitochondrial membrane becomes permeable by different apoptotic stimuli (such as

ROS) and because of this permeabilization, pro-apoptotic proteins (such as cytochrome c)

are released. Cytochrome C then forms a complex with apoptotic protease-activating factor-

1 (apaf-1) and activates procaspase-9. Through the catalyzing effect of procaspase-9,

effector caspase-3 is activated. Other apoptotic proteins furthermore help the caspase

activation by the neutralization of caspase inhibitors. It is this caspase-3 effector that

executes the final steps of apoptosis [61, 63, 64]. Bhaumik et al. [65] investigated the effect

of curcumin on AK-5 tumor cells and showed that curcumin could induce apoptosis in those

AK-5 cells through the activation of caspase-3 by the ROS pathway as described above.

12

Figure 4. Apoptosis through the mitochondrial pathway.

1.4.3. Anti-tumor activity through p53 The tumor suppressor protein p53 plays a central role in cell cycle regulation. This protein

has a supervisory role during the cell cycle and acts as a checkpoint control. When errors

happen during critical events (such as DNA replication or chromosome segregation) of the

cell cycle, checkpoints become activated and respond to this damage by arresting the cell

cycle to provide time for repair [66]. P53 is an important checkpoint that becomes activated

after DNA damage. The response of p53 is very heterogeneous depending on the cell type,

type of DNA damage, amount of p53 [64]. Furthermore, it is unclear when the cell goes into

cell cycle arrest or apoptosis under supervision of p53. However it is clear that p53 has a

central role during these processes. Jee et al. [67] performed an in vitro study with human

basal carcinoma (BCC) cells and were able to show apoptosis in these cells induced by

curcumin and a simultaneous increase in p53. Curcumin was able to promote de novo

synthesis of p53 or promote the synthesis of proteins for stabilization of p53, leading to

apoptosis. Other studies however show that curcumin is able to down-regulate the

expression of p53 at protein and mRNA level in in vitro studies on human colon cancer cells

(HCT-15) and BKS-2 cells (B cell lymphoma) [68, 69]. The protein p53 is mutated in almost

50% of all human cancers resulting in uncontrolled cell proliferation and progression of the

tumor growth, since curcumin was also able to down-regulate the production of p53, it could

thus also prevent cancer progression by preventing the uncontrolled cell proliferation that is

induces by the mutated p53 protein [68].

13

1.4.4. Anti-angiogenesis and anti-tumor activity via NF-κB Angiogenesis is the growth of new vascular capillaries from existing vessels and plays an

important role in processes such as wound healing and bone repair. However uncontrolled

angiogenesis is known to occur during e.g. tumor growth and rheumatoid arthritis. The

expansion of primary tumors and their metastasis to other organs is strongly influenced by

angiogenesis [70]. A central role during angiogenesis and tumor growth is reserved for

nuclear factor-κB as described above in section 1.3. Cytokines (such as IL-8) and growth

factors (TNF, VEGF) are produced by macrophages, neutrophils and other inflammatory

cells and are able to regulate angiogenesis. The production of these cytokines and growth

factors however is regulated through NF-κΒ and its activation. NF-κB also has effects toward

e.g. anti-apoptosis, inflammation and metastasis (Figure 5). Several studies show that

curcumin is able to down-regulate the activation of NF-κB and thus can suppress

downstream tumorigenic and angiogenic factors [71].

Figure 5. Expression of numerous genes regulated by the activation of NF-κB [72].

1.5. Curcumin analogues

The poor cellular absorption, metabolism and rapid elimination of curcumin are the major

reasons for the poor bioavailability of curcumin [2]. Because of this, different ways to provide

a better bioavailability have been proposed and investigated. A promising solution is the use

of certain curcumin analogues. Other possibilities are the use of nanoparticles, adjuvants or

emulsions containing micelles, liposomes or phospholipid complexes of curcumin. The use

of curcumin analogues will be discussed below.

14

Different curcumin analogues are naturally available. These analogues include

demethoxycurcumin, bisdemethoxycurcumin and cyclocurcumin or curcumin metabolites.

The last few decades different curcumin analogues with similar safety profiles as curcumin

have also been synthesized by man. This in order to overcome the low solubility and

bioavailability of curcumin and also to exploit information for further molecular design [22].

The instability of curcumin may be caused by the presence of its methylene and β-diketone

moiety. A deletion of the β-diketone moiety may thus result in an increased stability of the

curcuminoids [16]. Different types of curcumin analogues have been synthetized in the past.

The first group is characterized by changes in substitution on the aryl ring. This type of

analogue keeps the β-diketone structure. A second group of derivatives changes the β-

diketone structure. A mono-carbonyl arises if one of the carbonyl groups is removed.

Furthermore analogues with a substitution on the 1,3-keto-enol moiety and heterocyclic

curcumin analogues have been synthetized as well. Various curcumin analogues are shown

in Figure 6.

1.5.1. Changes in substitution on the aryl ring The antioxidative effects of curcumin and its analogues are influenced by the substitutions

on the aryl ring. As stated above, electron-donating groups at the ortho-position (such as

methoxy or ethoxy) are related to the antioxidant activity. An increased number of hydroxyl

groups on the aryl may contribute to an increased antioxidant activity [16].

Demethoxycurcumin and bisdemethoxycurcumin are 2 natural analogues of curcumin that

differ by the substitution on the aryl moiety. The difference with curcumin is that they lost

respectively one and two ortho-methoxy groups. Bisdemethoxycurcumin has the lowest

potency to suppress TNF-α-induced NF-κB activation followed by demethoxycurcumin, while

curcumin had the highest potency [73]. Furthermore curcumin was also found to be the most

active in a DNA cleavage reaction via the generation of reactive oxygen species followed by

demethoxycurcumin, which was more active than bisdemethoxycurcumin [59]. This suggests

the important role of the methoxy groups on the aryl. Yet, not in all cases curcumin is more

active. Other studies showed that demethoxycurcumin had the highest antiproliferative

activity against MCF-7 breast carcinoma cells while bisdemethoxycurcumin was more active

for cytotoxicity against ovarian cancer cells [74, 75]. Therefore, cell type and the

corresponding gene expression profiles may also determine the efficacy of the curcumin

analogues.

The substitution of the methoxy by an ethoxy group may also play a major role toward

cytotoxicity. In benzylpiperidin-4-one analogues (A and B) it may lead to higher toxicity in

certain cell lines (pancreatic, lung and renal cancer). However, the opposite was seen in a

non-small lung and colon cancer cell line [76].

15

Diphenyl difluoroketone C is an analogue fluorinated on the aryl, containing one carbonyl

function and a heterocyclic structure. C is shown to be one of the most active in anticancer

assays while still considerably less toxic than cisplatin, a commonly used chemotherapeutic

drug [77]. Diphenyl difluoroketone seems to have some of the same molecular targets like

curcumin. These molecular targets include COX-2, VEGF and IL-8 [77]. However another

study performed by [78] showed that C can disrupt the microtubule cytoskeleton while

curcumin did not show this activity. Thus, although curcumin and C have a very similar

molecular structure, they do not behave identically.

1.5.2. Mono-carbonyl analogues In mono-carbonyl analogues one of the carbonyl-functions has been removed. As stated

above that the methylene and β-diketone moiety contributes to the instability of curcumin,

the introduction of mono-carbonyl analogues may result in an increased stability.

Analogues were constructed including the deletion of a methylene group on top of the

carbonyl. In this way, a 5-carbon spacer was present between two aryl groups. The three

different types included a 1,5-diaryl-1,4-pentadiene-3-one (D), cyclopentanone (E) and

cyclohexanon (F) analogue. The cyclohexanon compounds were found to be more effective

in inhibiting LPS-induced proinflammatory cytokines TNF-α and IL-6 than the

cyclopentanone and 1,5-diaryl-1,4-pentadiene-3-one derived ones. With no change on the

aryl group, the cyclopentanone, cyclohexanone and 1,5-diaryl-1,4-pentadiene-3-one

analogue were more active than curcumin itself [11].

Various mono-carbonyl enone analogues of curcumin with 7, 5 or 3 carbons between the

aromatic rings have been investigated. Among 40 tested analogues with a 5-carbon spacer,

the most potent analogues toward the inhibition of TNF-α mediated activation of NF−κB

were the only two that contained a pyridine heterocyclic ring in the aryl (compounds G and

H). In general the 5-carbon spacer analogues were more active than the 3 and 7- carbon

spacers [79]. Note that C is also a 5-carbon spacer analogue, however this one was not

included in the experiments performed by Weber et al. [79].

1.5.3. Substitution on the 1,3-keto-enol moiety

Substitutions on the 1,3 keto-enol moiety include the formation of β-enaminoketone (I),

dioxime (J) and semicarbazone (K) analogues. A study by Simoni et al. [10] describes the

activities of β-enaminoketone and dioxime analogues. The β-enaminoketone analogues

containing a N-alkyl (methyl or ethyl) substituent show comparable (or little lower) cytotoxic

activities as curcumin. When an aromatic portion was present at the amine, the analogues

appeared to be inactive toward a breast cancer cell line (MCF-7 and MCF-7R) and a

hepatocellular carcinoma cell line (HA22T/VGH) in vitro. The dioxime derivatives show

higher cytotoxicity toward the tested cell lines than curcumin and the β-enaminoketone

16

analogues. The one carrying a benzyl compound was more active than the methyl

derivative. Furthermore the benzyl compound was able to reduce NF-κB activation (that up-

regulates various anti-apoptotic factors) in the hepatocellular carcinoma cell line whereas

this was not the case for the methyl analogue and curcumin. The latter two were however

able to decrease the anti-apoptotic factors to a lesser extent. This suggests a different

mechanism of action [10]. The curcumin semicarbazone was formed in order to up-regulate

its liposolubility. The antioxidative capacity and radical scavenging ability of this

semicarbazone was slightly higher than curcumin. Furthermore also its antiproliferative

activity toward the MCF-7 breast cancer cell line was higher than curcumin [80].

1.5.4. Heterocyclic analogues

The incorporation of the α,β−unsaturated keto group in a ring containing nitrogen (like an

isoxazole) seems desirable towards anti-cancer activity in different studies. An isoxazole (L) analogue was more active than curcumin and the dioxime derivative bearing a methyl group

in a breast cancer cell line. The dioxime bearing a benzyl group was more active than this

isoxazole analogue [10]. Piperidone analogues are another type of heterocyclic analogues.

They may be more active than curcumin. Compound M was found to be more active than

curcumin toward cancer cell lines of nine tumor groups (leukemia, non-small cell lung

cancer, colon cancer, central nervous system cancer, melanoma, ovarian cancer and renal

cancer) while other 4-piperidone compounds (for example N and O) may be less active [53].

The alkyl chain on the N of N-alkyl piperidones has furthermore an effect on the antioxidative

capacity. Increasing the length of this alkyl group (M versus N) resulted in an increased free

radical scavenging activity. This is also the case if the length of conjugation is extended (O

versus M and N) [53]. The antitumor activity in the in vitro tests in this last study went hand

in hand with their free radical scavenging activity (DPPH test) however no other tests (such

as for example ROS or NF-κB) were performed in this study. We should thus be careful to

say that an increased radical scavenging activity correlates with an increased antitumor

effect.

17

Figure 6. Molecular structure of different curcumin analogues (names of the various compounds are stated in the text above).

18

1.6. Cell line The in vitro experiments during this study are performed with different cell lines (Caco-2, HT-

29, HepG2, CHO and EA.hy926). A first advantage is that cell lines are easy to cultivate and

show a fast growth under standard culture conditions. Another advantage is that the used

cell lines are continuous which means that they can be used during the experiments up until

a passage number of 30 or even more. The use of primary cultures (cells freshly isolated

from tissues) is limited by their finite character leading towards fast senescence. Most of the

cell lines used in this study are derived from human or animal carcinomas and can give

preliminary information for the in vivo situation. The different cell lines used in this study are

described below.

1.6.1. Caco-2 The Caco-2 cell line is one originated from the human colon adenocarcinoma and was

obtained from the American Type Culture Collection (ATCC). Caco-2 cells are epithelial cells

and grow adherent. The epidermal growth factor receptor is expressed in this cell line [81].

When grown under standard culture conditions (37 °C, 10% CO2, DMEM, 10-20% FBS and

1% NEAA) the Caco-2 cells show a spontaneous structural and functional differentiation and

polarization after 21 days [82]. This differentiation starts when confluency is reached and is

one of the most relevant in vitro models to study the regulation and effects on intestinal

epithelial cells [83]. Upon differentiation, microvilli are visible at the apical side of the Caco-2

cell. The cells are asymmetric and show tight junctions, suggesting the polarization of the

monolayer. Furthermore also domes are found, which are the consequence of transport of

media compounds by the Caco-2 cells. The differentiation of the cells is characterized by

high levels of enzymes associated with the microvilli such as alkaline phosphatase and

sucrase isomaltase. The Caco-2 cells have low glucose consumption and only produce low

amounts of lactic acid. In contrast with in vivo enterocytes, they have a higher capacity of

storing energy as glycogen stores [84].

1.6.2. HT-29 The HT-29 cell line is also originated from a human colon adenocarcinoma and was

obtained from the American Type Culture Collection. Furthermore they also show an

epithelial morphology and grow adherent. Unlike Caco-2, HT-29 doesn’t show a

spontaneous differentiation, however when the glucose in the growth medium is removed or

substituted by galactose, inosine or uridine, a typical partial structural and functional

enterocytic differentiation can be seen. This differentiation is started when confluency is

reached (around the 15th day) and the cells form a polarized monolayer with the presence of

tight junctions and a brush border [84]. Another difference is that the HT-29 cells never form

domes when differentiated [82]. Hydrolases, which are present in the small intestine brush

19

border and in the brush border of the differentiated cells, include sucrose-isomaltase,

aminopeptidase N, dipeptidylpeptidase-IV and alkaline phosphatase. These enzymes in the

differentiated cells however show lower activities than observed in the normal small intestine

[84]. HT-29 cells have a high glucose consumption and lactic acid production and only

accumulate a medium degree of glycogen. Because of these findings, HT-29 cells are often

used when examining glucose metabolism [84]. Another difference with the Caco-2 cell line

is that HT-29 cells express high levels of COX-2 [85]. This might be relevant towards the

bioactivity of curcumin because curcumin was shown to be a potent COX-2 inhibitor [33].

1.6.3. HepG2 The HepG2 cell line originates from a human hepatocellular carcinoma and shows an

adherent, epithelial morphology. Human hepatocytes are suitable as an in vitro model for the

biotransformation in human liver and are important for toxicological and pharmaceutical

studies. A study by Wilkening et al. [86] showed that primary hepatocytes are more suitable

to predict metabolism in adult human liver cells than the HepG2 cell line because the drug-

metabolizing enzymes differ in the two different cell lines. Drug metabolism in the liver

happens in two phases. During the first phase the drug can undergo oxidation, reduction and

hydrolytic pathways adding functional groups. The second phase is characterized by a

modification to the added functional group such as the formation of O- and N-glucuronides

[86]. Phase I enzymes were shown to be less active and the expression was much lower in

the HepG2 cell line compared to the in vivo situation. Sulfotransferase 1A1 (a phase II drug

metabolism enzyme) was however expressed in a higher extent in HepG2 cells and is able

to sulphate curcumin [20]. Because the HepG2 cell line is easy to cultivate and the results

are reproducible, it is still often used.

1.6.4. EA.hy926 The EA.hy926 continuous cell line is obtained by hybridizing HUVEC (human umbilical vein

endothelial) cells with A549/8 human lung carcinoma cells. These cells contain many

endothelial properties such as the Von Willebrand factor and platelet-activating factor.

Furthermore, Weibel-Palade bodies have been identified in EA.hy926 [87]. These bodies are

endothelial cell specific storage granules. The Von Willebrand factor is located in these

bodies and plays a major role in haemostasis by platelet adhesion in case of vascular injury

[88]. The endothelial cell line grows adherent and is often used for studies on angiogenesis,

homeostasis, blood pressure and inflammation [89].

20

1.6.5. CHO The CHO cell line is one originated from the ovary of the chinese hamster (Cricetulus

griseus). The morphology is epithelial-like and the cells grow adherent and they require

proline in the growth medium in order to grow [90]. CHO cells are frequently used for

biomedical products and they can express a wide variety of recombinant proteins such as

human interleukin-6 or human follicle stimulating hormone [91]. An advantage of CHO cells

is that they have a high productivity [91] and that they grow fast when compared to Caco-2

cells, both grown under standard culture conditions.

21

2. Problem statement and objective Curcumin has been demonstrated to positively affect western-type diseases such as obesity

[5], diabetes [6], metabolic syndrome [7], Alzheimer’s disease [8] and cancer [9].

Curcumin is however not often used in the treatment of different diseases because it has a

rather low bioavailability. This is due to the low water solubility, poor absorption and rapid

degradation and metabolisation of curcumin [2]. Various techniques have been proposed to

improve this bioavailability. Nanoparticles [92], emulsions [93], and adjuvants [94] could offer

a solution. Another possibility that has already been suggested as an elegant way to

improve bioavailability is the use of curcumin analogues. The low bioavailability of curcumin

may be improved by removing the unstable β-diketone moiety [3]. The use of analogues

gives opportunities in unravelling the unique structural character of curcumin and may be

used for further molecular design and quantitative structure-activity relationship (QSAR)

analysis.

In this thesis, we will investigate the bioactivity of 15 newly formed nitrogen analogues and 3

non-nitrogen analogues of curcumin using a set of different cell lines and bioassays. To this

end, various nitrogen analogues of curcumin in which the β-diketone function was altered

were synthetized by ir. Sam Van Damme and ir. Maarten Van Bogaert under the guidance of

prof. dr. ir. Matthias D’hooghe and ir. Rob De Vreese.

The cytotoxicity of these compounds will be tested in vitro to show whether or not these

compounds show an improved biological effect towards various cancer cell lines in

comparison with curcumin. Non-cancer cells will also be included since cytotoxicity here is

not desirable. This will allow to determine which functional groups can improve the

cytotoxicity of curcumin towards tumor cells. Thereafter, the antioxidative capacities will be

discussed. DNA damage and subsequently cancer initiation and progression can be induced

by ROS [49]. Since ROS levels may be reduced by antioxidants, the antioxidative capacities

will be measured using the DPPH and FRAP assay. However, it has been described in

literature that curcumin can exhibit pro-oxidant and antioxidant activities depending on its

concentration and type of cancer cell [58-60], and that its anti-tumor activity may be

exhibited by ROS production [95]. Therefore, a fluorimetric assay that measures intracellular

ROS will be performed as well. This generation of ROS may be seen as positive in cancer

cells since ROS is also involved in apoptosis. This may lead to apoptotic tumor cell death as

described in the literature study in section 1.4.2.

22

Next, a rough estimation of the cellular uptake of the curcumin analogues will be performed

using analytical techniques in order to know whether the observed biological effects may be

the result of either the improved bioavailability of the analogue, or of the chemical properties

of the molecule itself. Finally, these results will be interpreted using multivariate statistical

techniques to conclude which analogues have the largest potential to be used as anti-cancer

agents.

23

3. Materials and methods

3.1. Reagents and cell lines Dimethylsulfoxide, triethylamine, trichloroacetic acid, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ),

trolox, sulforhodamine B (SRB), MEM non-essential amino acid solution (NEAA),

tris(hydroxymethyl)aminomethane, 2,2-diphenyl-1-picrylhydrazyl (DPPH), penicillin

/streptomycin (P/S), formic acid and 2,7-dichlorofluorescein diacetate were purchased from

Sigma-Aldrich. Both thiazolyl blue tetrazolium bromide and trypan blue were obtained from

Amresco. Glacial acetic acid, methanol, ethanol, isopropanol, acetonitrile, ethylacetate and

acetonitrile from Fisher Scientific. Dulbecco's phosphate-buffered saline (PBS-, no calcium

and no magnesium), Dulbecco's modified eagle medium (DMEM, high glucose + glutamax

and RPMI 1640) and trypsin/EDTA solution were obtained from Life Technologies and fetal

bovine serum (FBS) from Greiner Bio-one.

The cell lines, Caco-2 (intestinal), HT-29 (intestinal), CHO (ovarian), HepG2 (liver) and

EA.hy926 (endothelial) cells were obtained from ATCC. These cells were cultivated in the

same growth medium containing DMEM + glutamax, 1% NEAA, 1% P/S and 10% FBS in

order to avoid interferences of specific medium compounds with the results. Upon treatment,

no FBS was added to the medium. Curcumin and its compounds were dissolved in DMSO to

a concentration of 100 mM. When the cells were exposed, solutions of the compounds were

prepared from this stock solution in medium immediately before treatment. The final

concentration of DMSO in this growth medium was less than 0.1% to avoid cytotoxic effects

caused by DMSO.

3.2. Curcumin analogues Curcumin and the used analogues are shown in Figure 7 and were kindly provided by Prof.

dr. ir. Matthias D’hooghe and ir. Rob De Vreese. All compounds are nitrogen analogues of

curcumin, except for BDC, SVD61, SVD62 and curcumin itself. All compounds had a purity

of >99%, except for SVD62, which was a mixture of 90% SVD62a and 10% SV62b.

24

Figure 7. Curcumin and its analogues used in this research.

25

3.3. Cell viability and proliferation tests The MTT assay is based on the conversion of the yellow compound thiazolyl blue

tetrazolium bromide to formazan by the mitochondria of viable cells. This formazan has a

purple colour and absorbs maximally at a wavelength of 570 nm. Dead cells are not able to

convert this MTT dye. Therefore, by measuring the formation of the purple formazan, the

relative amount of viable cells can be obtained. An aliquot of cells were counted in a bürker

counting chamber after trypan blue staining. They were seeded at a concentration of 20000

cells in a 96 well plate and after one day the cells were treated without (control group) and

with curcumin or its analogues at a concentration of 1, 5, 10 and 25 µM. After 3 days of

treatment, 100 µL of culture medium was removed and 20 µL of MTT solution (5 mg MTT

/ml PBS) was added. After 2 hours of incubation in the dark at 37 °C and 10 % CO2 the

medium was removed and the 96 well plate was dried. Solubilisation of the formazan

crystals was achieved by adding 200 µL DMSO to each well. Finally, the absorbance was

recorded at 570 nm using a Bio-Rad Benchmark Plus microplate spectrophotometer. Each

condition was performed in triplicate.

The SRB assay is based on the measurement of cellular proteins. Sulforhodamine B can

bind electrostatically with basic amino acid residues if the cells are fixed with trichloroacetic

acid (TCA) and can be solubilized by weak bases [96]. Because of this quantitative staining

capacity of SRB, the assay is used to screen for cytoxicity and cell density. The cells were

seeded at a concentration of 20000 per well and treated with or without (control) compound

at 1, 5, 10 and 25 µM after 1 day. Three days after this treatment the cells were fixed by

addition of 50 µL of 50% TCA in milliQ-water for 1 hour at 4°C. The plate was cleaned 5

times with tap water and dried, after which the cells were stained with an SRB solution (0.4%

sulforhodamine B in 1% glacial acetic acid). After 30 minutes the plate was rinsed for 5 times

with 1% glacial acetic acid and dried. Tris buffer in a concentration of 10 mM was used to

dissolve the stain. Finally the absorbance was measured using the microplate

spectrophotometer at a wavelength of 490 nm. Each condition was performed in triplicate.

3.4. Solubility in exposure medium Solutions of 100 mM of curcumin and analogues in DMSO were diluted to 0.5, 5, 25 and 50

µM in exposure medium of 37 °C. The diluted solutions were transported to a 96 well plate

and incubated at 37 °C and 10% CO2. After 2 and 24 hours, crystal formation of the dilutions

was assessed using a phase-contrast light microscope (Motic AE31 equipped with a Zeiss

Axiocam camera) at a magnification of 10X10.

26

3.5. Antioxidant capacity 3.5.1. DPPH radical scavenging activity

The 1,1-diphenyl-2-picrylhydrazyl radical has a deep purple colour and shows a strong

absorption at a wavelength of 517 nm. Upon neutralization the purple changes to a more

yellow colour and the absorption peak at 517 nm disappears. A solution of 40 µg/ml DPPH in

methanol was prepared. This solution was mixed in a 9:1 ratio with the solution of curcumin

or its analogues in DMSO with a concentration ranging between 1 and 500 µM. This mixture

was incubated at 30 °C in the dark for 30 minutes. Subsequently, this mixture was added to

a 96 well plate and measured in the microplate spectrophotometer at a wavelength of 517

nm. Blanks without derivative and blanks without DPPH were run simultaneously. The last

blank was necessary to check for absorption of the derivatives themselves. Percentage

inhibition was calculated using the formula:

𝐼(%) = 𝐴𝑏𝑠𝑜𝑟𝑝𝑡𝑖𝑜𝑛𝑏𝑙𝑎𝑛𝑘 − 𝐴𝑏𝑠𝑜𝑟𝑝𝑡𝑖𝑜𝑛𝑐𝑜𝑚𝑝𝑜𝑢𝑛𝑑𝐴𝑏𝑠𝑜𝑟𝑝𝑡𝑖𝑜𝑛𝑏𝑙𝑎𝑛𝑘

∗ 100

3.5.2. Ferric reducing ability of plasma (FRAP) assay Acetate buffer of 300 mM was prepared by adding 3.1 g of sodium acetate trihydrate to 16

mL of acetic acid, which was diluted to 1000 mL with milliQ-water. The TPTZ (2,4,6-

tripyridyl-s-triazine) solution of 10 mM was prepared by adding 0.156 g of TPTZ to 50 mL of

ethanol. Lastly, a 20 mM solution of iron(III) chloride hexahydrate was prepared by mixing

0.5404 g of FeCl3.6H2O with 2 mL 37% HCl and 98 mL milliQ-water. The TPTZ and iron

solutions were freshly prepared on the day of the assay. These 3 mixtures were added in a

10:1:1 ratio to obtain the FRAP reagent. Finally, 100 µL of the samples were mixed with 900

µL of the FRAP reagent and after 4 minutes the absorbance was measured at 593 nm in a

Cary 50 bio UV-visible spectrophotometer. Trolox (a water soluble analogue of vitamin E)

was used as a standard and the FRAP value was calculated as trolox equivalent (µmol/L)

via a linear regression of the trolox standard curve. This method is similar to the protocol

used by previous workers [55, 97, 98].

3.6. Reactive oxygen species (ROS) generation assay 2’,7’-Dichloro-dihydro-fluorescein diacetate (DCFH-DA) is a non-fluorescent compound in

solution. This compound can be transported into the cell after which it can be hydrolyzed by

intracellular esterases. Subsequently, it can be converted to the highly fluorescent

dichlorofluorescein after oxidation through ROS. The cells were seeded in a black 96 well

plate with transparent bottom at a concentration of 20000 cells per well. When confluency

was reached, the cells were treated overnight with or without (control) curcumin or its

analogues. After 1 day, the cells were washed and incubated at 37°C and 10% CO2 with

DCFH-DA (20 µM) for 30 minutes. Hereafter hydrogen peroxide (0.25 mM) was added or not

27

(control) to induce oxidative stress. After 1 hour, the cells were washed again and culture

medium without phenol red was added. The plate was incubated again and was measured

after 1 hour with a Gemini XPS Microplate Reader with an excitation set at 485 nm and an

emission at 535 nm.

3.7. Cellular uptake Cells were cultivated in cell culture flasks and after confluency the cells were treated

overnight with a 10 µM solution of curcumin or its analogues. Subsequently the growth

medium was collected and the cell monolayer was washed for two times with cold PBS in

order to stop uptake and to wash the cells. Next, the cells were lysed by adding a PBS

solution with 10% ethanol. This solution was also collected and stored in a freezer no longer

than 7 days prior to extraction. Extraction reagent (95% ethylacetate, 5% methanol) was

added to the sample and mixing was performed thoroughly by vortexing for 1 minute.

Hereafter, the samples were centrifuged (3555 g) until good phase separation was obtained.

The organic phase was collected and the sample was washed for 2 times with extraction

reagent. Finally the samples were dried under nitrogen and resolved in 200 µL of mobile

phase A (90% methanol, 10% acetonitrile). The analogues were analyzed by reversed-

phase high-performance liquid chromatography with DAD detection (Ultimate 3000, Dionex,

Thermo). The wavelengths at which the compounds were detected are shown in the graphs

in Annex 1. A reversed phased YMC-pack C30 column (250 mm x 4.6 mm i.d., S-5µm) from

YMC (Schermbeck, Germany) was used with methanol:acetonitrile (9:1 v/v) as mobile phase

A and ethyl acetate with 0.25% triethylamine as mobile phase B. The flow rate was 1

mL/min, column temperature was 30 °C and the injection volume was 10 µL for the standard

solutions and 50 µL for the cellular extracts. The gradient profile was as follows: at 0 min

100% A, linear gradient to 75% A and 25% B in 10 min, isocratic for 5 min, linear gradient to

100% A over 2 min and finally isocratic for 4 minutes. Since curcumin could not be detected

using this protocol, a Dionex™ UltiMate™ 3000 UHPLC using a Waters Acquity BEH C18

column (2.1 mm × 150 mm, 1.7 µm particle size) and a Waters VanGuard Pre-column (2.1

mm × 5 mm) was used for this component. A constant flow of 0.3 mL/minute was used and

the injection volume was 10 µL. The column temperature was set at 35°C. Methanol (0.1%

formic acid) and MilliQ water (0.1% formic acid) were the mobile phases. The gradient profile

was as follows: at 0 min 5% methanol, isocratic for 5 min, linear gradient to 95% methanol in

15 min, isocratic for 5 min, linear to 5% methanol in 5 min and finally isocratic for min.

28

3.8. Statistics For the MTT, SRB, ROS, DPPH and FRAP assay a t-test was performed to statistically

compare the blank of every compound to the other tested condition. P-values lower than

0.05 are considered to be significant. For the ROS assay the modified Thompson tau test

was performed to identify and remove outliers. Principal component analysis and

hierarchical cluster analysis were performed in IBM SPSS Statistics 22.

29

4. Results

4.1. Cell viability and proliferation tests 4.1.1. Caco-2

Figure 8 and Figure 9 show the MTT and SRB results after 3 days of treatment on the

undifferentiated Caco-2 cell line. Curcumin showed a small response at 25 µM with an MTT

value of 90.01% ± 1.85% and SRB value of 86.73% ± 3.84%. The highest cytotoxicty was

seen for SVD90, SVD95, SVD99 and SVD130 with decreases of MTT values to 50% or

lower. The derivatives SVD1, SVD61 and SVD90 showed a significant increase for the MTT

at some concentrations. This increase may be due to cell stress. The SRB values showed a

similar profile. In general, concentrations lower than 25 µM often didn’t show a significant

response while 25 µM itself led to significantly lower cellular protein contents. Decreases of

20% or more at 25 µM could be seen for every compound except for curcumin and SVD61

(both are compounds without nitrogen).

In Figure 10 and Figure 11, the results for the MTT and SRB assay, tested at concentrations

of 1, 5, 10 and 25 µM, on the differentiated Caco-2 cell line are given. The optical density

never decreased to more than 75% in both assays. Curcumin itself was not able to

significantly alter the MTT and SRB values. In the MTT assay the compounds SVD62,

SVD66, SVD67 and SVD80 were able to significantly increase the mitochondrial acitivity at

some concentrations. Furthermore, significant decreases could be seen for MVB54, SVD2,

SVD29, SVD66, SVD99 and SVD130. The SRB values were very constant for almost every

component. The only significant decrease in cellular protein content was seen for SVD99.

Remark that SVD99 was also one of the components that gave a large decrease in the

undifferentiated Caco-2 cells. Furthermore, a small (<8%) but significant increase of cellular

protein content could be seen for SVD67 and SVD95 at 1 and 25 µM, respectively.

4.1.2. HT-29 First of all, curcumin did not show any significant changes in MTT value (p > 0.05) (Figure

12) at 1, 5, 10 and 25 µM when compared to the untreated cells. The derivatives MVB54,

MVB59, MVB74, MVB78, SVD1, SVD2, SVD29, SVD66, SVD67, SVD80 and SVD90 only

showed significant results at a concentration of 25 µM but not at lower concentrations. The

derivatives SVD95, SVD99 and SVD130 however also showed significant results at

concentrations lower than 25 µM. SVD99 already had a significant decrease at 1 µM.

Decreases to 50% or lower can be observed for MVB59, MVB74, SVD1, SVD80, SVD90,

SVD95 and SVD99. The SRB results (Figure 13) were again more or less similar as the

MTT results, however curcumin (10 µM), BDC (10 µM), MVB59 (1 µM) and SVD29 (1 and

10 µM) were able to significantly increase the cellular protein content. This increase was

30

always lower than 20%, except for MVB59 where the SRB values were increased to 138.50

± 9.93%. For curcumin, BDC and MVB59, this increase was again back to normal levels at

25 µM whereas for SVD29, a decrease to 78.35 ± 4.34 is seen at this concentration. This in-

and decrease may be explained by a stress response followed by cell death at higher

concentrations [99, 100].

4.1.3. CHO The results on the Chinese hamster ovary (CHO) cell line are presented in Figure 14 and

Figure 15. For most of the compounds a significant decrease can be seen with increasing

concentration in the MTT assay. BDC, MVB59, SVD61 and SVD62 showed only a very

small decrease with a value not lower than 80% when compared to the blank. Curcumin was

able to decrease the MTT value to 34.79 ± 3.68% at 25 µM. When comparing this value to

the other derivatives, MVB54, MVB74, MVB78, SVD90, SVD95 andSVD99 were also able to

decrease this value to 40% or lower. The SRB results showed a very similar pattern as the

MTT results. A decrease on both assays is associated with toxicity of the compound and

subsequent cell death [99, 100]. Note that SVD90, SVD95 and SVD99 gave a strong

decrease on the undifferentiated Caco-2, HT-29 and the CHO cell line.

4.1.4. EA.hy926 Finally, also an MTT and SRB were performed on the EA.hy926 endothelial cell line (Figure

16 and Figure 17). Curcumin was able to slightly but significantly decrease the MTT-value to

86 ± 5.66% at 25 µM. The addition of the compounds MVB59, MVB78, SVD1, SVD2,

SVD26, and SVD80 did not lead to a significant response by the cells. Highest cytotoxicity

was seen for SVD62, SVD95, SVD99 and SVD130 (MTT value <50% at 25 µM). On the

SRB, curcumin was not able to significantly affect the cells. No effect is seen for MVB54,

SVD26. Mostly, the same decreases are seen as in the MTT. The decrease in MTT value

may thus be explained by a decrease in number of cells. Important to notice is that SVD62

never showed strong cytotoxic effects on the other cell lines. Except for a decrease to 70%

on the MTT and SRB of the undifferentiated Caco-2 cell lines, SVD62 never decreased the

MTT and SRB values with more than 20%.

31

Figure 8. MTT values of curcumin and derivatives on the undifferentiated Caco-2 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and analogues tested at 0, 1, 5, 10 and 25 µM of compound. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

Figure 9. SRB values of curcumin and analogues on the undifferentiated Caco-2 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and analogues tested at 0, 1, 5, 10 and 25 µM of compound. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

*

** ** ** *

** ** *

** **

* *

*

**

** * **

* **

** **

* *

**

** * **

**

*

**

**

** ** **

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Op

tic

al

de

ns

ity

(%

)

Concentration (µM)

*

*

**

**

**

** * **

**

**

**

** ** **

* *

** **

**

**

**

**

** ** **

* *

**

**

**

*

*

** **

*

** **

**

** ** **

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Op

tic

al

de

ns

ity

(%

)

Concentration (µM)

32

Figure 10. MTT values of curcumin and derivatives on the differentiated Caco-2 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

Figure 11. SRB values of curcumin and derivatives on the differentiated Caco-2 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 490 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

* * ** ** *

* *

*

**

* *

* *

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Opt

ical

den

sity

(%)

Concentration (µM)

* *

**

0

20

40

60

80

100

120

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Opt

ical

den

sity

(%)

Concentration (µM)

33

Figure 12. MTT values of curcumin and derivatives on the HT-29 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

Figure 13. SRB values of curcumin and derivatives on the HT-29 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 490 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

** *

** **

* **

* *

**

*

** **

**

**

*

** ** **

*

**

0

20

40

60

80

100

120

140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Opt

ical

den

sity

(%)

Concentration (µM)

** * *

**

*

** **

* * **

** *

* * *

** ** *

* *

** ** **

**

**

** **

* * * **

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Opt

ical

den

sity

(%)

Concentration (µM)

34

Figure 14. MTT values of curcumin and derivatives on the CHO cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

Figure 15. SRB values of curcumin and derivatives on the CHO cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 490 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

** **

** **

* * ** **

**

**

** **

** **

** ** **

* *

**

** **

* ** **

* **

*

**

**

* ** *

*

**

**

**

**

**

**

**

* ** **

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

% O

ptic

al d

ensit

y

Concentration (µM)

** ** **

**

* * ** **

**

* * * **

**

* * *

*

*

*

* * ** *

**

* * * *

** ** **

* ** **

*

* **

**

** **

**

**

* ** **

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

% O

ptic

al d

ensit

y

Concentration (µM)

35

Figure 16. MTT values of curcumin and derivatives on the EA.hy926 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

Figure 17. SRB values of curcumin and derivatives on the EA.hy926 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 490 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

*

* *

**

** *

*

*

** **

*

*

* **

*

** ** ** ** **

**

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Opt

ical

den

sity

(%)

Concentration (µM)

**

* *

**

* * * * **

* * * *

** ** ** **

** * **

** **

** ** ** **

*

0 20 40 60 80

100 120 140

0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25 1 10 0 5 25

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Opt

ical

den

sity

(%)

Concentration (µM)

36

4.1.5. HepG2 First of all for the MTT assay on the HepG2 hepatocellular carcinoma cell line (Figure 18),

curcumin did not show any significant results (p < 0.05) when compared to the blank up to a

concentration of 25 µM. SVD66 and SVD95 showed significantly lower results starting at a

concentration of 5 µM and SVD26 and SVD67 at 10 µM. SVD61 however only showed a

significantly higher result (<15% increase) at 1 and 10 µM. The highest cytotoxicity can be

seen at 25 µM at which SVD95 could reduce the MTT value to 19.92 ± 1.42%. On the SRB (Figure 19), no significant results were obtained for curcumin and SVD61. For SVD66,

SVD67 and SVD95 the results were very similar to the MTT values. Finally for SVD26, a

small increase at 1 and 5 µM can be seen whereas this was not the case in the MTT.

Figure 18. MTT values of curcumin and derivatives on the HepG2 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 570 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). Structures can be seen in Figure 7.

Figure 19. SRB values of curcumin and derivatives on the HepG2 cell line after 3 days of treatment expressed as % of optical density compared to the untreated cells (at 490 nm). The x-axis indicates curcumin and the various analogues tested at 0, 1, 5, 10 and 25 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). Structures can be seen in Figure 7.

4.2. Solubility The lowest concentration at which crystals could still be observed was determined using

microscopy. Compounds were tested at 0.5, 5, 25 and 50 µM. Results are shown in Table 1.

For curcumin no crystals could be seen up till 50 µM. For every other derivative crystals

were observed and the solubility in aqueous solutions was thus lower. The lowest solubility

*

**

* * *

**

** *

* ** **

** 0

20 40 60 80

100 120 140

0 1 5 10 25 0 1 5 10 25 0 1 5 10 25 0 1 5 10 25 0 1 5 10 25 0 1 5 10 25

Curcumin SVD26 SVD61 SVD66 SVD67 SVD95

Op

tic

al

de

ns

ity

(%

)

Concentration (µM)

* *

** ** **

** ** **

** ** **

0 20 40 60 80

100 120 140

0 1 5 10 25 0 1 5 10 25 0 1 5 10 25 0 1 5 10 25 0 1 5 10 25 0 1 5 10 25

Curcumin SVD26 SVD61 SVD66 SVD67 SVD95

Op

tic

al

de

ns

ity

(%

)

Concentration (µM)

37

can be seen for MVB54, MVB74, SVD26, SVD29, SVD61, SVD62, SVD67 and SVD99 since

crystals were still identified at concentrations of 0.5 µM. After 1 day, in most cases the

concentration at which crystals were detectable was higher. This may be due to a higher

solubilisation in the growth medium or to the degradation of the compound.

Table 1. Lowest concentration at which crystals could still be identified after incubation of 2 hours or 1 day at 37°C and 10 % CO2.

Concentration (µM)

Concentration (µM)

2 hours 1 day

2 hours 1 day Curcumin /* /

SVD61 0.5 0.5

BDC 5 25

SVD62 0.5 0.5 MVB54 0.5 5

SVD66 5 25

MVB59 25 50

SVD67 0.5 5 MVB74 0.5 5

SVD80 25 25

MVB78 25 25

SVD90 25 25 SVD1 5 25

SVD95 25 25

SVD2 5 50

SVD99 0.5 5 SVD26 0.5 5

SVD130 5 5

SVD29 0.5 5 * / = No crystals could be detected at a concentration of at least 50 µM.

4.3. Antioxidant capacity 4.3.1. DPPH radical scavenging activity

The DPPH assay for curcumin and the analogues was tested at 5, 25, 50, 100 and 500 µM.

Results are shown in Figure 20. The compounds SVD26, SVD29, SVD61, SVD62, SVD67

and SVD80 never showed a significant percentage inhibition (p>0.05). BDC, SVD90, SVD95

and SVD95 only showed a significant increase in percentage inhibition (less than 15%) at a

concentration of 500 µM (p<0.01) It can be seen that curcumin, MVB54, SVD2, SVD66 and

SVD130 showed the largest antioxidant capacities with inhibition percentages ranging from

69 till 81% at 500 µM. Finally, MVB59, MVB74, MVB78 and SVD1 showed slightly lower

inhibition of approximately 50% at 500 µM.

4.3.2. Ferric reducing ability of plasma (FRAP) assay The results of the FRAP assay are given in Figure 21 expressed as trolox equivalent

(µmol/L). Trolox is a standard used widely in antioxidant assays and was measured at a

concentration ranging from 40 to 1600 µM. Firstly, the linear range was determined after

which a linear regression was performed with absorbance on the ordinate and concentration

(in µM) on the abscissa (linear range: 0-800 µM, y =0.0041*x; R2=0.99736). Trolox

equivalent (TE) was then calculated as TE (µM) =absorbance/0.0041. A trolox equivalent

comparable with the tested concentration means thus that the compound has an antioxidant

capacity equivalent with that of trolox. Curcumin and its analogues were measured at

38

concentrations of 50, 100 and 500 µM. The absorbance of the compounds was always

verified to be in the linear range. The results were comparable to the results obtained from

the DPPH radical scavenging assay. Curcumin, MVB74, MVB78, SVD2 showed similar or

even higher antioxidant capacities as trolox (TE > 500 µM at a concentration of 500 µM).

MVB54, MVB59, SVD1, SVD66, SVD99 and SVD130 showed mediocre antioxidant

capacities on the FRAP assay with trolox equivalents between 280 µM and 500 µM at 500

µM of the compound. BDC, SVD90 and SVD95 showed lower antioxidant activities. Finally,

SVD26, SVD29, SVD61, SVD62, SVD67 and SVD80 still showed significantly increased

trolox equivalents. However, since the increase was very low (TE<20 µM) these compounds

possess very small to non-existent antioxidant capacities.

4.4. ROS generation assay ROS values on the endothelial cell line expressed as percentage fluorescence compared to

the blank without oxidative stressor (H2O2) with DCFH-DA stain are demonstrated in Figure

22. Untreated conditions with or without H2O2 or stain were performed in every 96-well plate.

When H2O2 was added, similar or significantly increased ROS levels were obtained. ROS

values without stain always ranged between 30 and 82% compared to the blank with stain.

First of all, curcumin was the only tested compound that was not able to induce a significant

in- or decrease in ROS levels at any tested condition. For every tested compound, no

significant increases was detected at 1 µM (without oxidative stressor). Only for SVD80 and

SVD29 a significant decrease (p>0.01) to respectively 67.86 ± 8.52% and 60.07 ± 3.06%

was obtained. The compounds MVB74 (439%), MVB78 (267%), SVD26 (240%), SVD61

(189%), SVD62 (152%), SVD66 (191%), SVD90 (119%), SVD95 (414%) and SVD99

(299%) significantly increased ROS levels when 10 µM of the compound was added without

oxidative stressor. SVD80 again showed lower results with a decrease to 53 ± 3.49%. The

fluorescence after addition of 1 µM of the compounds and 0.25 mM of H2O2 was significantly

increased for BDC, MVB78, SVD26, SVD61, SVD90, SVD95, SVD99 and SVD130. So, the

addition of both an oxidative stressor and an analogue of curcumin at 1 µM could induce

increased ROS levels, whereas this was not the case without oxidative stressor. It cannot be

concluded that merely the H2O2 increases the ROS values since the ROS was not always

significantly increased by an addition of H2O2 in the blanks. This increase at 1 µM of

compound was however lower than with 10 µM of the compound with H2O2.

39

Figure 20. Percentage inhibition on the DPPH radical scavenging assay measured at 517 nm after 30 minutes of incubation in the dark at 30 °C tested at concentrations of 5, 25, 50, 100 and 500 µM. The x-axis indicates curcumin and the various analogues tested at 0, 5, 25, 50, 100 and 500 µM. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

Figure 21. Trolox equivalent (µM) on the ferric reducing ability of plasma (FRAP) measured at 593 nm with a concentration of 50, 100 and 500 µM of the different compounds. The x-axis indicates curcumin and the various analogues tested at 0, 50, 100, 500. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank). CUR = curcumin. Structures can be seen in Figure 7.

*

**

**

** ** **

**

**

**

** **

**

**

**

**

** ** **

**

** **

**

**

* **

**

** ** ** * **

**

**

-5

15

35

55

75

95

0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Perc

enta

ge in

hibi

tion

(%)

Concentration (µM)

** **

**

** ** **

** **

**

** **

**

** **

**

** **

**

** **

**

** **

**

** ** ** ** ** ** ** ** ** ** **

**

** * ** ** ** ** ** ** **

** **

** **

**

** **

**

0 100 200 300 400 500 600 700 800

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

0 100

CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Trol

ox e

quiv

alen

t (µM

)

Concentration (µM)

40

Figure 22. Percentage fluorescence on the ROS assay on the EA.hy.926 cell line relative to the blank with stain and without H2O2. Compounds were tested at 1 and 10 µM with or without H2O2 and measured at an excitation set at 485 nm and an emission at 535 nm. Error bars indicate standard deviations. *: p<0.05, **: p<0.01 (t-test compared to blank), OX = with oxidative stressor (0.25 µM H2O2), / = without H2O2, S = with stain (20 µM), NS = without stain. CUR = curcumin. Structures can be seen in Figure 7.

Figure 23. Percentage fluorescence on the ROS assay on the CHO cell line relative to the blank with stain and without H2O2. Compounds were tested at 1 and 10 µM with or without H2O2 and measured at an excitation set at 485 nm and an emission at 535 nm. Error bars indicate standard deviations. *:p<0.05, **: p<0.01 (t-test compared to blank), OX = with oxidative stressor (0.25 µM H2O2), / = without H2O2, S = with stain (20 µM), NS = without stain. CUR = curcumin. Structures can be seen in Figure 7.

** ** **

** ** **

**

** **

**

**

**

** * **

*

**

**

* *

**

**

** ** ** ** **

** ** * **

** **

**

**

**

**

**

**

** **

0

100

200

300

400

500

1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10

/ OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX

S NS CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Fluo

resc

ence

(%)

** **

* ** * **

** * * * * * * ** ** * *

* * *

** ** ** * * ** **

* **

**

** **

0

50

100

150

200

250

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

/ OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX / OX

S NS CUR BDC MVB54 MVB59 MVB74 MVB78 SVD1 SVD2 SVD26 SVD29 SVD61 SVD62 SVD66 SVD67 SVD80 SVD90 SVD95 SVD99 SVD130

Fluo

resc

ence

(%)

41

Interestingly, a huge (>400%) increase in ROS value can be seen for MVB74, MVB78,

SVD26, and SVD95. Note also that only SVD29 and SVD80 were able to significantly

decrease the ROS values while those two compounds did not show antioxidative or radical

scavenging activities on the DPPH and FRAP assay.

The ROS assay was also performed on the CHO cell line for which the results are presented

in Figure 23. With the addition of 0.25 mM H2O2 no significant increase in ROS level was

seen for the blank. Curcumin, SVD29, SVD67 and SVD90 were not able to significantly

increase or decrease the ROS values at all tested conditions. No decreases in ROS were

identified on the CHO cell line whereas SVD29 and SVD80 were able to significantly

decrease the intracellular ROS in the endothelial cell line. In contrast, SVD80 and MVB54

were the only analogues that increased the ROS value in all tested conditions. The

compounds MVB54, MVB59, MVB74, MVB78, SVD61, SVD80, SVD99 and SVD130 were

already able to exert an effect at 1 µM of analogue added without oxidative stressor. In

general, the increases on the CHO cell line were smaller than on the EA.hy926 cells. A

maximum increase of 193 ± 21% can be seen with 0.25 mM H2O2 and 10 µM SVD99. The

strong increase in ROS value (>300%) on the EA.hy926 cell line by MVB74, MVB78,

SVD26, SVD61, SVD90, SVD95 were not as high in the CHO cell line, where SVD90 did not

increase ROS at all.

4.5. Cellular uptake Cellular uptake of the various compounds was assessed through comparing both the

retention time and DAD spectra of the standard solutions and cellular extracts. This setup

allows only a rough estimation of the cellular uptake, as optimization of the extraction and

analysis parameters for each of the analogues was not performed due to time restrictions.

Therefore, only a qualitative interpretation of the results was possible. Nevertheless, this

approach may give some useful information about cellular uptake. The compounds curcumin,

SVD26, SVD61, SVD66, SVD67 and SVD95 were added to the EA.hy926, CHO and

differentiated Caco-2 cell line, whereas the other compounds were only tested with the

EA.hy926 cell line. The chromatograms, DAD spectra and peak areas are specified in annex

1 and 2. Table 2 summarizes the retention time, similarity of the DAD spectra of the

compounds extracted from the cells with the reference and the assessment of cellular

uptake. SVD61 could neither be detected with the HPLC or UHPLC protocol. Curcumin was

detected with the UHPLC and the other compounds with the HPLC.

42

Table 2. Assessment of cellular uptake of curcuminoids by reversed phase HPLC with DAD detection. The DAD spectra as well as the retention time were used to evaluate cellular uptake.

Standard EA.hy926 cell line CHO cell line Differentiated caco-2 cell line

tR (min) tR

DAD** Uptake tR

DAD Uptake tR

DAD Uptake (min) (min) (min)

Curcumin 21.09 /* / N* NT* NT NT / / N BDC 3.326 / / N NT NT NT NT NT NT MVB54 3.493 3.486 N Y* NT NT NT NT NT NT MVB59 3.466 / / N NT NT NT NT NT NT MVB74 3.593 / / N NT NT NT NT NT NT MVB78 3.317 / / N NT NT NT NT NT NT SVD1 3.593 / / N NT NT NT NT NT NT SVD2 3.275 / / N NT NT NT NT NT NT SVD26 3.741 3.771 N Y 3.756 N Y 3.764 N Y SVD29 3.478 3.496 N Y NT NT NT NT NT NT SVD61 ? ? ? ? ? ? ? ? ? ? SVD62 (1)* 3.055 ? ? ? NT NT NT NT NT NT SVD62 (2)* 3.188 ? ? ? NT NT NT NT NT NT SVD62 (3)* 3.489 ? ? ? NT NT NT NT NT NT SVD66 3.545 3.5 Y Y 3.583 Y Y 3.558 Y Y SVD67 3.82 3.844 N Y / / N / / N SVD80 (1) 3.259 3.293 N Y NT NT NT NT NT NT SVD80 (2) 3.446 3.474 N Y NT NT NT NT NT NT SVD80 (3) 3.627 / / / NT NT NT NT NT NT SVD90 3.259 3.298 Y Y NT NT NT NT NT NT SVD95 3.106 3.097 Y Y 3.134 N Y / / / SVD99 3.326 3.355 N Y NT NT NT NT NT NT SVD130 4.394 4.369 Y Y NT NT NT NT NT NT

* / = No peak at same retention time as standard, N = No, Y= Yes, NT= not tested, ? = results not sufficient to asses cellular uptake, (1) = first peak, (2) = second peak, (3) = third peak ** DAD spectra of cellular extract compared with standard. Y = similar DAD spectra, N = different DAD spectra

43

The standard solutions of SVD62 showed 3 major peaks. Since SVD62 consists of 90%

SVD62a and 10% SVD62b (see Figure 7) 2 peaks were expected. During storage of the

compound in DMSO, preparation or analysis some of the compound may have been

degraded. The loss of an acetoxy group (of both SVD62a or SVD62b) or benzyloxy group (of

SVD62b) could have resulted in the various peaks. However, in some cases two peaks that

are not fully separated may lead to a third peak in the middle. The result of the EA.hy926

cellular extract was furthermore qualitatively not good. Therefore cellular uptake was not

assessed for SVD62.

SVD80 also had 3 peaks. This is probably due to the loss of 1 and 2 acetoxy groups.

Furthermore, no major changes are seen in the DAD spectra of the 3 peaks. The

degradation of the compound did thus not result in changed UV/VIS absorption.

Confirmation by mass spectrometry is however needed for both SVD62 and SVD80. When

no peak could be identified at the same retention time as the standard, we assumed that no

cellular uptake has occurred. For SVD66 (on the EA.hy926, CHO and differentiated Caco-2

cells) and for SVD90, SVD95 and SVD130 (on the EA.hy926 cells) the same DAD spectra

were seen as the standard at the same retention time. Hence we concluded that cellular

uptake had occurred.

The compounds MVB54, SVD26, SVD29, SVD67, SVD80, SVD95 and SVD99 showed

peaks at the same retention time as the standard while the DAD spectra were not

completely identical. The first part of the spectra up till a wavelength of approximately 400

nm was comparable for the standard and cellular extract (see annex 1). At higher

wavelengths, the absorption was somewhat lower for the pure compounds than for the

compounds extracted from the cells. Small modifications (resulting in the same retention

time) of the structure through cellular activity, degradation in the cell culture medium or co-

elution of other medium compounds in very low concentrations may have resulted in this

higher absorbance. Although the DAD spectra of the modified compounds were thus not

completely identical, they still resembled those of the standards rather well. Furthermore,

since the retention time was also very similar, a cellular uptake was still concluded for those

compounds. Mass spectrometry could provide more information about the cellular uptake,

metabolisation and degradation products and should be performed in future experiments.

Further optimization of the HPLC protocol is advised for most of the structures. These

results however already give an indication of possible cellular uptake.

44

4.6. Principal component analysis To allow interpretation of the results, multivariate statistics, including principal component

analysis and hierarchical clustering analysis, was used to parts of the dataset generated by

the bioassays described above. Prior to principal component analysis, the dataset of the

toxicity tests was reduced to the IC50 and IC80 values, which refer to the concentration at

which the MTT or SRB value was decreased to respectively 50% and 80% of the untreated

condition. In Figure 24 the PCA plot of this dataset is shown. In this plot, the results of the

MTT and SRB assay on the differentiated Caco-2 cell line are not included since the

variability in those inhibition concentrations was too low, as the IC50 and IC80 were in general

higher than 25. Also the HepG2 cell line was not included in this test, because there were

too few analogues tested on this cell line. First the scree plot (see Figure 25) was screened

to see how many principal components should be selected. The scree plot displays the

eigenvalues in function of the principal components. These eigenvalues indicate how much

variance there is in the data in the direction of that principal component. Often the principal

components are retained with an eigenvalue higher than 1. Four principal components could

thus be selected based on this scree plot (Figure 25). Based on the fact that 60% of the

variance [101] was already explained by the first two principal components and in order to

graphically interpret the results, 2 components will be selected. The result of the PCA is

shown in Figure 24.

Figure 24. PCA plot based on the inhibition concentration on both the MTT and SRB assay. 50 and 80 refer to the IC50 and IC80. CHO, EA, ATCC and HT refer to respectively the CHO, EA.hy926, undifferentiated Caco-2 and HT-29 cell line on which the MTT and SRB assay were performed. “S_” preceding the name of the cell line means the SRB values were used. If not, the MTT values were used.

45

Figure 25. Scree plot based on the IC50 and IC80 values of the MTT and SRB assay on the CHO, EA.hy926, HT-29 and undifferentiated caco-2 cell line.

First it can be seen that the EA.hy926 endothelial cell line is differently influenced by the

various compounds since the variables containing information about the cytotoxicity on this

cell line are separated from the other cell lines. The undifferentiated Caco-2 cell line forms a

group close to the HT-29 cell line whereas the CHO cell line is also separated from the

intestinal cell lines HT-29 and Caco-2. The IC50 values are grouped together with the HT-29

cell line whereas the IC80 values show a separate group. When observing the results of the

MTT and SRB on the CHO line (see Figure 14 and Figure 15), it can be seen that if the

values decreased, this often already started at 1 µM. This was not the case for most of the

compounds tested on the other cell lines. This may explain the separation of the IC80 values

on the CHO lines. The cell line thus plays an important role towards activity of the various

compounds.

Principal component 1 (PC1) can therefore be considered as a measurement for the

inhibition concentration on the CHO and HT-29 cell line. The cytotoxicity on the EA.hy926 is

correlated with principal component 2 (PC2) (Figure 24 and Annex 3). The undifferentiated

Caco-2 cell line is correlated with both principal components. The factor scores of curcumin

and its analogues were calculated based on those 2 principal components and a hierarchical

cluster analysis was performed. Four major clusters were identified (Ward’s method, p<0.05

tested by 1-way Anova, see Annex 3 for dendrogram). The factor scores of curcumin and its

analogues are shown in Figure 26. The graph at the top left of Figure 26 includes the names

of the curcumin compounds. The following graphs show the substitution at R1 (top right), R2

(bottom left) and R3 (bottom right) position. The positions of these functional groups are

shown in Figure 27. For MVB54, SVD62 and SVD130 another chemical basic structure was

used. MVB54 only contains the functional groups R2 and R3 on one of its aromatic rings.

46

SVD62 carries 2 benzyl groups and SVD130 consists of a heterocyclic ring at the R1

position. Hence its R1 group was named isoxazole.

Figure 26. Factor scores based on the PCA of the cytotoxicity (MTT and SRB) data on the CHO, EA.hy926, undifferentiated Caco-2 and HT-29 cell line as shown in Figure 24. The factor scores were clustered by a hierarchical cluster analysis (Ward’s method, p<0.05). The graph at the top left displays the factor scores with the name of the chemical compounds. The substitution at R1 (top right), R2 (bottom left) and R3 (bottom right) position are included in order to graphically interpret the effect of certain substitutions. Figure 27 displays the position of the three functional groups. AcO = acetoxy, MeO = methoxy, OH = hydroxyl, H = hydrogen. Structures can be seen in Figure 7.

47

Figure 27. Positions of the R1, R2 and R3 functional groups used in Figure 26 and Figure 28. For MVB54, SVD62 and SVD130 a different basic structure is used. SVD130’s R1 functional group was named isoxazole. The functional groups themselves and their name can be seen in Figure 7, Figure 26 and Figure 28.

The factor scores are normalized values. A factor score of 0 therefore means that the

cytotoxicity of the compound is close to the average of all the compounds. A high factor

score for PC1 means that the inhibition concentration on the CHO and HT-29 cell line is high

and the cytotoxicity thus low in comparison to the other compounds. A high value for PC2

means that the IC value on the EA.hy926 cell line is low in comparison with the other

compounds (IC values on the EA.hy926 were negatively correlated with PC2 as seen in

Figure 24). Analogues situated in the second quadrant (factor score 1<0, factor score 2>0)

also show high toxicity on the undifferentiated Caco-2 cell line. It can be seen that SVD95

and SVD99 (the yellow cluster) score high for PC2 and low for PC1. This means that their IC

values were low when compared with the mean of the IC values. It can thus be concluded

that in general SVD95 and SVD99 showed highest cytotoxicity on the cell line included in the

PCA. BDC, SVD61 and SVD62 showed low cytotoxicity on the CHO, HT-29 and

undifferentiated Caco-2 cell line (a high IC value) whereas SVD62 showed high cytotoxicity

on the endothelial cell line. From Figure 7 and Figure 26 it can be seen that these 3

compounds do not contain nitrogen in the R1 functional group (also not in another functional

group). Curcumin is the only other compound without nitrogen, which is situated quite close

to the other 3. In general, the compounds without nitrogen did not result in high cytotoxicity.

Furthermore it can be seen that MVB59, MVB74, SVD66, SVD67, SVD90 and SVD130 (the

red cluster) show higher cytotoxicity than curcumin, MVB54, MVB78, SVD1, SVD2, SVD26,

SVD29 and SVD80. SVD99 (which had the highest toxicity) carries a cyclohexylamine at R1

position while this functional group is also situated close to the average of the cytotoxicity.

The same remark can be made for SVD95. The change in cytotoxicity may be attributed to

another structural property. When comparing the structures of the various compounds

48

(Figure 7 and Figure 26), it can be seen that BDC, SVD90, SVD95 and SVD99 are the 4

compounds with merely a hydroxyl on the aromatic ring. When this was combined with

nitrogen in the R1 functional group this led to higher cytotoxicity.

A second PCA and hierarchical cluster analysis was performed to link ROS generation

inside the cells with cytotoxicity on the same cell line. The first two principal components

were able to explain 74% of the total variance for the data on the EA.hy926 cell line. This is

considered to be a good value [101]. The scree plot (see Annex 4) was also used to obtain

the number of principal components that were extracted. Two principal components were

retained. PC1 is a measure for the intracellular ROS in the EA.hy.926 cell line both with and

without oxidative stressor. PC2 is a measure for cytotoxicity (see Annex 4). A high value for

PC2 relates to a high inhibition concentration and thus a low cytotoxicity. A standardized

factor score was then calculated for curcumin and its analogues based on principal

component 1 and 2. The result is shown in Figure 28. The hierarchical clustering analysis

identified three major clusters (see Annex 4 for dendrogram, p<0.05 tested by 1-way Anova).

The analogues in the first and fourth quadrant showed high ROS values in comparison with

the compounds in quadrant II and III. The upper half of the graph (quadrant I and II) shows

the compounds with a relatively low cytotoxicity while the lower half contains the more toxic

compounds. It can thus be seen that SVD62 and SVD99 (which formed the red cluster) were

most toxic on the EA.hy926 cell line (as was already seen in the first PCA analysis).

However, now we can see that a relatively high intracellular ROS concentration did not

always lead to a high cytotoxicity (compounds in quadrant I). Furthermore, also a high

cytotoxicity was not always correlated with high ROS values (compounds in quadrant III).

Cluster 1 (blue) contains the compounds that do not induce ROS in the endothelial cell line

combined with a low cytotoxicity on this cell line. Cluster 2 (green) contains the ones that are

able to induce intracellular ROS and show a relatively low cytotoxicity. Finally, the third

cluster (red) with SVD62 and SVD99 has a high cytotoxicity independently of whether the

compound can induce ROS or not.

49

Figure 28. Factor scores based on the PCA including cytotoxicity and ROS data on the EA.hy926 cell line. The factor scores were clustered by a hierarchical cluster analysis (Ward’s method, p<0.05). The graph at the top left displays the factor scores with the name of the chemical compounds. The substitution at R1 (top right), R2 (bottom left) and R3 (bottom right) position are included in order to graphically interpret the effect of certain substitutions. Figure 27 displays the position of the three functional groups. AcO = acetoxy, MeO = methoxy, OH = hydroxyl, H = hydrogen. Structures can be seen in Figure 7. Third, a principal component analysis was also performed on data for the cytotoxicity and

ROS assays of the CHO cell line. When all the results of the ROS assay (1 and 10 µM of

compound with and without H2O2) were used (as for the EA.hy926 cell line), interpretation of

the sense of the principal components was very difficult. This is due to the low correlation

between ROS at 1 and 10 µM (see Annex 5). Since the correlation was highest between 10

µM with and without H2O2 and between 1 µM with and without H2O2 two PCA’s were

performed for 1 and 10 µM (see Annex 5). The graphs indicate that compounds with the

50

possibility to increase intracellular ROS did not always have a high cytotoxicity. Furthermore,

also the compounds with a low cytotoxicity could either show a high intracellular ROS

increase or not. Correlations between a structural property and the ability to increase ROS

could not be detected.

51

5. Discussion Curcumin, a plant polyphenol extracted from the curcuma plant, has been shown to be a

potential agent for cancer prevention and progression in vitro and in vivo [9, 102]. Curcumin

is not often used in the treatment of cancer because it has a rather low bioavailability. This is

due to the low water solubility, poor absorption and rapid degradation and metabolisation of

curcumin [2]. Therefore, some research groups have tried to modify the curcumin molecular

structure in order to promote its potential beneficial effects [10, 11]. The low bioavailability of

curcumin may be improved by removing the unstable β-diketone moiety [3]. The use of

curcumin analogues gives opportunities in unravelling the unique structural character of

curcumin and may be used for further molecular design.

As shown in the cell viability and proliferation tests, curcumin was able to decrease the

viability of only the CHO ovarian cancer cell line to a large extent. For the other tested

hepatic, colon and endothelial cell lines (EA.hy926, differentiated and undifferentiated caco-

2, HT-29 and HepG2) the effect of curcumin on cell viability and proliferation was very small

or even non-existent. SVD61 was the only compound that was never able to show large

cytotoxicity with SRB and MTT values always higher than 80%. Curcumin, BDC, SVD61 and

SVD62 were the only tested compounds that did not contain nitrogen. BDC and SVD62 were

in most of the tested cell lines also not cytotoxic however decreases could be observed on

the undifferentiated Caco-2 and EA.hy926. In general these compounds without nitrogen

(Figure 29) were less toxic than the ones containing nitrogen. With the exception of

curcumin, they formed a cluster (Figure 26) in the principal component analysis including the

cytotoxicity (MTT and SRB) data on the CHO, HT-29, undifferentiated Caco-2 and EA.hy926

endothelial cell line. The substitution of the β−diketone moiety with a nitrogen functional

group may be beneficial in reducing cell viability of carcinoma cell lines.

Figure 29. Curcumin and analogues without nitrogen were in general less toxic than the analogues containing a nitrogen.

52

Compound SVD99 was the compound showing the highest cytotoxicity. MTT and SRB

values decreased with more than 50% for every tested cell line with the exception for the

differentiated Caco-2 intestinal cell line. SVD90, SVD95 and SVD130 (Figure 30) were 3

other compounds that exhibited high cytotoxicity among the various tested analogues as can

be seen in the PCA analysis in Figure 26. SVD130 is the only tested compound containing

an isoxazole. Other authors also found that an isoxazole analogue was more active than

curcumin itself in a breast cancer cell line [10]. Furthermore SVD90, SVD95 and SVD99

together with BDC are the only 4 compounds (out of the 19 tested compounds) that only

contain a hydroxyl group on the aryl. The combination of a β-enaminoketone analogue of

curcumin with an aryl with one hydroxyl group without methoxy may show improved

cytotoxic effects against cancer cells.

Figure 30. Most toxic curcumin analogues in vitro

SVD95, SVD2 and SVD29 are β-enaminoketone analogues of curcumin bearing an N-2-

butyl group. The substitutions on the aryl group can be seen in Figure 7. SVD2 and SVD29

showed very comparable cytotoxic effects in most of the tested cell lines (Figure 26) and

were in general not very toxic (MTT > 50% compared with blank for every cell line). An N-2-

butyl analogue could thus only lead to a higher cytotoxic capacity without methoxy or

acetoxy group present on the aryl.

SVD99, SVD66 and SVD67 all contain an N-cyclohexyl. SVD66 however has an additional

methoxy group next to the hydroxyl on the aryl while for SVD67 the hydroxyl is substituted

for an acetoxy group. SVD99 was always more cytotoxic. SVD66 and SVD67 were situated

closely together in the PCA analysis using the MTT and SRB data. Again, the N-cyclohexyl

analogues only improve cytotoxicity when they do not bear a methoxy or acetoxy group on

the aryl. The substitution of the hydroxyl by an acetoxy can also be seen between curcumin

and SVD61 and between SVD1 and SVD26. SVD61 was never really toxic while curcumin

was on the CHO cell line. SVD1 and SVD26 were mostly not cytotoxic. SVD26 however was

able to induce cytotoxicity in the HT-29 cell line and both were cytotoxic in the CHO cell line.

Based on these results, the effect of a substitution of a hydroxyl group on the aryl with an

53

acetoxy group is difficult to predict. SVD62 and SVD80 also both contain an acetoxy group

and showed effects in some cell lines. In general, the substitution of the hydroxyl group with

an acetoxy group did not lead to the most toxic compounds. It has been suggested that the

methoxy group on the aryl is associated with the potential NF-κB activity since

bisdemethoxycurcumin has the lowest potency to suppress TNF-induced NF-κB activation

followed by demethoxycurcumin and curcumin had the highest potency [73]. This suggests

the important role of the methoxy groups on the aryl. Other studies showed that

demethoxycurcumin had the highest antiproliferative activity against MCF-7 breast

carcinoma cells whereas bisdemethoxycurcumin was more active for cytotoxicity against

ovarian cancer [74, 75]. Furthermore, bisdemethoxycurcumin had the highest potency in

inhibiting human fibrosarcoma cells invasion followed by demethoxycurcumin and curcumin

[103]. Our results suggest that highest antiproliferative activity is achieved without methoxy

group. The hydroxyl group is furthermore necessary to express those high cytotoxic effects

since SVD80 (which only carries an acetoxy group on the aryl) showed average cytotoxic

effects (Figure 26). Finally, also the substitution of the β−diketone moiety with a nitrogen

functional group is necessary considering the lower anti-cancer activity of

bisdemethoxycurcumin in our in vitro study.

The differentiated Caco-2 cell is one of the most relevant in vitro models to study the

regulation and effects on intestinal epithelial cells [83]. Because of this, toxic effects on this

cell line should be minimal since we want to address cancer cells with curcumin and its

analogues. Curcumin did not induce stress or toxicity in the intestinal Caco-2 cells. Some

components were capable of inducing cytotoxicity. The effects however were small when

compared to the other tested cell lines. These small effects may also be due to biological or

technical variation rather than cytotoxic effects of the compounds since the cells had to be

maintained for 3 weeks while differentiating. The components that decreased the MTT value

until maximum 75% were MVB54, SVD29, SVD66, SVD99 and SVD130.

The EA.hy926 endothelial cell line is differently influenced by curcumin and the various

analogues since the variables containing information about the cytotoxicity on this cell line

(see PCA in Figure 24) are somewhat separated from the other cell lines. The

undifferentiated Caco-2 cell line forms a group close to the HT-29 cell line while the CHO

cell line is quite separated from HT-29 and Caco-2. The IC50 values are grouped together

with the HT-29 cell line while the IC80 values show a separate group. The EA.hy926 cell line

is a hybrid of HUVEC (human umbilical vein endothelial) cells with A549/8 human lung

carcinoma cells and is thus not fully of cancer origin. The HT-29 and undifferentiated Caco-2

54

cell lines are both from colon carcinomas while the CHO cells originated from Chinese

hamster ovary. The CHO cells were in general more susceptible for toxic effects by

curcumin and its analogues with lower IC80 values than the other cell lines. Hence the

separation of the IC80 values on the CHO cell line in the PCA analysis in Figure 24. The

EA.hy926 cells were in general less prone for toxic effect by curcumin and its analogues with

the exception of SVD62, SVD95, SVD99 and SVD130, which induced high toxicity. The

toxicity on colon cancer cell lines HT-29 and cCco-2 were intermediate. The cell line thus

plays an important role toward activity of the various compounds.

SVD90, SVD95, SVD99 and SVD130 had the highest antiproliferative effects in vitro while

their solubility in exposure medium was diverse. More specific, SVD99 and SVD130 still

showed crystals at 0.5 and 5 µM respectively, whereas SVD90 and SVD95 had a higher

solubility. They only showed crystals starting at 25 µM. A high solubility (curcumin for

example was highly soluble relative to the other compounds) did not lead to high cytotoxicity

toward the cancer cell lines. Although SVD99 showed very low solubility with crystals still

appearing at 0.5 µM, it was the only compound that could reduce all (except for the

differentiated Caco-2 cell line) MTT and SRB values to less than 50%. Furthermore,

curcumin showed relatively high solubility whereas it’s cytotoxicity was rather low in

comparison with the other compounds. On the other hand, when the solubility of the

compounds that showed medium or high cytotoxicity together with a low solubility, would be

higher, this could perhaps lead to an even higher cytotoxicity. The solubility and

bioavailability of low soluble hydrophobic agents could be improved by using polymer-based

nanoparticles [104] or emulsions [105]. Inert nanoparticles have been used to deliver

hydrophobic natural products [106]. The particles themselves were found to be not cytotoxic

in vitro and in vivo [107]. On top of increasing the solubility in aqueous solutions, the

incorporation of curcumin in nanoparticles resulted in an increased cellular uptake, an

increased potency to induce apoptosis in tumor cells and a higher activity in suppressing

NF-κB in vitro [104, 108]. Furthermore, a higher bioavailability and longer half-life was

shown in vivo in mice [104]. Liposomes (consisting of a hydrophilic aqueous core

surrounded with a phospholipid multi-layer [109]) and emulsomes (a core consisting of solid

fats surrounded by one or more phospholipid bilayers [109]) also improve the solubility of

curcumin [105, 110]. CurcuEmulsomes led to significantly prolonged biological activity in

hepatocellular carcinoma cells in vitro [110] whereas liposomal curcumin inhibited pancreatic

carcinoma growth and induced apoptosis in vitro in an equal or better way than curcumin

alone [105].

55

Highest antioxidant activities were seen on the DPPH radical scavenging assay for all the

compounds with a methoxy group next to a hydroxyl group on the aromatic system. When

this methoxy group was removed, a small antioxidant effect was still observed (BDC,

SVD90, SVD95 and SVD99). No significant antioxidant effect is seen when the aryl is not

carrying a hydroxyl group. This corroborates with the evidence from literature that

polyphenols inactivate radicals through the donation of an electron or hydrogen atom of the

hydroxyl [50]. Furthermore the methoxy group plays a major role in the antioxidant activity of

curcumin and its analogues since it leads towards an improved removal of the hydrogen

from the hydroxyl group [16]. Also other electron-donating groups at the ortho-position (such

as an ethoxy group) are able to improve the radical scavenging effect [53]. The substitution

of the hydroxyl group with an acetoxy group in analogues with (SVD26, SVD29, SVD61,

SVD62 and SVD67) or without (SVD80) methoxy group never led to antioxidant activity on

the DPPH assay. The FRAP values of these compounds were significantly increased,

however, only to a very low extent. It has been suggested that the DPPH radical site

appears to be difficult to access due to steric hindrance. This implies that the accessibility of

the radical position controls the neutralization reaction more than the chemical properties of

the antioxidant itself [54]. Therefore also a FRAP assay was performed. In the FRAP assay,

the components without significant radical scavenging activities on the DPPH assay,

significantly increased the trolox equivalent, these rises are however very small (trolox

equivalent < 20 µM at 500 µM of the compound). Intermediate antioxidant levels are again

seen for BDC, SVD90 and SVD95. A higher trolox equivalent is seen for SVD99 (which does

not contain a methoxy group). However this TE is still lower then every compound bearing a

methoxy on the aromatic ring. From the compounds containing a methoxy group next to the

hydroxyl group, MVB54 has the lowest antioxidant activity on the FRAP assay. This may be

attributed to the fact that MVB54 contains one aryl without functional groups. The N-allyl and

N-alkyl substitution on the β-diketone moiety did not alter the antioxidant effect a lot. MVB74

with an N-n-propyl functional group showed lower antioxidant effects on the DPPH when

compared with the FRAP. This might be due to the steric hindrance to access the DPPH

radical since this effect was smaller for MVB59, which has a smaller N-n-propyl functional

group. Due to the small changes between the DPPH and FRAP antioxidant assay, the

effects of the N-allyl and N-alkyl groups on the antioxidant capacity is difficult to predict.

These effects were however always smaller than the substitution of a methoxy or acetoxy

group.

A study by Youssef and El-Sherbeny [53] showed that N-alkyl piperidone curcumin

analogues with high radical scavenging effects (tested by the DPPH radical scavenging

assay) also showed high in vitro anti-tumor effects in cell lines from various cancer origins.

56

In our experiments a high antioxidant activity did not lead to high cytotoxic effects per se.

SVD90, SVD95 and SVD99 showed highest cytotoxic effects while lower antioxidant

capacities were seen for those compounds due to the deletion of the methoxy group.

SVD130, which also showed high cytotoxic effects, was a good antioxidant. The analogues

SVD26, SVD29, SVD61, SVD62, which showed low DPPH and FRAP values, were often not

able to induce toxicity or stress in the tested cell lines in vitro. However not every analogue

(for example SVD67) with a low antioxidant capacity had low cytotoxic effects. The activity of

various analogues towards the inhibition of NF-κB activation did not correlate with

antioxidant activity as concluded in a study by Weber et al. [79]. Since NF-κB plays a major

role in cancer [72], cytotoxicity of cancer cells by curcumin may thus not be correlated with

antioxidation but with the suppression of NF-κB activation. It has been shown that ROS are

needed for TNF-induced NF-κB activation [111]. Atsumi et al. [112] showed that curcumin

induced ROS inside both human gingival fibroblasts primary cells (normal) and human

submandibular gland carcinoma cells. In the normal cells a plateau of ROS was seen at 15

µM of curcumin, while ROS kept rising in the carcinoma cells up till a 50 µM of curcumin.

ROS levels were approximately tripled or quintupled at 15 µM in respectively the carcinoma

and normal cells. The ROS levels in the carcinoma cells were however more than double

than those in the human gingival fibroblasts at 50 µM. However, in a study by Sandur et al.

[113] curcumin was shown to produce ROS while TNF-mediated NF-κB activation was

inhibited. The authors suggested a concentration effect of ROS or the possibility that the

apoptotic effects of curcumin are mediated through ROS while NF-κB activation was

supressed in other ways. Since ROS is involved in apoptosis through the mitochondrial

intrinsic pathway [61] curcumin may induce apoptosis of tumor cells through this pathway

[65]. The insight that ROS can induce apoptotic cell dead leads to the question whether

elevated ROS levels resulted into cytotoxicity and whether constant or decreased cellular

ROS levels resulted in normal cell viability in our experiments. In our study, curcumin was

not able to induce elevated ROS levels inside an endothelial hybrid (EA.hy926) and Chinese

hamster ovary carcinoma cell line at 10 µM while cytotoxicity is clearly seen in the CHO cell

line. Elevated cellular ROS levels were seen for other compounds after overnight incubation.

This did not always result in cytotoxicity tested after three days. Furthermore, when

cytotoxicity could be observed, this did not always translate to increased ROS levels. In fact,

no general trend between cytotoxicity and induction of ROS was identified. The fact that

curcumin was not able to significantly increase ROS levels during these experiments might

be due to the exposure time. It may be that ROS levels were already back to normal since

other authors measured curcumin-induced ROS after 1 to 5 hours of incubation [112, 113].

In a study performed by Sandur et al. [73] the antiproliferative responses on various

57

carcinoma cell lines by curcumin, demethoxycurcumin, BDC and tetrahydrocurcumin were

shown to be independent of intracellular ROS generation and suppression of NF-κB. Our

experiments did not explore correlations with NF-κB. Future experiments could focus on

measuring NF-κB and caspase-3 activity since this caspase-3 effector executes the final

steps in mitochondrial apoptosis [61].

As stated in section 4.5, cellular uptake of the various compounds was assessed through

comparing both the retention time and DAD spectra of the standard solutions and cellular

extracts. Optimization of the extraction and analysis parameters for each of the analogues

was not performed due to time restrictions. Therefore, only a qualitative interpretation of the

results is possible. Since cellular uptake was assessed in the endothelial EA.hy926 cell line,

the principal component analysis on the ROS and cytotoxicity data of this cell line (see

Figure 28) was used to link cellular uptake and bioactivity as seen in Figure 31. The most

toxic compounds (SVD95, SVD99 and SVD130) were detected in the cell lysate. For SVD62

however, further optimization of the protocol is advised. From these results, it looks like

cellular uptake is necessary to induce toxic effects. SVD67 however could not be detected in

the CHO cell line (see Table 2) while toxic effects could be seen for this analogue on this cell

line (see Figure 14 and Figure 15). The cellular uptake study did not determine degradation

or metabolisation products. Hence, a possible explanation is that SVD67 was already

degraded or metabolized before analytical analyses were performed. Another explanation is

that SVD67 actually was incorporated in the cell but in a concentration under the limit of

detection. Furthermore, since curcumin is mainly incorporated in the cell membrane [114],

binding of SVD67 with components of the membrane could have led to a bad extraction.

A study by Kunwar et al. [114] showed cellular uptake by curcumin. Curcumin was mainly

detected in the cell membrane but also in the cytoplasm and nucleus. Furthermore, cellular

uptake was higher in carcinoma (from lymphoma and breast origin) cell lines than in

lymphocytes. This improved cellular uptake resulted in a higher cytotoxicity and may thus

indicate that curcumin or its analogues could target cancer cells. Curcumin was not detected

in the EA.hy926 cell line or in the differentiated Caco-2 cell line in this study whereas some

of the analogues were. Interestingly, only the nitrogen compounds were detected

intracellular. The compounds without nitrogen were not detected either because of problems

with the protocol (SVD61 and SVD62) or because there was no cellular uptake (curcumin

and BDC). The use of analogues may thus increase cellular uptake depending on the cell

line and give rise to an increased cytotoxicity. This was also seen for a

diarylidenylpiperidone analogue of curcumin [115].

58

Figure 31. Factor scores based on the PCA including cytotoxicity and ROS data on the EA.hy926 cell line as in Figure 28. Factor score 1 is a normalized measure for the intracellular ROS generation in the EA.hy.926 cell line (a high value for factor score 1 means a relatively high generation of intracellular ROS) both with and without oxidative stressor. PC2 is a measure for cytotoxicity (a high value for factor score 2 means a relatively low cytotoxicity on the EA.hy926 cell line) (see Annex 4). Structures can be seen in Figure 7.

59

6. Conclusion In general it can be concluded that the differentiated Caco-2 cell line (a relevant in vitro

model for intestinal enterocytes [83]) and the EA.hy926 endothelial cell line were not very

susceptible to curcumin and its analogues for stress or toxic effects. The Chinese hamster

ovary cell line was in general more prone to cytotoxic effects and the HT-29 and

undifferentiated Caco-2 colon carcinoma cells showed intermediate effects. Since the first

two cell lines are less cancerous than the other cells, curcumin and its analogues may target

cells with adenocarcinoma origin while normal cells are less affected. The analogues

bearing an N-alkyl functional group without a methoxy group next to the hydroxyl group on

the aryl moiety demonstrated the highest cytotoxic effects with SVD99>SVD95>SVD90

(Figure 32). The isoxazole derivative SVD130 also showed high cellular cytotoxic activity.

The non-nitrogen analogues curcumin, BDC, SVD61 and SVD62 were the compounds with

an under average cytotoxic effect. SVD62 however was able to induce cellular death in the

endothelial cell line.

Figure 32. The analogues bearing an N-alkyl functional group without a methoxy group next to the hydroxyl group on the aryl moiety, together with the isoxazole curcumin analogue, are most promising to be studied in vivo.

High antioxidant activities were observed on both the FRAP and DPPH radical scavenging

assay for the compounds with an ortho-methoxy and para-hydroxyl group. They inactivate

radicals through the donation of an electron or hydrogen atom of the phenol [50]. The

methoxy group plays an important role since it forms an intramolecular hydrogen bond with

the hydrogen atom of the phenol, leading towards an improved removal of the hydrogen

atom [16] and thus an improved stabilization of the radical. When this methoxy group was

removed, small antioxidant effects were still observed. The removal of the para-hydroxyl

group however resulted in no antioxidant activity.

60

The cytotoxicity effects in vitro were not associated with increased intracellular reactive

oxygen species. Although some authors [65] proved that curcumin induces apoptosis in

some tumor cells through ROS, other authors [73] showed that the antiproliferative effects of

curcumin, BDC, tetrahydrocurcumin and demethoxycurcumin were independent of

intracellular ROS. The mechanism of apoptosis and cytotoxicity in our study remains unclear

but may be the result of other pathways as described in section 1.4.3 and 1.4.4.

The solubility of the analogues was decreased relatively to curcumin. This decreased

solubility did however not lead to lower activity against tumor cells. SVD99, which showed

very low solubility, was the only compound that could reduce all (except for the differentiated

Caco-2 cell line) MTT and SRB values to less than 50%. Finally, altering the chemical

structure of curcumin by introducing an amine group led to the detection of the analogues

intracellular. This improved bioavailability can lead to an increased cytotoxicity. However,

detection of the compound in the cell lysate did not always lead to cytotoxic effects after 3

days of incubation.

61

7. Future developments Since NF-κB activity has been linked with cancer and curcumin in various publications, it

may be of interest to perform experiments measuring the activity of this nuclear factor in the

various cell lines and to link them with cytotoxicity on the same cell line. In our experiments

ROS generation was not correlated with cytotoxicity while curcumin may induce apoptosis

through caspase-3 and the mitochondrial pathway depending on the cell line and

concentration. Therefore, the measurement of caspase-3 may help to see if the

mitochondrial apoptotic pathway is involved in apoptosis in the CHO, Caco-2, EA.hy926 or

HT-29 cell line.

The nitrogen analogues with only a hydroxyl on the aryl group showed high toxic effects in

the tumor cells, while only limited effects were seen in the non-cancer cells. An in vivo study

may clarify if these analogues provide opportunities for future cancer treatment.

The bioavailability of the compounds assessed through high performance liquid

chromatography has to be optimized to determine intracellular uptake quantitatively.

Furthermore, formation of metabolites or degradation products should be determined as

well. It could be of interest to perform transport studies with differentiated Caco-2 cells to

predict uptake of the analogues or metabolites by enterocytes and transport of the

compounds to the basolateral compartment. By doing this, uptake in the blood stream in vivo

could be predicted and the activity of these metabolites could then be tested on a carcinoma

cell line (e.g. hepatocellular, renal, breast, ovarian, lung,...) to see if these metabolites still

show anti-cancer activity.

The results from these experiments may serve in quantitative structure-activity relationship

(QSAR) analysis. This QSAR will link the activity quantitatively with the molecular structure

and may then be used for further molecular design of curcumin analogues. This QSAR may

also enhance to comprehend the unique structural character and effects of curcumin itself.

62

References 1. Shishodia S, Chaturvedi MM, Aggarwal BB. Role of curcumin in cancer therapy. Curr Prob Cancer. 2007;31(4):243-305. 2. Anand P, Kunnumakkara AB, Newman RA, Aggarwal BB. Bioavailability of curcumin: Problems and promises. Mol Pharm. 2007;4(6):807-18. 3. Bukhari SNA, Jantan IB, Jasamai M, Ahmad W, Amjad MWB. Synthesis and Biological Evaluation of Curcumin Analogues. J Med Sci. 2013;13(7):501-13. 4. Shishodia S, Sethi G, Aggarwal BB. Curcumin: Getting back to the roots. Ann Ny Acad Sci. 2005;1056:206-17. 5. Bradford PG. Curcumin and obesity. Biofactors. 2013;39(1):78-87. 6. Zhang DW, Fu M, Gao SH, Liu JL. Curcumin and Diabetes: A Systematic Review. Evid-Based Compl Alt. 2013. 7. Panahi Y, Hosseini MS, Khalili N, Naimi E, Majeed M, Sahebkar A. Antioxidant and anti-inflammatory effects of curcuminoid-piperine combination in subjects with metabolic syndrome: A randomized controlled trial and an updated meta-analysis. Clinical nutrition. 2015. 8. Hishikawa N, Takahashi Y, Amakusa Y, Tanno Y, Tuji Y, Niwa H, et al. Effects of turmeric on Alzheimer's disease with behavioral and psychological symptoms of dementia. Ayu. 2012;33(4):499-504. 9. Shishodia S, Chaturvedi MM, Aggarwal BB. Role of curcumin in cancer therapy. Curr Probl Cancer. 2007;31(4):243-305. 10. Simoni D, Rizzi M, Rondanin R, Baruchello R, Marchetti P, Invidiata FP, et al. Antitumor effects of curcumin and structurally beta-diketone modified analogs on multidrug resistant cancer cells. Bioorg Med Chem Lett. 2008;18(2):845-9. 11. Liang G, Li XK, Chen L, Yang SL, Wu XD, Studer E, et al. Synthesis and anti-inflammatory activities of mono-carbonyl analogues of curcumin. Bioorg Med Chem Lett. 2008;18(4):1525-9. 12. Shen L, Ji HF. Theoretical study on physicochemical properties of curcumin. Spectrochim Acta A. 2007;67(3-4):619-23. 13. Manolova Y, Deneva V, Antonov L, Drakalska E, Momekova D, Lambov N. The effect of the water on the curcumin tautomerism: A quantitative approach. Spectrochim Acta A. 2014;132:815-20. 14. Aggarwal BB, Kumar A, Bharti AC. Anticancer potential of curcumin: Preclinical and clinical studies. Anticancer Res. 2003;23(1A):363-98. 15. Anand P, Sundaram C, Jhurani S, Kunnumakkara AB, Aggarwal BB. Curcumin and cancer: An "old-age" disease with an "age-old" solution. Cancer Lett. 2008;267(1):133-64. 16. Bairwa K, Grover J, Kania M, Jachak SM. Recent developments in chemistry and biology of curcumin analogues. Rsc Adv. 2014;4(27):13946-78. 17. Berginc K, Trontelj J, Basnet NS, Kristl A. Physiological barriers to the oral delivery of curcumin. Pharmazie. 2012;67(6):518-24. 18. Wang YJ, Pan MH, Cheng AL, Lin LI, Ho YS, Hsieh CY, et al. Stability of curcumin in buffer solutions and characterization of its degradation products. J Pharmaceut Biomed. 1997;15(12):1867-76. 19. Cheng AL, Hsu CH, Lin JK, Hsu MM, Ho YF, Shen TS, et al. Phase I clinical trial of curcumin, a chemopreventive agent, in patients with high-risk or pre-malignant lesions. Anticancer Research. 2001;21(4B):2895-900. 20. Ireson CR, Jones DJL, Orr S, Coughtrie MWH, Boocock DJ, Williams ML, et al. Metabolism of the cancer chemopreventive agent curcumin in human and rat intestine. Cancer Epidem Biomar. 2002;11(1):105-11. 21. Pan MH, Huang TM, Lin JK. Biotransformation of curcumin through reduction and glucuronidation in mice. Drug Metab Dispos. 1999;27(4):486-94. 22. Anand P, Thomas SG, Kunnumakkara AB, Sundaram C, Harikumar KB, Sung B, et al. Biological activities of curcumin and its analogues (Congeners) made by man and Mother Nature. Biochem Pharmacol. 2008;76(11):1590-611. 23. Goel A, Kunnumakkara AB, Aggarwal BB. Curcumin as "Curecumin": From kitchen to clinic. Biochem Pharmacol. 2008;75(4):787-809. 24. Shishodia S, Singh T, Chaturvedi MM. Modulation of transcription factors by curcumin. Adv Exp Med Biol. 2007;595:127-48. 25. Liu ZG. Molecular mechanism of TNF signaling and beyond. Cell Res. 2005;15(1):24-7. 26. Bharti AC, Donato N, Singh S, Aggarwal BB. Curcumin (diferuloylmethane) down-regulates the constitutive activation of nuclear factor-kappa B and I kappa B alpha kinase in human multiple myeloma cells, leading to suppression of proliferation and induction of apoptosis. Blood. 2003;101(3):1053-62. 27. Karin M, Liu ZG, Zandi E. AP-1 function and regulation. Curr Opin Cell Biol. 1997;9(2):240-6.

63

28. Ameyar M, Wisniewska M, Weitzman JB. A role for AP-1 in apoptosis: the case for and against. Biochimie. 2003;85(8):747-52. 29. Park CH, Lee JH, Yang CH. Curcumin derivatives inhibit the formation of Jun-Fos-DNA complex independently of their conserved cysteine residues. Journal of Biochemistry and Molecular Biology. 2005;38(4):474-80. 30. Aggarwal BB, Kumar A, Aggarwal MS, Shishodia S. Curcumin derived from turmeric (Curcuma longa): a spice for all seasons. Phytopharmaceuticals in Cancer Chemoprevention. 2005;23:351-87. 31. Evans JF, Kargman SL. Cancer and cyclooxygenase-2 (COX-2) inhibition. Curr Pharm Design. 2004;10(6):627-34. 32. Plummer SM, Holloway KA, Manson MM, Munks RJL, Kaptein A, Farrow S, et al. Inhibition of cyclo-oxygenase 2 expression in colon cells by the chemopreventive agent curcumin involves inhibition of NF-kappa B activation via the NIK/IKK signalling complex. Oncogene. 1999;18(44):6013-20. 33. Goel A, Boland CR, Chauhan DP. Specific inhibition of cyclooxygenase-2 (COX-2) expression by dietary curcumin in HT-29 human colon cancer cells. Cancer Lett. 2001;172(2):111-8. 34. Pan MH, Lin-Shiau SY, Lin JK. Comparative studies on the suppression of nitric oxide synthase by curcumin and its hydrogenated metabolites through down-regulation of I kappa B kinase and NF kappa B activation in macrophages. Biochem Pharmacol. 2000;60(11):1665-76. 35. Rockey DC, Chung JJ, McKee CM, Noble PW. Stimulation of inducible nitric oxide synthase in rat liver by hyaluronan fragments. Hepatology. 1998;27(1):86-92. 36. Camacho-Barquero L, Villegas I, Sanchez-Calvo JM, Talero E, Sanchez-Fidalgo S, Motilva V, et al. Curcumin, a Curcuma longa constituent, acts on MAPK p38 pathway modulating COX-2 and NOS expression in chronic experimental colitis. Int Immunopharmacol. 2007;7(3):333-42. 37. Yang FJ, Li XY, Sharma M, Sasaki CY, Longo DL, Lim B, et al. Linking beta-catenin to androgen-signaling pathway. J Biol Chem. 2002;277(13):11336-44. 38. Choi HY, Lim JE, Hong JH. Curcumin interrupts the interaction between the androgen receptor and Wnt/beta-catenin signaling pathway in LNCaP prostate cancer cells. Prostate Cancer P D. 2010;13(4):343-9. 39. Rusch V, Mendelsohn J, Dmitrovsky E. The epidermal growth factor receptor and its ligands as therapeutic targets in human tumors. Cytokine & growth factor reviews. 1996;7(2):133-41. 40. Nicholson RI, Gee JMW, Harper ME. EGFR and cancer prognosis. Eur J Cancer. 2001;37:S9-S15. 41. Chen A, Xu J, Johnson AC. Curcumin inhibits human colon cancer cell growth by suppressing gene expression of epidermal growth factor receptor through reducing the activity of the transcription factor Egr-1. Oncogene. 2006;25(2):278-87. 42. Binion DG, Otterson MF, Rafiee P. Curcumin inhibits VEGF-mediated angiogenesis in human intestinal microvascular endothelial cells through COX-2 and MAPK inhibition. Gut. 2008;57(11):1509-17. 43. Wu GF, Luo JC, Rana JS, Laham R, Sellke FW, Li J. Involvement of COX-2 in VEGF-induced angiogenesis via P38 and JNK pathways in vascular endothelial cells. Cardiovasc Res. 2006;69(2):512-9. 44. Schottelius AJG, Moldawer LL, Dinarello CA, Asadullah K, Sterry W, Edwards CK. Biology of tumor necrosis factor-alpha - implications for psoriasis. Exp Dermatol. 2004;13(4):193-222. 45. Pasparakis M, Alexopoulou L, Episkopou V, Kollias G. Immune and inflammatory responses in TNF alpha-deficient mice: a critical requirement for TNF alpha in the formation of primary B cell follicles, follicular dendritic cell networks and germinal centers, and in the maturation of the humoral immune response. The Journal of experimental medicine. 1996;184(4):1397-411. 46. Zhong Y, Yu WY, Feng H, Fan ZC, Li JF. Curcumin suppresses tumor necrosis factor-alpha-induced matrix metalloproteinase-2 expression and activity in rat vascular smooth muscle cells via the NF-kappa B pathway. Exp Ther Med. 2014;7(6):1653-8. 47. Cho JW, Lee KS, Kim CW. Curcumin attenuates the expression of IL-1 beta, IL-6, and TNF-alpha as well as cyclin E in TNF-alpha-treated HaCaT cells; NF-kappa B and MAPKs as potential upstream targets. Int J Mol Med. 2007;19(3):469-74. 48. Lobo V, Patil A, Phatak A, Chandra N. Free radicals, antioxidants and functional foods: Impact on human health. Pharmacognosy reviews. 2010;4(8):118-26. 49. Waris G, Ahsan H. Reactive oxygen species: role in the development of cancer and various chronic conditions. Journal of carcinogenesis. 2006;5(1):14. 50. Leopoldini M, Russo N, Toscano M. The molecular basis of working mechanism of natural polyphenolic antioxidants. Food Chem. 2011;125(2):288-306. 51. Ak T, Gulcin I. Antioxidant and radical scavenging properties of curcumin. Chem-Biol Interact. 2008;174(1):27-37.

64

52. Masuda T, Hidaka K, Shinohara A, Maekawa T, Takeda Y, Yamaguchi H. Chemical studies on antioxidant mechanism of curcuminoid: Analysis of radical reaction products from curcumin. J Agr Food Chem. 1999;47(1):71-7. 53. Youssef KM, El-Sherbeny MA. Synthesis and antitumor activity of some curcumin analogs. Arch Pharm. 2005;338(4):181-9. 54. Xie J, Schaich KM. Re-evaluation of the 2,2-Diphenyl-1-picrylhydrazyl Free Radical (DPPH) Assay for Antioxidant Activity. J Agr Food Chem. 2014;62(19):4251-60. 55. Benzie IFF, Strain JJ. The ferric reducing ability of plasma (FRAP) as a measure of ''antioxidant power'': The FRAP assay. Anal Biochem. 1996;239(1):70-6. 56. Simon HU, Haj-Yehia A, Levi-Schaffer F. Role of reactive oxygen species (ROS) in apoptosis induction. Apoptosis. 2000;5(5):415-8. 57. Bayir H. Reactive oxygen species. Crit Care Med. 2005;33(12):S498-S501. 58. Banerjee A, Kunwar A, Mishra B, Priyadarsini KI. Concentration dependent antioxidant/pro-oxidant activity of curcumin Studies from AAPH induced hemolysis of RBCs. Chem-Biol Interact. 2008;174(2):134-9. 59. Ahsan H, Parveen N, Khan NU, Hadi SM. Pro-oxidant, anti-oxidant and cleavage activities on DNA of curcumin and its derivatives demethoxycurcumin and bisdemethoxycurcumin. Chem-Biol Interact. 1999;121(2):161-75. 60. Gupta SC, Hevia D, Patchva S, Park B, Koh W, Aggarwal BB. Upsides and downsides of reactive oxygen species for cancer: the roles of reactive oxygen species in tumorigenesis, prevention, and therapy. Antioxidants & redox signaling. 2012;16(11):1295-322. 61. Circu ML, Aw TY. Reactive oxygen species, cellular redox systems, and apoptosis. Free Radical Bio Med. 2010;48(6):749-62. 62. Sak K. Cytotoxicity of dietary flavonoids on different human cancer types. Pharmacognosy reviews. 2014;8(16):122-46. 63. Cheng YY, Yang JS, Tsai SC, Liaw CC, Chung JG, Huang LJ, et al. The newly synthesized 2-(3-hydroxy-5-methoxypheny1)-6,7-methylenedioxyquinolin-4-one triggers cell apoptosis through induction of oxidative stress and upregulation of the p38 MAPK signaling pathway in HL-60 human leukemia cells. Oncol Rep. 2012;28(4):1482-90. 64. Sa G, Das T. Anti cancer effects of curcumin: cycle of life and death. Cell Div. 2008;3. 65. Bhaumik S, Anjum R, Rangaraj N, Pardhasaradhi BVV, Khar A. Curcumin mediated apoptosis in AK-5 tumor cells involves the production of reactive oxygen intermediates. Febs Lett. 1999;456(2):311-4. 66. Elledge SJ. Cell cycle checkpoints: Preventing an identity crisis. Science. 1996;274(5293):1664-72. 67. Jee SH, Shen SC, Tseng CR, Chiu HC, Kuo ML. Curcumin induces a p53-dependent apoptosis in human basal cell carcinoma cells. J Invest Dermatol. 1998;111(4):656-61. 68. Shehzad A, Lee J, Huh TL, Lee YS. Curcumin induces apoptosis in human colorectal carcinoma (HCT-15) cells by regulating expression of Prp4 and p53. Molecules and cells. 2013;35(6):526-32. 69. Han SS, Chung ST, Robertson DA, Ranjan D, Bondada S. Curcumin causes the growth arrest and apoptosis of B cell lymphoma by downregulation of egr-1, C-myc, Bcl-X-L, NF-kappa B, and p53. Clin Immunol. 1999;93(2):152-61. 70. Maheshwari RK, Singh AK, Gaddipati J, Srimal RC. Multiple biological activities of curcumin: a short review. Life sciences. 2006;78(18):2081-7. 71. Lin YG, Kunnumakkara AB, Nair A, Merritt WM, Han LY, Armaiz-Pena GN, et al. Curcumin inhibits tumor growth and angiogenesis in ovarian carcinoma by targeting the nuclear factor-kappaB pathway. Clin Cancer Res. 2007;13(11):3423-30. 72. Aggarwal BB. Nuclear factor-kappa-B: The enemy within. Cancer Cell. 2004;6(3):203-8. 73. Sandur SK, Pandey MK, Sung B, Ahn KS, Murakatni A, Sethi G, et al. Curcumin, demethtoxycurcumin, bisdemethoxycurcumin, tetrahydrocurcumin and turmerones differentially regulate anti-inflammatory and anti-proliferative responses through a ROS-independent mechanism. Carcinogenesis. 2007;28(8):1765-73. 74. Simon A, Allais DP, Duroux JL, Basly JP, Durand-Fontanier S, Delage C. Inhibitory effect of curcuminoids on MCF-7 cell proliferation and structure-activity relationships. Cancer Lett. 1998;129(1):111-6. 75. Syu WJ, Shen CC, Don MJ, Ou JC, Lee GH, Sun CM. Cytotoxicity of curcuminoids and some novel compounds from Curcuma zedoaria. J Nat Prod. 1998;61(12):1531-4. 76. Karthikeyan NS, Sathiyanarayanan KI, Aravindan PG, Giridharan P. Synthesis, crystal structure, and anticancer properties of cyclic monocarbonyl analogs of curcumin. Med Chem Res. 2011;20(1):81-7.

65

77. Subramaniam D, May R, Sureban SM, Lee KB, George R, Kuppusamy P, et al. Diphenyl difluoroketone: A curcumin derivative with potent in vivo anticancer activity. Cancer Res. 2008;68(6):1962-9. 78. Thomas SL, Zhong DS, Zhou W, Malik S, Liotta D, Snyder JP, et al. EF24, a novel curcumin analog, disrupts the microtubule cytoskeleton and inhibits HIF-1. Cell Cycle. 2008;7(15):2409-17. 79. Weber WM, Hunsaker LA, Roybal CN, Bobrovnikova-Marion EV, Abcouwer SF, Royer RE, et al. Activation of NF kappa B is inhibited by curcumin and related enones. Bioorgan Med Chem. 2006;14(7):2450-61. 80. Dutta S, Padhye S, Priyadarsini KI, Newton C. Antioxidant and antiproliferative activity of curcumin semicarbazone. Bioorg Med Chem Lett. 2005;15(11):2738-44. 81. ATCC. Caco-2 [Caco2] (ATCC® HTB-37™): American type culture collection; 2014 [cited 2015 April 11]. Available from: http://www.lgcstandards-atcc.org/products/all/HTB-37.aspx?geo_country=be#characteristics. 82. Pinto M, Robineleon S, Appay MD, Kedinger M, Triadou N, Dussaulx E, et al. Enterocyte-Like Differentiation and Polarization of the Human-Colon Carcinoma Cell-Line Caco-2 in Culture. Biol Cell. 1983;47(3):323-30. 83. Jumarie C, Malo C. Caco-2 Cells Cultured in Serum-Free Medium as a Model for the Study of Enterocytic Differentiation Invitro. J Cell Physiol. 1991;149(1):24-33. 84. Rousset M. The Human-Colon Carcinoma Cell-Lines Ht-29 and Caco-2 - 2 Invitro Models for the Study of Intestinal Differentiation. Biochimie. 1986;68(9):1035-40. 85. Lev-Ari S, Strier L, Kazanov D, Madar-Shapiro L, Dvory-Sobol H, Pinchuk I, et al. Celecoxib and curcumin synergistically inhibit the growth of colorectal cancer cells. Gastroenterology. 2005;128(4):A296-A. 86. Wilkening S, Stahl F, Bader A. Comparison of primary human hepatocytes and hepatoma cell line HEPG2 with regard to their biotransformation properties. Drug Metab Dispos. 2003;31(8):1035-42. 87. Edgell CJ, Haizlip JE, Bagnell CR, Packenham JP, Harrison P, Wilbourn B, et al. Endothelium specific Weibel-Palade bodies in a continuous human cell line, EA.hy926. In vitro cellular & developmental biology : journal of the Tissue Culture Association. 1990;26(12):1167-72. 88. Rondaij MG, Bierings R, Kragt A, van Mourik JA, Voorberg J. Dynamics and plasticity of Weibel-Palade bodies in endothelial cells. Arterioscl Throm Vas. 2006;26(5):1002-7. 89. ATCC. EA.hy926 (ATCC® CRL-2922™): American type culture collection; 2014 [cited 2015 April 11]. Available from: http://www.lgcstandards-atcc.org/products/all/CRL-2922.aspx?geo_country=be. 90. ATCC. CHO-K1 (ATCC® CCL-61™): American type culture collection; 2014 [cited 2015 April 11]. Available from: http://www.lgcstandards-atcc.org/products/all/CCL-61.aspx?geo_country=be#characteristics. 91. Yang M, Butler M. Effects of ammonia on CHO cell growth, erythropoietin production, and glycosylation. Biotechnol Bioeng. 2000;68(4):370-80. 92. Bisht S, Feldmann G, Soni S, Ravi R, Karikar C, Maitra A, et al. Polymeric nanoparticle-encapsulated curcumin ("nanocurcumin"): a novel strategy for human cancer therapy. J Nanobiotechnology. 2007;5:3. 93. Yu HL, Huang QR. Enhanced in vitro anti-cancer activity of curcumin encapsulated in hydrophobically modified starch. Food Chem. 2010;119(2):669-74. 94. Kakarala M, Brenner DE, Korkaya H, Cheng C, Tazi K, Ginestier C, et al. Targeting breast stem cells with the cancer preventive compounds curcumin and piperine. Breast Cancer Res Tr. 2010;122(3):777-85. 95. Yang C-L, Ma Y-G, Xue Y-X, Liu Y-Y, Xie H, Qiu G-R. Curcumin induces small cell lung cancer NCI-H446 cell apoptosis via the reactive oxygen species-mediated mitochondrial pathway and not the cell death receptor pathway. DNA and cell biology. 2012;31(2):139-50. 96. Voigt W. Sulforhodamine B assay and chemosensitivity. Methods in molecular medicine. 2005;110:39-48. 97. Liao HB, Dong WQ, Shi XJ, Liu HL, Yuan K. Analysis and comparison of the active components and antioxidant activities of extracts from Abelmoschus esculentus L. Pharmacognosy magazine. 2012;8(30):156-61. 98. Thaipong K, Boonprakob U, Crosby K, Cisneros-Zevallos L, Byrne DH. Comparison of ABTS, DPPH, FRAP, and ORAC assays for estimating antioxidant activity from guava fruit extracts. J Food Compos Anal. 2006;19(6-7):669-75. 99. Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. Journal of immunological methods. 1983;65(1-2):55-63.

66

100. Vichai V, Kirtikara K. Sulforhodamine B colorimetric assay for cytotoxicity screening. Nat Protoc. 2006;1(3):1112-6. 101. Beaumont R, Dracup C. An introduction to Principal Component Analysis & Factor Analysis Using SPSS 19 and R (psych package) 2012 [cited 2015 May 23]. Available from: http://www.floppybunny.org/robin/web/virtualclassroom/stats/statistics2/pca1.pdf. 102. Hong JH, Ahn KS, Bae E, Jeon SS, Choi HY. The effects of curcumin on the invasiveness of prostate cancer in vitro and in vivo. Prostate Cancer P D. 2006;9(2):147-52. 103. Yodkeeree S, Chaiwangyen W, Garbisa S, Limtrakul P. Curcumin, demethoxycurcumin and bisdemethoxycurcumin differentially inhibit cancer cell invasion through the down-regulation of MMPs and uPA. J Nutr Biochem. 2009;20(2):87-95. 104. Anand P, Nair HB, Sung BK, Kunnumakkara AB, Yadav VR, Tekmal RR, et al. Design of curcumin-loaded PLGA nanoparticles formulation with enhanced cellular uptake, and increased bioactivity in vitro and superior bioavailability in vivo. Biochem Pharmacol. 2010;79(3):330-8. 105. Li L, Braiteh FS, Kurzrock R. Liposome-encapsulated curcumin - In vitro and in vivo effects on proliferation, apoptosis, signaling, and angiogenesis. Cancer. 2005;104(6):1322-31. 106. Bala I, Bhardwaj V, Hariharan S, Kharade SV, Roy N, Kumar MNVR. Sustained release nanoparticulate formulation containing antioxidant-ellagic acid as potential prophylaxis system for oral administration. J Drug Target. 2006;14(1):27-34. 107. Bisht S, Mizuma M, Feldmann G, Ottenhof NA, Hong SM, Pramanik D, et al. Systemic Administration of Polymeric Nanoparticle-Encapsulated Curcumin (NanoCurc) Blocks Tumor Growth and Metastases in Preclinical Models of Pancreatic Cancer. Mol Cancer Ther. 2010;9(8):2255-64. 108. Chuah LH, Roberts CJ, Billa N, Abdullah S, Rosli R. Cellular uptake and anticancer effects of mucoadhesive curcumin-containing chitosan nanoparticles. Colloid Surface B. 2014;116:228-36. 109. Amselem A, Zawoznik E, Yogev A, Friedman D. Emulsomes, a new type of lipid assembly. Handbook of Nonmedical Applications of Liposomes. 1996;3:209-23. 110. Ucisik MH, Kupcu S, Schuster B, Sleytr UB. Characterization of CurcuEmulsomes: nanoformulation for enhanced solubility and delivery of curcumin. J Nanobiotechnol. 2013;11. 111. Garg AK, Aggarwal BB. Reactive oxygen intermediates in TNF signaling. Mol Immunol. 2002;39(9):509-17. 112. Atsumi T, Tonosaki K, Fujisawa S. Induction of early apoptosis and ROS-generation activity in human gingival fibroblasts (HGF) and human submandibular gland carcinoma (HSG) cells treated with curcumin. Arch Oral Biol. 2006;51(10):913-21. 113. Sandur SK, Ichikawa H, Pandey MK, Kunnumakkara AB, Sung B, Sethi G, et al. Role of pro-oxidants and antioxidants in the anti-inflammatory and apoptotic effects of curcumin (diferuloylmethane). Free Radical Bio Med. 2007;43(4):568-80. 114. Kunwar A, Barik A, Mishra B, Rathinasamy K, Pandey R, Priyadarsini KI. Quantitative cellular uptake, localization and cytotoxicity of curcumin in normal and tumor cells. Bba-Gen Subjects. 2008;1780(4):673-9. 115. Dayton A, Selvendiran K, Kuppusamy ML, Rivera BK, Meduru S, Kalai T, et al. Cellular uptake, retention and bioabsorption of HO-3867, a fluorinated curcumin analog with potential antitumor properties. Cancer Biol Ther. 2010;10(10):1027-32.

67

Annex

7.1. Annex 1. Chromatograms and DAD spectra

Figure 33. Chromatogram of curcumin standard solution (top), differentiated caco-2 (middle) and endothelial cell line (bottom) extract.

Figure 34. Chromatogram of BDC standard solution (top) and endothelial cell line (bottom) extract.

Figure 35. Chromatogram of MVB54 standard solution (top) and endothelial cell line (bottom) extract.

Figure 36. DAD spectra of MVB54 standard solution (left) and endothelial cell line (right) extract.

18,00 18,50 19,00 19,50 20,00 20,50 21,00 21,50 22,00 22,50 23,00 23,50 24,00-10,0

-5,0

0,0

5,0

10,0

15,0

20,0

25,0

1 - 20150429 UROLITHIN EXTRACTION + TRANSPORT #66 Endo curc UV_VIS_42 - 20150429 UROLITHIN EXTRACTION + TRANSPORT #67 Caco curc UV_VIS_43 - 20150429 UROLITHIN EXTRACTION + TRANSPORT #63 Curcumine UV_VIS_4mAU

min

3

2

1

WVL:360 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-20

25

50

75

100

125

160

1 - 20032015Sandercurcumin #66 endo BDC UV_VIS_12 - 20032015Sandercurcumin #91 BDC UV_VIS_1mAU

min

2

1

WVL:410 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-20

0

20

40

60

80

100

1 - 20032015Sandercurcumin #67 endo mvb54 UV_VIS_12 - 20032015Sandercurcumin #92 MVB54 UV_VIS_1mAU

min

2

1

WVL:410 nm

Peak #40 100% at 3.49 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

391.4

50% at 3.45 min: 919.73 -50% at 3.55 min: 906.04

Peak #36 100% at 3.49 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

446.4

50% at 3.35 min: 152.54 -50% at 3.55 min: 998.12

68

Figure 37. Chromatogram of MVB59 standard solution (top) and endothelial cell line (bottom) extract.

Figure 38. Chromatogram of MVB74 standard solution (top) and endothelial cell line (bottom) extract.

Figure 39. Chromatogram of MVB78 standard solution (top) and endothelial cell line (bottom) extract.

Figure 40. Chromatogram of SVD1 standard solution (top) and endothelial cell line (bottom) extract.

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-10,0

0,0

10,0

20,0

30,0

40,0

1 - 20032015Sandercurcumin #68 endo mvb59 UV_VIS_12 - 20032015Sandercurcumin #93 MVB59 UV_VIS_1mAU

min

2

1

WVL:410 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-10,0

0,0

10,0

20,0

30,0

1 - 20032015Sandercurcumin #69 endo mvb74 UV_VIS_12 - 20032015Sandercurcumin #94 MVB74 UV_VIS_1mAU

min

2

1

WVL:410 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-20,0

0,0

12,5

25,0

37,5

50,0

70,0

1 - 20032015Sandercurcumin #70 endo mvb78 UV_VIS_12 - 20032015Sandercurcumin #89 mvb78 25 µM UV_VIS_1mAU

min

2

1

WVL:410 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-10,0

0,0

10,0

20,0

30,0

40,0

1 - 20032015Sandercurcumin #71 endo svd1 UV_VIS_12 - 20032015Sandercurcumin #95 SVD1 UV_VIS_1mAU

min

2

1

WVL:410 nm

69

Figure 41. Chromatogram of SVD2 standard solution (top) and endothelial cell line (bottom) extract.

Figure 42. Chromatogram of SVD26 standard solution (top), CHO (second from top) differentiated caco-2 (second from bottom) and endothelial (bottom) cell line extract.

Figure 43. DAD spectra of SVD26 standard (top left), CHO (top right), differentiated caco-2 (bottom left) and endothelial (bottom right) cell line extract.

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-30

0

20

40

60

80

100

1 - 20032015Sandercurcumin #72 endo svd2 UV_VIS_12 - 20032015Sandercurcumin #96 SVD2 UV_VIS_1mAU

min

2

1

WVL:410 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-60

0

50

100

150

200

250

1 - 20032015Sandercurcumin #31 endo svd26 cells UV_VIS_12 - 20032015Sandercurcumin #47 caco svd26 cells UV_VIS_13 - 20032015Sandercurcumin #81 cho svd26 UV_VIS_14 - 20032015Sandercurcumin #59 SVD26 UV_VIS_1mAU

min

4

3

2

1

WVL:410 nm

Peak #35 100% at 3.74 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

384.7

50% at 3.70 min: 909.97 -50% at 3.78 min: 916.15

Peak #25 100% at 3.76 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

410.1

50% at 3.71 min: 956.57 -50% at 3.81 min: 968.52

Peak #26 100% at 3.76 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

442.2

50% at 3.72 min: 996.74 -50% at 3.81 min: 997.05

Peak #29 100% at 3.77 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

443.6

50% at 3.73 min: 996.98 -50% at 3.82 min: 997.66

70

Figure 44. Chromatogram of SVD29 standard solution (top) and endothelial cell line (bottom) extract.

Figure 45. DAD spectra of SVD29 standard solution (left) and endothelial cell line (right) extract.

Figure 46. Chromatogram of SVD62 standard solution (top) and endothelial cell line (bottom) extract.

Figure 47. Chromatogram of SVD66 standard solution (top), CHO (second from top) differentiated caco-2 (second from bottom) and endothelial (bottom) cell line extract.

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-40

0

50

100

160

1 - 20032015Sandercurcumin #73 endo svd29 UV_VIS_12 - 20032015Sandercurcumin #97 SVD29 UV_VIS_1mAU

min

2

1

WVL:410 nm

Peak #36 100% at 3.48 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

379.7

50% at 3.44 min: 948.34 -50% at 3.52 min: 914.61

Peak #25 100% at 3.50 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

424.2

50% at 3.43 min: 991.75 -50% at 3.55 min: 979.10

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-6,0

0,0

5,0

10,0

16,0

1 - 20032015Sandercurcumin #74 endo svd62 UV_VIS_42 - 20032015Sandercurcumin #90 svd62 UV_VIS_4mAU

min

2

1

WVL:390 nm

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-40

0

25

50

75

100

120

1 - 20032015Sandercurcumin #35 endo svd66 cells UV_VIS_12 - 20032015Sandercurcumin #51 caco svd66 cells UV_VIS_13 - 20032015Sandercurcumin #83 cho svd66 UV_VIS_14 - 20032015Sandercurcumin #61 SVD66 UV_VIS_1mAU

min

4

3

2

1

WVL:410 nm

71

Figure 48. DAD spectra of SVD66 standard solution (top left), CHO (top right), differentiated caco-2 (bottom left) and endothelial (bottom right) cell line extract.

Figure 49. Chromatogram of SVD66 standard solution (top), CHO (second from top), differentiated caco-2 (second from bottom) and endothelial (bottom) cell line extract.

Figure 50. DAD spectra of SVD67 standard solution (left) and endothelial cell line (right) extract.

Peak #35 100% at 3.54 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

396.8

50% at 3.50 min: 928.29 -50% at 3.62 min: 912.60

Peak #26 100% at 3.58 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

455.0

50% at 3.50 min: 999.76 -50% at 3.69 min: 999.51

Peak #26 100% at 3.56 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

455.1

50% at 3.51 min: 999.54 -50% at 3.63 min: 999.52

Peak #36 100% at 3.50 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

429.6

50% at 3.45 min: 997.67 -50% at 3.57 min: 978.71

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-100

0

100

200

300

1 - 20032015Sandercurcumin #37 endo svd67 cells UV_VIS_12 - 20032015Sandercurcumin #53 caco svd67 cells UV_VIS_13 - 20032015Sandercurcumin #84 cho svd67 UV_VIS_14 - 20032015Sandercurcumin #62 SVD67 UV_VIS_1mAU

min

4

3

2

1

WVL:410 nm

Peak #38 100% at 3.82 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

384.4

50% at 3.78 min: 895.12 -50% at 3.86 min: 927.76

Peak #42 100% at 3.84 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

432.8

50% at 3.79 min: 991.07 -50% at 3.89 min: 993.10

72

Figure 51. Chromatogram of SVD80 standard solution (top) and endothelial cell line (bottom) extract.

Figure 52. DAD spectra of SVD80 standard solution peak 1 (top left), peak 2 (top right), peak 3 (middle left) and endothelial cell line extract peak 1 (middle right) and peak 2 (bottom right).

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-15,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

1 - 20032015Sandercurcumin #75 endo svd80 UV_VIS_22 - 20032015Sandercurcumin #99 SVD80 UV_VIS_2mAU

min

2

1

WVL:450 nm

Peak #35 100% at 3.26 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

384.1

50% at 3.22 min: 949.77 -50% at 3.30 min: 943.98

Peak #36 100% at 3.45 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

383.5

50% at 3.40 min: 941.88 -50% at 3.49 min: 934.22

Peak #37 100% at 3.63 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

371.6

50% at 3.60 min: 999.41 -50% at 3.66 min: 998.91

Peak #29 100% at 3.29 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

428.9

50% at 3.25 min: 991.35 -50% at 3.34 min: 988.22

Peak #30 100% at 3.47 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

433.8

50% at 3.43 min: 994.27 -50% at 3.52 min: 992.65

73

Figure 53. Chromatogram of SVD90 standard solution (top) and endothelial cell line (bottom) extract.

Figure 54. DAD spectra of SVD90 standard solution (left) and endothelial cell line (right) extract.

Figure 55. Chromatogram of SVD95 standard solution (top), CHO (second from top) differentiated caco-2 (second from bottom) and endothelial cell line (bottom) extract

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-40

0

50

100

160

1 - 20032015Sandercurcumin #76 endo svd90 UV_VIS_12 - 20032015Sandercurcumin #100 SVD90 UV_VIS_1mAU

min

2

1

WVL:410 nm

Peak #31 100% at 3.26 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

373.0

50% at 3.22 min: 999.74 -50% at 3.30 min: 999.82

Peak #29 100% at 3.30 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

384.6

50% at 3.26 min: 951.14 -50% at 3.34 min: 882.78

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-50

0

50

100

150

200

250

1 - 20032015Sandercurcumin #39 endo svd95 cells UV_VIS_12 - 20032015Sandercurcumin #55 caco svd95 cells UV_VIS_13 - 20032015Sandercurcumin #85 cho svd95 UV_VIS_14 - 20032015Sandercurcumin #63 SVD95 UV_VIS_1mAU

min

4

3

2

1

WVL:410 nm

74

Figure 56. DAD spectra of SVD95 standard solution (top left), CHO (top right) and endothelial (bottom) cell line extract.

Figure 57. Chromatogram of SVD99 standard solution (top) and endothelial cell line (bottom) extract.

Figure 58. DAD spectra of SVD99 standard solution (left) and endothelial cell line (right) extract.

Peak #28 100% at 3.11 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

376.3

50% at 3.07 min: 992.47 -50% at 3.14 min: 988.22

Peak #27 100% at 3.13 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

432.1

50% at 3.07 min: 999.07 -50% at 3.19 min: 992.32

Peak #22 100% at 3.10 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

359.9

50% at 3.06 min: 947.70 -50% at 3.14 min: 953.95

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-50

0

50

100

150

200

250

1 - 20032015Sandercurcumin #77 endo svd99 UV_VIS_12 - 20032015Sandercurcumin #101 SVD99 UV_VIS_1mAU

min

2

1

WVL:410 nm

Peak #33 100% at 3.33 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

379.9

50% at 3.29 min: 972.41 -50% at 3.37 min: 960.34

Peak #32 100% at 3.36 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

420.6

50% at 3.31 min: 992.18 -50% at 3.40 min: 974.81

75

Figure 59. Chromatogram of SVD130 standard solution (top) and endothelial cell line (bottom) extract.

Figure 60. DAD spectra of SVD130 standard solution (left) and endothelial cell line (right) extract.

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0 9,5 10,0-4,0

0,0

2,5

5,0

7,5

10,0

12,0

1 - 20032015Sandercurcumin #78 endo svd130 UV_VIS_42 - 20032015Sandercurcumin #102 SVD130 UV_VIS_4mAU

min

2

1

WVL:390 nm

Peak #47 100% at 4.39 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

346.7

50% at 4.33 min: 941.37 -50% at 4.51 min: 860.62

Peak #32 100% at 4.37 min

-10,0

0,0

10,0

20,0

30,0

40,0

50,0

60,0

190 250 300 350 400 450 500 550 600 650 700 790

%

nm

365.5

50% at 4.31 min: 992.09 -50% at 4.45 min: 991.67

76

7.2. Annex 2. Cellular uptake assessed by HPLC Table 3. Assessment of cellular uptake of curcuminoids by reversed phase HPLC with DAD detection. The DAD spectra as well as the retention time were used to evaluate cellular uptake.

Standard (25 µM) EA.hy926 cell line CHO cell line Differentiated caco-2 cell line

tR (min) Abs area (mAU*t)

tR Abs area DAD** Uptake

tR Abs area DAD Uptake

tR Abs area DAD Uptake

(min) (mAU*t) (min) (mAU*t) (min) (mAU*t) Curcumin 21.09 1.24 /* / / N* NT* NT NT NT / / / N BDC 3.326 11.76 / / / N NT NT NT NT NT NT NT NT MVB54 3.493 11.49 3.486 1.38 N Y* NT NT NT NT NT NT NT NT MVB59 3.466 6.92 / / / N NT NT NT NT NT NT NT NT MVB74 3.593 3.59 / / / N NT NT NT NT NT NT NT NT MVB78 3.317 9.65 / / / N NT NT NT NT NT NT NT NT SVD1 3.593 7.56 / / / N NT NT NT NT NT NT NT NT SVD2 3.275 15.28 / / / N NT NT NT NT NT NT NT NT SVD26 3.741 12.95 3.771 0.83 N Y 3.756 5.61 N Y 3.764 0.91 N Y SVD29 3.478 14.67 3.496 0.46 N Y NT NT NT NT NT NT NT NT SVD61 ? ? ? ? ? ? ? ? ? ? ? ? ? ? SVD62 (1)* 3.055 0.64 ? ? ? ? NT NT NT NT NT NT NT NT SVD62 (2)* 3.188 0.72 ? ? ? ? NT NT NT NT NT NT NT NT SVD62 (3)* 3.489 0.82 ? ? ? ? NT NT NT NT NT NT NT NT SVD66 3.545 11.02 3.5 0.26 Y Y 3.583 0.78 Y Y 3.558 0.60 Y Y SVD67 3.82 17.47 3.844 1.54 N Y / / / N / / / N SVD80 (1) 3.259 3.26 3.293 0.84 N Y NT NT NT NT NT NT NT NT SVD80 (2) 3.446 5.71 3.474 0.98 N Y NT NT NT NT NT NT NT NT SVD80 (3) 3.627 0.77 / / / / NT NT NT NT NT NT NT NT SVD90 3.259 11.46 3.298 2.40 Y Y NT NT NT NT NT NT NT NT SVD95 3.106 13.26 3.097 1.06 Y Y 3.134 1.34 N Y / / / / SVD99 3.326 19.43 3.355 0.52 N Y NT NT NT NT NT NT NT NT SVD130 4.394 0.60 4.369 2.41 Y Y NT NT NT NT NT NT NT NT

* / = No peak at same retention time as standard, N = No, Y= Yes, NT= not tested, ? = results not sufficient to asses cellular uptake, (1) = first peak, (2) = second peak, (3) = third peak ** DAD spectra of cellular extract compared with standard. Y = similar DAD spectra, N = different DAD spectra

77

7.3. Annex 3. PCA analysis on MTT and SRB data

Table 4. Pattern (left) and structure (right) matrix of PCA analysis based on MTT and SRB data on the EA.hy926, HT-29, undifferentiated caco-2 and CHO cell line

Pattern Matrix

Structure Matrix

Component

Component

1 2

1 2 CHO_50 0,663 -0,035

CHO_50 0,672 -0,199

CHO_80 0,864 0,506

CHO_80 0,739 0,293 ATCC_50 0,458 -0,49

ATCC_50 0,579 -0,602

ATCC_80 0,278 -0,662

ATCC_80 0,441 -0,731 HT_50 0,564 -0,262

HT_50 0,629 -0,401

HT_80 0,806 -0,087

HT_80 0,828 -0,286 EA_50 0,003 -0,813

EA_50 0,204 -0,814

EA_80 -0,02 -0,915

EA_80 0,206 -0,91 S_CHO_50 0,603 -0,29

S_CHO_50 0,675 -0,439

S_CHO_80 0,889 0,383

S_CHO_80 0,795 0,164 S_ATCC_50 0,521 -0,46

S_ATCC_50 0,634 -0,589

S_ATCC_80 0,493 -0,447

S_ATCC_80 0,603 -0,568 S_HT_50 0,696 -0,042

S_HT_50 0,706 -0,213

S_HT_80 0,72 -0,141

S_HT_80 0,755 -0,318 S_EA_50 -0,219 -0,752

S_EA_50 -0,033 -0,698

S_EA_80 0,19 -0,627

S_EA_80 0,345 -0,674

Figure 61. Cluster analysis based on Ward’s method based on MTT and SRB data on the EA.hy926, HT-29, undifferentiated caco-2 and CHO cell line

78

7.4. Annex 4. PCA analysis on ROS and cytotoxicity data on EA.hy926 cells

Figure 62. Scree plot based on the IC50 and IC80 values of the MTT and SRB assay and on the ROS assay on the EA.hy926 cell line.

Table 5. Pattern (left) and structure (right) matrix of PCA analysis based on MTT, SRB and ROS data on the EA.hy926,

Pattern Matrixa

Structure Matrix

Component

Component

1 2

1 2 ROS_EN_10_OX 0,965 0,126

ROS_EN_10_OX 0,942 -0,05

ROS_EN_1_OX 0,907 0,162

ROS_EN_1_OX 0,877 -0,003

ROS_EN_1 0,816 -0,158

ROS_EN_1 0,845 -0,307 ROS_EN_10 0,776 -0,213

ROS_EN_10 0,815 -0,355

EA_50 0,032 0,912

EA_50 -0,134 0,906 S_EA_50 0,053 0,871

S_EA_50 -0,105 0,861

EA_80 -0,01 0,839

EA_80 -0,163 0,841 S_EA_80 -0,083 0,715

S_EA_80 -0,213 0,73

Figure 63. PCA plot based on ROS and cytotoxicity data on EA.hy926 cells

79

Figure 64. Cluster analysis based on Ward’s method based on MTT, SRB and ROS data on the EA.hy926 cell line.

7.5. Annex 5. PCA analysis on ROS and cytotoxicity data on CHO cells Table 6. Correlation matrix of results of ROS assay on CHO cell line

Table 7. Pattern (left) and structure (right) matrix of PCA analysis based on MTT, SRB and ROS data (at 1 µM) on the CHO cell line.

Pattern Matrix

Structure Matrix

Component

Component

1 2

1 2

S_CHO_80 0,918 -0,194

S_CHO_80 0,939 -0,294 CHO_50 0,875 0,11

CHO_80 0,874 -0,349

CHO_80 0,846 -0,257

CHO_50 0,863 0,015 S_CHO_50 0,666 0,174

S_CHO_50 0,647 0,102

ROS_CHO_1_OX -0,216 0,852

ROS_CHO_1_OX -0,309 0,876

ROS_CHO_1 0,141 0,781

ROS_CHO_1 0,056 0,765

ROS_CHO_

1 ROS_CHO_1

0 ROS_CHO_1_O

X ROS_CHO_10_O

X

Correlation ROS_CHO_1 1,000 ,282 ,419 ,040

ROS_CHO_10 ,282 1,000 -,037 ,864

ROS_CHO_1_OX ,419 -,037 1,000 -,226

ROS_CHO_10_OX

,040 ,864 -,226 1,000

80

Figure 65. Cluster analysis based on Ward’s method based on MTT, SRB and ROS (at 1 µM) data on the CHO cell line

Figure 66. Factor scores based on PCA analysis of ROS (at 1 µM), MTT and SRB data on CHO cell line (p>0.05 tested by 1-way Anova for clustering).

Table 8. Pattern (left) and structure (right) matrix of PCA analysis based on MTT, SRB and ROS data (at 10 µM) on the CHO cell line.

Pattern Matrix

Structure Matrix

Component

Component

1 2

1 2

S_CHO_80 0,92 0,192

S_CHO_80 0,935 0,262 CHO_50 0,866 -0,009

CHO_50 0,865 0,057

81

CHO_80 0,846 0,24

CHO_80 0,865 0,305 S_CHO_50 0,707 -0,417

S_CHO_50 0,676 -0,363

ROS_CHO_10_OX 0,011 0,959

ROS_CHO_10_OX 0,085 0,96

ROS_CHO_10 0,115 0,914

ROS_CHO_10 0,184 0,923

Figure 67. Factor scores based on PCA analysis of ROS (at 10 µM), MTT and SRB data on CHO cell line (p>0.05 tested by 1-way Anova for clustering).

Figure 68. Factor scores based on PCA analysis of ROS (at 10 µM), MTT and SRB data on CHO cell line.