Cyclotides from an Extreme Habitat: Characterization of Cyclic Peptides from Viola Abyssinica of Ethiopian Highlands

To my late grandmother

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Yeshak, MY., Burman, R., Eriksson, C., Göransson, U. Opti- mization of Cyclotide Extraction Parameters. Submitted to Phy- tochemistry Letters.

II Yeshak, MY., Burman, R., Asres, K., Göransson, U. (2011) Cyclotides from an Extreme Habitat: Characterization of Cyclic Peptides from Viola abyssinica of Ethiopian Highlands. Journal of Natural Products, 74:727-731. Reprinted with permission. Copyright © 2011 The American Chemical Society and Ameri- can Society of Pharmacognosy.

III Yeshak, MY., Strömstedt, A., Burman, R., Göransson, U., Cy- clotide stability in Pseudomonas aeruginosa and Staphylococ- cus aureus Secreted Proteases. In manuscript.

IV Yeshak, MY., Göransson, U., Burman, R., Hellman, B., Geno- toxicity and Cellular Uptake of Cyclotides: Evidence for Multi- ple Mode of Action. Accepted. Journal of Mutation Research - Genetic Toxicology and Environmental Mutagenesis.

Contents

Introduction ...... 11 Natural products ...... 11 Cyclotides ...... 12 The cyclotide framework...... 13 Natural sources and biosynthesis...... 15 Biological activities ...... 16 Potential applications of cyclotides ...... 17 Aims of this study...... 18 Biochemical studies on cyclotides ...... 19 Extraction and isolation...... 19 Small-scale extraction...... 19 Large-scale extraction...... 22 Identification and sequence determination...... 27 Mass spectrometry ...... 27 Ambiguities to solve ...... 29 Amino acid analysis...... 32 Quantitative analysis...... 33 Stability in proteases...... 34 Cyclotides and cells...... 38 Cytotoxicity ...... 38 Genotoxicity ...... 40 Cell permeation...... 44 Radiolabeling ...... 45 Microautoradiographic analysis...... 45 Discussion ...... 50 Cyclotide production ...... 50 Nature as a source of novel templates ...... 51 Diverse sequences...... 51 Lessons from the knot...... 53 Multiple modes of action...... 54 Concluding remarks and future perspectives ...... 56 Popular science summary...... 58

አጭር መግለጫ...... 60 Acknowledgements ...... 66 References ...... 70 Appendices...... 80 Appendix I ...... 80 Appendix II...... 81

Abbreviations

3D three dimensional AAA (quantitative) amino acid analysis AMPs antimicrobial peptides APG Angiosperm phylogeny group AQC 6-amino-quinolyl-N-hydroxysuccinimidyl carbamate ASL above sea level BPI base peak ion C-18 carbon-18 CCK cyclic cystine knot cyO18 cycloviolacin O18 cyO19 cycloviolacin O19 cyO2 cycloviolacin O2 DMSO dimethyl sulfoxide ESI electro spray ionization HPLC high performance liquid chromatography HTS high throughput screening IC50 the concentration that inhibited survival by 50% kB1 kalata B1 LC liquid chromatography MARG microautoradiography MS mass spectrometry MS-MS tandem mass spectrometry PBS phosphate-buffered saline solution PE pseudomonas elastase Q-TOF quadropole time of flight RP reversed phase RPMI Roswell Park Memorial Institute medium SCG single cell gel SCX strong cation exchange chromatography SPE solid phase extraction TC top concentration (corresponding to IC50 in FMCA) TDNA tail DNA Xaa an amino acid

Amino acids Amino acid Three letter code One letter code Polarity/Net charge Alanine Ala A hydrophobic Arginine Arg R positively charged Asparagine Asn N hydrophilic Aspartic acid Asp D negatively charged Cysteine Cys C hydrophilic Glutamic acid Glu E negatively charged Glutamine Gln Q hydrophilic Glycine Gly G hydrophobic Histidine His H positively charged Isoleucine Ile I hydrophobic Leucine leu L hydrophobic Lysine Lys K positively charged Methionine Met M hydrophobic Phenylalanine Phe F hydrophobic Proline Pro P hydrophobic Serine Ser S hydrophilic Threonine Thr T hydrophilic Tryptophan Try W hydrophobic Tyrosine Tyr Y hydrophobic Valine Val V hydrophobic

Introduction

Natural products atural products are chemical compounds derived from plants, animals, N and microorganisms; the term is usually reserved for secondary me- tabolites that are not involved in primary metabolic pathways (Canell 1998; Kinghorn et al., 2011; Pawlik, 2011). The discovery and use of natural products by human beings dates back to pre-historic periods (Koehn and Carter, 2005; Pawlik, 2011); natural products have also been an essential source of drug discovery, inspiring scientists for millennia (Clardy and Walsh, 2004; Hong, 2011). They can be utilized for the pharmacological actions they induce by their own right or as pharmacophores upon which semi-synthetic or synthetic analogues can be designed (Koehn and Carter, 2005). The lessons synthetic and medicinal chemists take from the architectural platforms of biosynthesis used by nature are also invaluable (Clardy and Walsh, 2004). Natural products are the richest source of novel compound classes for biological/pharmacological studies due to their wide structural and functional diversity, biochemical specificity, and desirable molecular properties (Koehn and Carter, 2005; Shen et al., 2003). About 50% of the currently marketed drugs discovered between the period of 1981 to 2010 owe their origin to natural products or natural product derivatives; the role of natural products is more pronounced in the areas of anti-cancer and anti-infective agents, where almost two-thirds of the drugs are derived from natural products (Newman and Cragg, 2012). Despite this success, the interest of the pharmaceutical industry in natural products for drug discovery has declined during recent years in favor of newer strategies such as high throughput screening (HTS), combinatorial chemistry, and genome mining, and also for commercial rea- sons (Hong, 2011; Kingston, 2011). These strategies first offered a simpler technique of addressing a defined molecular target and optimizing drug-like structures (Carter, 2011; Koehn and Carter, 2005; Newman and Cragg, 2012); however, the expectations of the pharmaceutical industry did not come to realization but rather to a much higher rate of failure than predicted (Carter, 2011; Koehn and Carter, 2005). For instance, in 30 years of time, combinatorial chemistry could provide only one compound approved as a drug, i.e., the antitumor compound known as sorafenib (Newman and Cragg, 2012). On the other hand, the technological advances in isolation and characterization, function-oriented synthesis, and biotechnology are addressing such limitations as chemical complexity, difficulty in access, and supply of

11 natural products (Carter, 2011; Clardy and Walsh, 2004; Koehn and Carter, 2005; O'Keefe, 2001). Hence, the field of natural product research is once again prevailing in drug discovery (Carter, 2011; Kingston, 2011). In fact, there are suggestions that for the pharmaceutical industry to improve, what is required is digging into natural product research (Carter, 2011; Clardy and Walsh, 2004; Kingston, 2011; Newman and Cragg, 2012); nature will un- doubtedly continue its influence on chemical biology and drug discovery, with its untapped potential of providing novel structures that could be used as drug leads or scaffolds (Carter, 2011; Hong, 2011; Kingston, 2011). The cyclotides that are the subject of this thesis illustrate just one example of this fact.

Cyclotides The cyclotides are a class of gene-encoded circular proteins derived from plants. The discovery of about 70% of medicinal compounds from plant natural products is a result of ethnopharmacological information (Kinghorn et al., 2011). The story of the cyclotides is no different: they were discovered following two independent observations of the ethnopharmacological use of the plant known as Oldenlandia affinis (R&S) DC. The first report came in 1965 from Finn Sandberg, the former professor at the Division of Pharma- cognosy, Uppsala University, where he reported the use of a decoction of wetegere (vernacular name of O. affinis in the Gbaya language) in the Cen- tral African Republic to facilitate childbirth (Sandberg, 1965). Later, in 1970, the Norwegian physician Lorents Gran reported the observation of a Red Cross team working in Congo about the use of a drug called kalata- kalata (in the Tshiluba language) by the Lulua tribe to accelerate uterine contractions; the drug was found to be an aqueous decoction of O. affinis (Gran, 1970). It was prepared in such a way that a handful of the powdered, dried material was boiled in about 1 L of water for 30 min. The resulting green decoction was then taken either orally or directly applied to the birth canal during labor (Gran et al., 2000). The mode of preparation and route of administration indicated that the active principle(s) of the drug is/are both thermally stable and orally bioavail- able. Gran then brought a collection of O. affinis to the Department of Phar- macognosy, University of Oslo, where the active principles were determined to be peptides (Gran, 1973a). Further chemical studies established the partial sequence of the main peptide, which was given the name kalata B1 (Gran, 1973b; Sletten and Gran, 1973). However, it took more than two decades before the unique structure of kalata B1 was fully understood. During the early 1990s, reports about cyclic peptides from plants with the same unique structure as kalata B1's began to appear (Broussalis et al., 2001; Gustafson et al., 1994; Göransson et al., 1999; Schöpke et al., 1993;

12 Witherup et al., 1994). The macrocyclic peptides, which were eventually given the name “cyclotides” (cyclo-peptides) (Craik et al., 1999), are today the largest existing family of circular proteins (see Cybase, the data base of cyclic proteins, www.cybase.org.au) (Mulvenna et al., 2006a; Wang et al., 2008) . The cyclotides have drawn the interest of research ever since, mainly because of their unique, exceptionally stable topological framework but also because of the inherent biological/pharmacological properties they exhibit across a wide range of biological systems.

The cyclotide framework The cyclotides demonstrate a head-to-tail cyclized backbone, i.e., there is no N or C terminal in the structure since the termini are joined by a peptide bond. Moreover, each member of the cyclotide family contains six conserved cysteines that make three disulfide bonds. The cyclic backbone together with the three disulfide bonds create the unique topology of the cyclotides which is termed the cyclic cystine knot (CCK). The knot is formed in a way that two of the disulfide bonds form an embedded ring that is pene- trated by the third disulfide bond (Craik et al., 1999; Göransson and Craik, 2003; Wang et al., 2009) (Figure 1). Besides being a defining topology of this class of circular proteins, the CCK motif endows the cyclotides with exceptional stability; hence they are resistant against the action of common protein denaturing treatments such as heat, enzymes and chemicals (Colgrave and Craik, 2004). The number of amino acid residues found in a single cyclotide ranges between 28-37 (Mulvenna et al., 2006a; Wang et al., 2008); regardless of this small size, the cyclotides are considered to be proteins because of their gene- encoded origin and also because of their well defined three dimensional (3D) structure. The cyclotides further fall into two sub-families depending on the connectivity of their backbone structure. In the first sub-family, the backbone has no twist, like a bracelet, and consequently named the bracelet sub- family. The backbone in the second sub-family, on the other hand, is cyclized with a twist, forming a möbius strip, hence named the Möbius sub- family. The twist in the Möbius cyclotides is a result of the cis-amide Xaa- Pro bond in loop 5; the bracelets lack a Pro residue at the corresponding position (Craik et al., 1999). In addition to this structural difference, the Möbius and bracelet cyclotides differ in size, amino acid composition, and the nature of their inter-cysteine loops. A typical bracelet cyclotide contains one or two more residues than its Möbius counterpart. Moreover, the bracelets contain a higher number of cationic residues. According to the Cybase two thirds of the so far identified cyclotides are bracelets (Mulvenna et al., 2006a; Wang et al., 2008).

Figure 1. Topology of cyclotides illustrating the cyclic cystine knot (CCK). Note that the cyclotides further fall into two main sub-families based on a conceptual topological difference. Here, Kalata B1 (kB1), varv A and vaby A represent the Möbius (A) whereas cycloviolacin O2 (cyO2), cycloviolacin O18 (cyO18) and cycloviolacin O19 (cyO19) represent the bracelets (B). Structures were built from PDB files: 1NB1 (Rosengren et al., 2003) and 2KNM (Göransson et al., 2009). The six conserved cysteines are designated I-VI, the sequence between each cysteine is referred to as a loop; these are marked 1-6.

For most of the known cyclotides, there is a strong sequence similarity in certain loops within but not between the subfamilies. There are a few cyclotides that feature a hybrid nature, i.e., their sequences in loops 2 and 3 resemble those of the Möbius but they lack the Pro residue in loop 5. In fact, their sequence in loop 5 is similar to that of the bracelets (Daly et al., 2006; Gerlach et al., 2010a). Also there have been a few linear cyclotides, i.e., cyclotides with uncyclized backbone (Ireland et al., 2006a; Gerlach et al., 2010a; Nguyen et al., 2011a).

14 Natural sources and biosynthesis

To date, the natural source of cyclotides is the kingdom Plantae. Cyclotides have been isolated from nearly 40 different plant species (Mulvenna et al., 2006a; Wang et al., 2008) belonging to the families Rubiaceae, Violaceae and Fabaceae (Craik and Conibear, 2011). Rubiaceae, also known as the coffee family is the source of the prototype cyclotide kB1. However, only a few of the so far investigated species in the family were found to contain cyclotides, with the content restricted to certain tribes within the family (Gruber et al., 2008). Violaceae (the violet family), on other hand, is a cyclotide-rich family. Cyclotides have been found in all of the violets hitherto screened; moreover, the large share of the cyclotides known today is isolated from Violaceae spp. Cyclotides were recently found in Clitoria ternatea (Butterfly pea) (Fabaceae) (Nguyen et al., 2011b; Poth et al., 2011). Apart from these families cyclotide-like gene sequences have been discovered in the economically significant family of the grains and the cereals, i.e., Poaceae (Basse, 2005; Mulvenna et al., 2006b). Small proteins containing six Cys residues were also isolated from Apocynaceae, a family closely related to Rubiaceae (Gruber et al., 2008). The distribution pattern of the cyclotides within the Plant kingdom is not yet fully understood. Rubiaceae and Violaceae are families from phyloge- netically distant plant orders, i.e., Gentianales and Malpighiales, respectively (Appendix II). The isolation of cyclotides from Fabaceae that is a closer to Violaceae narrows the gap, however, a wider investigation of Plantae is needed to reveal the pattern and the extent of distribution of these circular proteins. The cyclotides are biosynthesized within the plants as gene products (Dutton et al., 2004; Jennings et al., 2001; Mulvenna et al., 2005; Simonsen et al., 2005). This makes them one of the few classes of naturally derived circular peptides (Daly et al., 2009; Trabi and Craik, 2002). The precursor genes, depicted in Figure 2, consist of an endoplasmic reticulum (ER) signal domain, a pro-region and and N-terminal repear (NTR) region followed by mature cyclotide domain. The NTR and the mature cyclotide domain can be repeated up to three times encoding the same or different cyclotides. A cyclotide sequence is obtained after cleaving after Lys/Gly/Asn at the N- terminal and cleaving after and Asn/Asp residue at the C-terminal (Herrmann et al., 2008; Jennings et al., 2001; Mulvenna et al., 2005; Simon- sen et al., 2005; Zhang et al., 2009). The post-transitional biosynthesis, including cyclization and formation of the disulfide bonds is not yet fully understood; plausible hypotheses are that the C-terminal cleavage and the cyclization of the backbone are simultane- ously catalyzed by an asparaginyl endoproteinase enzyme (Gillon et al.,

15 2008; Saska et al., 2007), and a protein-disulfide isomerase is involved in the oxidative folding by re-shuffling the disulfide bonds (Gruber et al., 2006; 2007).

Figure 2. The linear precursor protein of cyclotides. The precursor consists of an endoplasmic reticulum signal domain (ER), a pro-region, an N-terminal repeat signal (NTR), the mature cyclotide sequence and a short tail at the C-terminal. The NTR and cyclotide regions can be repeated up to three times resulting in the encoding of the same or different cyclotides.

Biological activities The first report on activity of cyclotides is the one associated with the discovery of the prototype cyclotide kalata B1. The uterotonic activity of the extract from 1 g of O. affinis leveled that of 2 units of oxitocin (Gran, 1970). Bio-assay guided fractionation studies by independent research groups over the past two decades have shown the cyclotides to be the active principles of various plant extracts with a range of activities. For instance, the cyclotide violapeptide I was isolated as the haemolytic principle of and Viola arvensis Murray (Schöpke et al., 1993). Witherup et al. (1994) then isolated cyclopsychotride A, a neurotensin binding-inhibitor cyclotide from Psychotria longipes Müll.Arg. These were followed by the discovery of cyclotides with anti HIV properties; the circulins A-F from Chassalia pravifolia K. Schum. (Gustafson et al., 1994; 2000), the cycloviolins A-D from Leonia cymosa Mart. (Hallock et al., 2000) and palicourein from Palicourea condensata Cappel (Bokesch et al., 2001). Similar investigations on activity have shown the cyclotides to have insecticidal (Barbeta et al., 2008; Jennings et al., 2001), molluscicidal (Plan et al., 2008), anthelmintic (Colgrave et al., 2009), anti-microbial (Tam et al., 1999) and immunosuppressant (Grundemann et al., 2012) effects. Since 1995, the Division of Pharmacognosy, Uppsala University has been involved in plant polypeptide research, of which cyclotides are the core interest. Our group has reported on the antifouling effect against barnacle lar-

16 vae (Göransson et al., 2004a) and the antimicrobial activity (Pränting et al., 2010) of cycloviolacin O2, and the cytotoxic effects of more than 20 cyclotides against human lymphoma cells (Burman et al., 2011a; Gerlach et al., 2010a; Herrmann et al., 2008; Lindholm et al., 2002; Svangård et al., 2004). The pattern of biological activities of the cyclotides indicates that their natural function within the plants that produce them is very likely as defense molecules against pathogens and pests.

Potential applications of cyclotides Biologically active proteins isolated from natural sources such as microorganisms, animals, plants and marine organisms present a large potential as resources in drug design. The field has been significantly challenged, however, with limitations such as suceptibility to proteolytic degradation and poor bioavailability (O'Keefe, 2001). The cyclotides, with their exceptional stability against thermal, enzymatic or chemical degradation, do not have such a limitation and have now thus become proteins of interest with a significant potential application in drug design. Their unique CCK structure that endows them with their extraordinary stability can be used as a scaffold into which otherwise susceptible bioactive sequences can be grafted and stabilized. This approach has been supported by proof-of-concept studies; Clark et al. (2006) have reported on the plasticity of the cyclotide framework where replacing the hydrophobic residues of kB1 with hydrophilic ones was tolerated. Other, therapeutically relevant grafting studies are the successful grafting of anti-angiogenic and pro-angiogenic sequences to kB1 (Chan et al., 2011; Gunasekera et al., 2008). Cyclotides also find applications associated with their inherent biological activities, particularly their anti-HIV, anti-microbial and anti-tumor (Gerlach et al., 2010b; Lindholm et al., 2002) properties. In non-therapeutic areas, one approach is to employ them as natural pesticides by utilizing their host- defense function in plants. Their reversible and non-toxic anti-fouling effect makes them potential candidates for inhibiting bacterial or fungal growth either in agricultural fields or on surfaces (Göransson et al., 2004a). An effort to insert the cyclotide gene to crops so as to improve their defense has also been already realized (Gillon et al., 2008). Research on employing the cyclotides for drug design is a relatively recent but a rapidly developing field. It has attracted researchers around the world; a number of patent applications have been filed (Smith et al., 2011) and the time for seeing a cyclotide-related drug at the stage of a clinical trial may not be too far-off.

17 Aims of this study

The work presented in this thesis was part of an ongoing research in the Pep- tide Chemical Biology group at the Division of Pharmacognosy, Uppsala University. The ultimate long-term aim of the group is to understand and utilize nature and natural processes, with particular emphasis on peptides, for possible pharmaceutical and medical applications. Discovery, design and activity studies on the cyclotides are at the core of the group's activities.

The specfic objectives of this thesis were to:

• optimize the extraction methods of cyclotides from a plant biomass in order to get a maximum yield of cyclotide cocktail

• characterize cyclotide content of a plant collected from a previously unexplored habitat to see cyclotide sequence diversity with the environment

• determine the stability of the cyclotide framework of both sub-families in potent proteases and in serum in order to see the potential use of cyclotides in development of antibacterial agents

• determine the genotoxicity of cyclotides to know the level of safety of using these proteins in drug design

• undertake a microautoradiographic study to gain more insight into cyclotide-cell interaction and mechanism of action of cyclotides

18 Biochemical studies on cyclotides

The topics in this chapter discuss the studies carried out to optimize parameters around the cyclotides (Papers I-III). Particular areas of interest were the cyclotide extraction protocol, any possible sequence diversity with site of collection of plant material and the stability of the cyclotide backbone in bacterial proteases.

Extraction and isolation Extraction is a critical step in procuring natural products from their respec- tive biological sources. It should be designed in an efficient and economical way that meets the purpose of the work. The two main purposes of cyclotide extraction are to screen a possible cyclotide content of a plant and to isolate of individual cyclotides once cyclotide content is established.

Small-scale extraction Small-scale extraction of cyclotides is the choice of extraction method when one aims to determine the possible cyclotide content of a plant. Extraction is carried out by macerating 10-100 mg of dried plant material in 2 mL of buffer solution (60% MeCN in 0.1% HCOOH) for 24 h (3x). Cyclotides are then captured using a solid phase extraction (SPE) on a 500 mg C-18 column, and then eluted with 60% MeCN in 0.1% HCOOH. After lyophiliza- tion and re-dissolving, an amount corresponding to approximately 1 mg of plant material) is analyzed by LC-MS with a gradient of 10 to 60% MeCN in 0.1% HCOOH (Burman et al., in manuscript). Late eluting peaks, i.e., between 30-60% MeCN, and having mass range between 2500 and 4000 Da indicate the possible presence of cyclotides (Gruber et al., 2008). In our effort to screen the Ethiopian flora for cyclotides, ten different Rubiaceae spp. and one Violaceae (Viola abyssinica Steud. ex Oliv.) were screened using this method. Peaks correlating to cyclotides were observed for V. abyssinica (Figure 3, Paper II). On the other hand, the Rubiaceaeous samples, interest- ingly enough even an Oldenlandia sp., did not test positive. The Rubiaceae spp. screened in this study are listed in Appendix I.

Figure 3. Base peak ion (BPI) chromatogram of Viola abyssinica after small-scale extraction. The late eluting peaks with molecular weight ranging between 2500 and 4000 Da indicate the presence of cyclotides. The small-scale extraction is a fast and economical method for the preliminary screening of plants. A very significant application of this extraction method is that, by exploiting the ultra-stable nature of cyclotides, it can even be utilized on herbarium specimens (Burman et al., in manuscript). Screen- ing of Violaceae spp. from Ethiopian highlands was carried out on herbarium specimens kindly provided by the National Herbarium, Addis Ababa University. Hits were found in the violets, which were as old as 55 years (Table 1).

21 Large-scale extraction When the purpose of extracting a plant biomass is to isolate and characterize individual cyclotides, there is a need to perform large-scale extraction. Two major approaches have been developed in our laboratory for the purpose (Figure 4). The first strategy involves prior defatting of the plant material with CH2Cl2, drying out the MeCl2 and performing the actual extraction with 50% EtOH (Claeson et al., 1998). This method had limitations such as evaporation of CH2Cl2 from the plant material and polyamide filtration to remove tannins, which was not found to be necessary. Hence, it was replaced with a better strategy where the plant material is extracted by 50% aqueous solution of either EtOH or MeOH, after which the extract solution is defatted by liquid-liquid extraction with CH2Cl2 before proceeding to isolation of individual cyclotides (Broussalis et al., 2001; Göransson et al., 1999; 2004b; Herrmann et al., 2008).

Figure 4. Two major approaches developed in our laboratory for the extraction and isolation of cyclotides from a plant biomass. Note that the first approach (left) involves a lengthy-set of procedures; with more knowledge attained about the nature of cyclotides, it was possible to replace it with a simpler method (right). After hydrophobic principles such as plant pigments are removed by liquid-liquid extraction with CH2Cl2, the extract is a mixture of the target compounds, i.e., the cyclotides, and polar ingredients such as phenolic compounds. The polar ingredients are removed using RP-SPE whereby the cyclotides are captured by eluting with 60% MeCN. The resulting extract is now ready for fractionation on preparative RP- HPLC.Cyclotides vaby A-E and varv E were isolated from V. abyssinica using the approach to the right (Paper II).

The extraction method needed to be further fine-tuned in terms of which concentration/type of hydro alcoholic solvent to use, time of extraction, plant material to solvent ratio and number of re-macerations required, to get a maximum quantitative yield from a cyclotide containing plant. Paper I ad- dressed this issue. An optimized extraction protocol not only makes the isolation of cyclotides efficient in research labs but can also be employed in the industrial-scale production of the cyclotides. Viola odorata L. (sweet violet), was selected as a model cyclotide-bearing plant; its cyclotide content has been well characterized by our group and others (Craik et al., 1999; Ireland et al., 2006b; Svangård et al., 2003; Trabi et al., 2004); it is known to express

23 up to 30 cyclotides (Colgrave et al., 2010). Extracts were made in ninety-one different ways, each in a triplicate, and a total of 273 samples (91 x 3) were analyzed by LC-MS (Figure 5). Four cyclotides, namely, cyO2, cyO18, cycyO19 and varv A which have been previously isolated from the plant were used as marker compounds. Extractive yields were compared by taking the integrals of the peaks in the LC-MS chromatograms. The integral values were compared with each other by setting a value of 1 to the highest integral value and by dividing the others against it.

Figure 5. Analyses of extracts of V. odorata. Ninety one types of extracts were prepared by varying solvent type (0-100% v/v of either MeOH or EtOH in water), time of extraction (0.5-18 h), number of re-macerations (1-4X) and plant material to solvent ratio (0.5-2g/10 mL solvent). Chromatograms shown are examples of LCMS analyses of the extracts, note that a higher yield of cyclotides was obtained with solvents of medium polarity (e.g. 30% EtOH) and the yield was very low with a polar solvent i.e., water (Paper I). (V. odorata photo courtesy of Dr Erika Svedlund)

Analyses of the LC-MS chromatograms have shown that hydroalcoholic solutions of medium polarity between the range of 30% and 60% of MeOH give a comparatively high yield. The case was true also for 20% EtOH, one of the cyclotides, i.e., cyO2 even had a maximum yield at this EtOH solution (Figure 6). On the other hand, solvents of extreme polarity/hydrophobicity in this study, for e.g water or the pure alcohols, resulted in low yields of all of the four cyclotides. For instance, extraction with pure EtOH gave only 5% of the yield of cyO2 at 20% EtOH (where maximum yield was obtained),

24 similarly water and pure MeOH gave only 15% and 22%, respectively. The same also holds true for varv A where maximum yield was obtained between 20% and 40% EtOH and the other solvents i.e., water and either of pure EtOH and MeOH yielded only 19%, 3%, and 11% as much, respectively.

Figure 6. Extractive yield from V.odorata by eleven different hydroalcoholic solvents as compared by taking four cyclotides as marker compounds. Legends: W- water, M1-M10 =10% MeOH-100% MeOH, E1-E10 =10%-100% EtOH (Paper I).

Extraction is a critical step in isolating natural products from their sources, and the extraction solvent plays a key role in obtaining the target constitu- ents in a desired quality and quantity (Samuelsson et al., 1985; Sasidharan et al., 2011). The major factor affecting the quantitative yield in this kind of extraction, i.e., maceration, is the ability of the solvent to solubilize ingredients of interest from a plant material. The surface structure of the cyclotides is a property of both their hydrophobic regions and their content of charged residues giving the cyclotides an amphipatic property (Figure 7). Hence it is not a surprise that a higher amount of cyclotides was extracted by solvents of medium polarity and very little/none with highly polar or hydrophobic ones.

Figure 7. Surface structures of four cyclotides. The color scheme of the models is: yellow for hydrophobic residues (Ala, Leu, Ile, Pro, Trp and Val), blue for cationic residues (Arg and Lys), red for anionic residues (Glu) and white for the rest of the amino acids. Note that the cyclotide surface is a mixture of hydrophobic and hydrophilic patches that gives the proteins an amphipatic nature. Structures of cyO2 was built from the PDB file 2KNM. SWISS-MODEL (Arnold et al., 2006; Guex and Peitsch, 1997; Schwede et al., 2003) was used to build the 3D structures of the other three cyclotides based on the PDB file of cyO2 (for cyO18 and cyO19) and that of kalata B1 (1NB1) for varv A.

Once the solvents for the highest possible yield of the cocktail were established, five solvents (water, and 30% and 60% of MeOH or EtOH) were chosen to optimize the other extraction parameters. After the analysis of the complete data, the optimum condition for the extraction of cyclotides was

26 determined as single maceration with at a ratio of 0.5 g plant material per 10 mL solvent for 6 h (Paper I). High yield of the cyclotide cocktail was obtained at 20% EtOH and 50% MeOH, however, extraction with 50% MeOH may be preferred since MeOH is a relatively safe solvent, not possible for user abuse, and has been widely employed in natural product laboratories.

Identification and sequence determination The step that follows the extraction of cyclotides is the isolation of individual cyclotides from the extract and determination of their sequence. It is accomplished by combining the information obtained from MS with data from quantitative amino acid analysis (AAA).

Mass spectrometry Identification and sequence determination of individual cyclotides from V.abyssinica (Paper II) were achieved by MS and MS-MS; the techniques are known to provide highly accurate information on molecular weights of intact peptides and enzymatic digests (Trauger et al., 2002; Wysocki et al., 2005 ). The molecular weights of the purified cyclotides were directly established by MS on a Micromass Q-Tof micro MS system. However, as the CCK motif is very stable, chemical modification is required before sequences could be determined by MS-MS. Modifying cyclotides in order to make them suit- able for MS-MS sequencing involves two steps; the first of which is reduction of the cystines to cysteines using a reducing agent such as dithiothreitol. The second step is alkylation of the cysteines to prevent re-formation of the disulfide knot. Iodoacetamide and iodoacetate are among the commonly used alkylating agents in protein analysis (Creighton, 1980); the former is preferred in modification of cyclotides in that it converts the cysteines to S- carbamidomethylcysteines without introducing a charge difference into the protein whereas the latter produces S-carboxymethyl derivative of cysteines, effectively introducing new negative charges. Reduction and alkylation of each cysteine result in an increase of mass by 58 Da. A total increase of mass by 348 (6 x 58) Da, which is a way to confirm the presence of six cysteines, should be observed for cyclotides. The modified cyclotides can be isolated from the native ones by means of RP-HPLC. The modified ones elute earlier because removal of the CCK results in unfolding of the proteins and their hydrophobic surface is no more exposed. The cyclic backbone is now susceptible to enzymatic digestion, making it possible to make fragments using different proteases. Since all cyclotides contain a conserved Glu residue, a single product can be obtained using Endoproteinase Glu C, which selectively cleaves from the C-terminal of Glu. Fragments can also be ob-

27 tained by digestion with trypsin, whereby specific quantitative cleavage at the C-terminus of the positively charged amino acids i.e., Lys or Arg, occurs. The stability study on the cyclotide framework showed that the cyclic backbone of reduced alkylated cyclotides is also susceptible to proteases like elastase and aureolysin (Paper III). Cyclotides from V. abyssinica were modified according to the above given procedure and two enzymatic digests were prepared for each cyclotide using trypsin and Glu C (Figure 8). The samples were dissolved in 50% MeOH, 1% HCOOH, and MS-MS was performed by Nanospray, using a Protana NanoES source mounted on a Thermo Finnigan LCQ ion trap MS or alternatively by infusing samples through a PicoTip® emitter into a Micro- mass Q-Tof micro-MS system using a NanoLockspray ion source.

Figure 8. Modification of cyclotides to make them amenable for MS-MS sequencing as illustrated by using vaby A as an example. The native structure is given at the top left; reduction with dithiothreitol (DTT) breaks the disulfide bonds and alkylation with iodoacetamide (IAM) converts the cysteines to carbamidomethylcystienes (CAM) hence blocking re-formation of the knot. The cyclic backbone is now susceptible to enzymatic cleavage. Glu C and trypsin digests were used to achieve full sequence coverage of the vaby cyclotides (Paper II).

28 It was possible to complete sequences of six cyclotides, five novel and one already known, from the V. abyssinica extract (Paper II). All of the isolated cyclotides belong to the Möbius sub-family. The novel cyclotides were named vaby A-E following the naming system proposed by Broussalis et al. (2001) where a pronounceable and indicative name is constructed from the binomial scientific name of the plant from which the cyclotide was first isolated and the order of isolation is indicated by the letters A-Z. Sequences of isolated cyclotides are listed in Table 2. Cyclotides vaby A-C demonstrated a further novel feature in that they contained an Ala residue in loop 2 which has not been observed in other cyclotides.

Table 2. Sequence alignment of vaby A-E and varv E isolated from V. abyssinica. Vaby A-E are novel sequences; varv E was previously isolated from V.arvensis (Göransson et al., 1999); its isolation from V.abyssinica is reported for the first time in Paper II. Cyclotide Sequence alignment Net charge vaby A GLPVCGETCAGGTCNTPGCSC-SWPICTRN 0 vaby B GLPVCGETCAGGTCNTPGCSC-TWPICTRN 0 vaby C GLPVCGETCAGGRCNTPGCSC-SWPVCTRN +1 vaby D GLPVCGETCFGGTCNTPGCTCDPWPVCTRN -1 vaby E GLPVCGETCFGGTCNTPGCSCDPWPVCTRN -1 varv E GLPICGETCVGGTCNTPGCSCS-WPVCTRN 0

Ambiguities to solve The leucine isoleucine mystery is a very common challenge encountered in sequence determination of cyclotides, and any protein. The amino acids have identical molecular masses making the MS, which relies on mass-to-charge ratio measurement of analytes, unable to distinguish between them. Enzy- matic digestion of reduced and alkylated peptide with chymotrypsin can in some cases solve the problem, since it selectively cleaves after Leu but not Ile. But as with the cases of the vaby cyclotides where Leu is followed by Pro, cleavage with chymotrypsin is nearly impossible. Information from AAA would not help, since it only provides information on quantitative composition. Hence, positions of Ile and Leu in the cyclotide sequences are determined by homology with other cyclotides, for which data is available from Edman degradation, NMR or sequencing at genetic level (e.g. cDNA). A second phenomenon that could make sequence determination tricky, al- though rather uncommon in cyclotides, is the non-enzymatic deamidation of Asn (to Asp or iso-Asp) and of Gln to Glu. Surprisingly, the former was encountered, but in only one of the vaby cyclotides. During reduction and alkylation, both Asn residues (residual mass 114.1) of vaby C were deami- dated, yielding either Asp (115.0) or isoAsp (115.0) and consequently result-

29 ing in the gain of mass by two units (Paper II). Asn deamidation is pro- moted by alkaline conditions, particularly when the Asn is followed by either a Gly or a Thr residue (Ahern and Klibanov, 1985; Bischoff and Kolbe, 1994; Weintraub and Deverman, 2007) The reduction and alkylation reactions were carried out at pH of 8.5, and either Gly or Thr follows the Asn residues in vaby C. The chemical reaction in the deamidation process of Asn is given in Scheme1.

Scheme 1. Deamidation of Asn to Asp or IsoAsp. A succinimide intermediate is formed during deamidation. The intermediate is then rapidly hydrolyzed to either an aspartate or to an isoaspartate (Ahern and Klibanov, 1985; Bischoff and Kolbe, 1994; Weintraub and Deverman, 2007).

31 Amino acid analysis Quantitative amino acid analysis (AAA) provides information on the amino acid composition of a protein and helps to confirm information obtained from MS. The method relies on vapor/liquid phase hydrolysis of the protein/peptide and subsequent chromatographic separation of the hydrolysate amino acids. Pre- or post-column derivatization of the amino acids is required to determine the amino acid type. In case of the vaby cyclotides, samples were sent to the Amino Acid Analyses Center, Department of Biochemistry and Organic Chemistry, Upp- sala University. They were hydrolyzed with 6 N HCl containing 2 mg/mL phenol (100 °C, 24 h), and analyzed using an LKB amino acid analyzer (Al- pha Plus model 4151) with ninhydrin detection. The procedure is an example of separation of the free amino acids by ion-exchange chromatography followed by post-column derivatization with ninhydrin. Another way of performing AAA is pre-column derivatization of hydrolysate amino acids. These techniques are very sensitive requiring only between 0.5 and 1.0 μg of protein sample (compared with about 5 and 10 μg of protein needed in post-column derivatization). One of the reagents used in this method is 6-amino-quinolyl-N-hydroxysuccinimidyl carbamate (AQC). AQC reacts quantitatively with all primary and secondary amino acids in a few seconds and forms single, very stable derivatives that are amenable to direct analysis by HPLC (Cohen and Michaud, 1993) (Scheme 2). Attempts have been made to determine the amino acid composition of cyclotides using this method. The cyclotide kB1 was first hydrolyzed by vapor phase hydrolysis using constant boiling HCl (6 M) containing 70% thioglycolic acid (105 °C, 24-36 h). The hydrolysate (100 µL) was dissolved in 20 mM HCl and 60 µL borate buffer (0.4 M, pH 9.3). Twenty µL of AQC reagent were added to the solution to tag the amino acids. Analysis on a Shimadzu LC 10 HPLC system equipped with a photodiodide array detector using GROM column (GROM Sapphire 110 C-18, 3 µM) eluted at a linear gradient of 2% of 50mM Na acetate (eluent A) to 75% of 70% MeCN in 30% 50 mM Na acetate (eluent B) in 70 min showed that the method is feasible for determin- ing amino acid composition of cyclotides. Achieving a complete hydrolysis of the peptide was the major success-limiting step to optimizing and setting up the method.

Scheme 2. Derivatization of amino acids with 6-amino-quinolyl-N- hydroxysuccinimidyl carbamate (AQC). AQC reacts quantitatively with primary and secondary amino acids within seconds and to form single, stable derivatives that are amenable to direct analysis by HPLC (A). Excess of the reagent dissociates to give 6-aminoquinoline, N-hydroxysuccinimide and carbondioxide (B), which do not in- terefere with the chromatographic analysis (Cohen and Michaud, 1993)

Quantitative analysis Determination of the relative quantity of a given cyclotide in either a plant extract or an analyte solution was an important task (Papers I, III). It was

33 performed using LC-MS technique on a nanoAcquity UPLC system (MA, USA) coupled with a Micromass Q-Tof micro MS (Waters, USA). ESI-Q- Tof mass spectrometers are known to perform quantitative analysis of peptides at a highly accurate level (Domon and Aebersold, 2006). The calibration curve (Figure 9) constructed for cyO2 demonstrated that the system could effectively be employed for quantitative determination of the cyclotides as well.

12000 y = 905166x + 1011,2 R² = 0,95893 8000 Peak area 4000

0 0 0,002 0,004 0,006 0,008 0,01 0,012

CyO2 in micrograms

Figure 9. Calibration curve constructed for the quantitative analysis of cyclotides on a nanoAcquity UPLC system (MA, USA) coupled with a Micromass Q-Tof micro MS (Waters, USA). peak integrals were computed with MassLynx 4.1 software (Wa- ters Corporation, MA, USA). Triplicates of seven different concentrations of cyO2 were analyzed. Samples were auto-injected.

Stability in proteases As mentioned in the introduction section, the way in which the traditional drug kalata-kalata is prepared, i.e., boiling it with water, showed that the active ingredient(s), now known to be cyclotides, are resistant to thermal degradation. A stability study by Colgrave and Craik (2004) on the prototype cyclotide kB1 also determined the stability of the protein to thermal, chemical, and enzymatic degradation. The presence of cyclotides in almost 200- year-old herbaria specimens screened by our group (Burman et al., in manuscript) further shows the exceptional stability of these proteins. The aim of Paper III was to determine the stability of the cyclotide framework of both sub-families in proteases and in serum to see the potential use of cyclotides in the development of anti-bacterial agents. The study was

34 conducted on kb1 (a Möbius), cyO2 (a bracelet), and MCoTI-II, a cyclic trypsin inhibitor that has a similar CCK to the cyclotides (Chiche et al., 2004). The proteases were from two pathogenic bacteria, namely, Staphylo- coccus aureus and Pseudomonas aeruginosa. S. aureus and P. aeruginosa are practically resistant to most antimicrobial drugs available on the market, reaching the rank of “superbugs” (Breidenstein et al., 2011; Kuroda et al., 2001). Secretion of potent proteases is one of the mechanisms these bacteria use to enhance their virulence and resistance (Peschel and Sahl, 2006). The S. aureus-secreted aureolysin is a potent protease known to effectively de- grade the human antimicrobial peptide LL-37 as well as proteins involved in first-host defense barriers against infection (Sieprawska-Lupa et al., 2004; Laarman et al., 2011), whereas P. aeruginosa produces a protease called pseudomonas elastase (PE) which degrades surfactant proteins, antibacterial peptides, cytokines, chemokines, and immunoglobulins that are components of the immune system (Breidenstein et al., 2011; Kuang et al., 2011). Each of the three proteins, native and the reduced-alkylated form, were incubated in commercially obtained aureolysin (BioCol GmbH, Michendorf, Germany) and PE (Merck KGaA, Darmstadt, Germany) for 8 h at 37 °C (Paper III) and also in culture supernatants of the two bacteria prepared in our lab. The culture supernatants would contain a cocktail of proteases. Sta- bility was also determined in serum following the same procedure. The amount of protein left was quantified by LC-MS. The results showed that the native forms of the proteins were found to be highly stable in commercial enzyme treatments, in S. aureus culture, and in serum, MCoTI-II being comparatively less stable than cyO2 and kB1 (Table 3, Paper III).

Table 3. Stability of cyO2, kB1 and MCoT-II in bacterial proteases, bacterial culture supernatants and serum after incubation over time. Aliquotes were taken at different time points and the amount of protein remained was quantitatively determined by LC-MS (Paper III).

Protein Incubation Amount of protein remained (percent ± SEM) after time (h) treatment with: aureolysin elastase SA PA serum cyO2 0.5 103±0.14 92±0.01 108±0.01 103±0.25 134±0.19 1 95±0.15 102±0.15 132±0.02 107±0.19 NA1 2 103±0.02 101±0.05 104±0.01 104±0.08 114±0.23 4 99±0.16 107±0.14 102±0.01 100±0.19 127±0.30 8 100±0.20 92±0.07 98±0.00 100±0.11 122±0.20 kB1 0.5 94±0.05 108±0.22 123±0.10 115±0.19 107±0.25 1 91±0.08 103±0.26 91±0.25 98±0.12 NA1 2 94±0.11 103±0.29 96±0.26 107±0.17 100±0.12 4 103±0.18 101±0.22 89±0.16 120±0.31 80±0.21 8 83±0.05 102±0.10 128±0.24 117±0.24 100±0.21 MCoTI-II 0.5 92±0.25 89±0.11 76±0.01 97± 0.09 96±0.04 1 75±0.18 90±0.03 66±0.11 98±0.01 NA1 2 73±0.16 93±0.01 100±0.04 101±0.02 96±0.14 4 71±0.22 92±0.06 100±0.08 104±0.02 96±0.17 8 69±0.13 98±0.04 84±0.15 0.1±0.00 76±0.05 SA=S.aureus cultute supernatant, PA= P.aeruginosa culture supernatant 1 aliquots were not taken at this time point

After incubation with P. aeruginosa culture, two of the cyclotides, cyo2 and kB1, were not affected, whereas a slight shift in retention time was observed for MCoTI-II after 30 min, i.e., it eluted earlier, and the peptide was completely lost after incubation for 8 h (Figure 10). MCoTI-II is one of the two cyclic trypsin inhibitors that share a little sequence homology with the cyclotides except for CCK motif. The presence of longer and flexible loops in this protein (Felizmenio-Quimio et al., 2001) could render the general structure less constrained and therefore more accessible to protease cleavage. The increase in flexibility in one loop might well result in an increase of flexibility in the whole structure. Once the circular backbone is cleaved, the constrained structure is relaxed and more targets might become accessible, leading to full degradation. The stability of MCoTI-II in PE of P. aeruginosa but its susceptibility to the culture supernatant of the bacteria shows that the supernatant contains proteases other than PE.

Figure 10. LC-MS chromatograms of MCoTI-II, kB1 and cyO2 after incubation in P. aeruginosa culture supernatant for 8 h. Note that kB1 and cyO2 have the same profile indicating the stability of these proteins aginst proteases secreted by the bacteria. MCoT-II, on the other hand, was fully degraded after 8 h.

37 Cyclotides and cells

Understanding the cyclotide-cell interaction is an essential step in the move toward the development of a cyclotide-based drug. The ability of the proteins to interact with membranes has been previously shown by cellular and membrane model studies. In this thesis, it was possible to show that there are also other cyclotide targets within cells. The details of the studies are sum- marized in the following sections where the cytotoxic property, as reported by other members of our group and in Paper I, as well as genotoxic and cell penetrating effects (Paper IV) of the proteins are discussed.

Cytotoxicity The first report on the cytotoxicity of cyclotides came from our group where the potent cytotoxic effect of the cyclotides against ten human tumor cell lines, including solid and resistant tumors, was demonstrated (Lindholm et al., 2002). The cytotoxic profile of the cyclotides was dose dependent and also distinctively different from those of clinically used anticancer drugs. This finding heralded the emergence of these circular proteins as novel type of cytotoxic agents (Lindholm et al., 2002). Further studies are being pur- sued in our group, both to build a library of cytotoxic cyclotides and also to understand the mode of action behind the cytotoxic effect. The cytotoxic property of cyclotides is determined in the fluorometric microculture cytotoxicity assay (FMCA). FMCA is a short-term microplate- based cell viability assay used to determine the cytotoxic and/or cytostatic effect of a test substance in vitro. The mechanism of the assay is that cells with intact plasma membranes can hydrolyze the probe, fluorescein diacetate (FDA) to fluorescein, which is a highly fluorescent compound. Experimental plates are prepared with a small amount of the test substance, after which cells are seeded on the plates and incubated at 37 ºC and 5% CO2 for 72 h. The cells are then washed with phosphate-buffered saline solution (PBS) and re-incubated for 40 min with FDA. The generated fluorescence, which is directly proportional to the number of surviving cells, is measured with a scanning fluorometer. Data is presented as a concentration versus survival index (SI) curve from which the IC50 (the concentration that inhibited survival by 50%) value is extrapolated (Larsson and Nygren, 1989; Lindhagen et al., 2008).

38 So far the cytotoxicity of more than 20 cyclotides, including vaby A and vaby D from V. abyssinica has been determined using the human cancer cell line U937-GTB. Almost all of the tested cyclotides have exhibited a cytotoxic property at micromolar concentrations (Table 4). The potency of the cyclotides is comparable to the activity of known anti-cancer agents such as doxorubicin, cisplatin and paclitaxel, which have IC50 values of 0.09, 0.49 and 0.003 µg/mL, respectively (Gullbo et al., 2004).

Table 4. Cytotoxic activities and sequences of cyclotides against the human cancer cell line U937-GTB as determined using the FMCA.

3 Protein Sequence Net IC50 Ref. Charge (µM)

Bracelet cyO2 G-IP-CGESCVWIPC-ISSAIGCSCKSKV-CYRN +2 0.26-1.8 1-3 cyO19 GTLP-CGESCVWIPC-ISSVVGCSCKSKV-CYKD +1 0.52 3 vitri A G-IP-CGESCVWIPC-ITSAIGCSCKSKV-CYRN +2 0.6 4 psyle E GVIP-CGESCVFIPC-ISSVLGCSCKNKV-CYRD +1 0.76 5 vibi G GTFP-CGESCVFIPC-LTSAIGCSCKSKV-CYKN +2 0.96 6 vibi H GLLP-CAESCVYIPC-LTTVIGCSCKSKV-CYKN +2 1.6 6 vibi E G-IP-CAESCVWIPCTVTALIGCGCSNKV-CY-N 0 3.2 6 vodo O G-IP-CAESCVFIPCTITALLGCGCSNKV-CY-N 0 3.2 3 cyO2(kyn)1 G-IP-CGESCVWIPC-ISSAIGCSCKSKV-CYRN +2 5.2 3

Hybrid kalata B8 GSVLNCGETCLLGTC---YTTGCTCNKYRVCTKD +1 18 3 psyle A G-IA-CGESCVFLGC---FIPGCSCKSKV-CYFN +1 2 5

Acyclic psyle C ---KLCGETCFKFKC---YTPGCSC-SYPFC-K- +3 3.5 5

Möbius kB22 G-LPVCGETCFGGTC---NTPGCSCTWPI-CTRD -1 2.6 3 vaby D G-LPVCGETCFGGTC---NTPGCTCDPWPVCTRN -1 2.8 II kB132 G-LPVCGETCFGGTC---NTPGCACDPWPVCTRD -2 3.8 3 varv A G-LPVCGETCVGGTC---NTPGCSCSWPV-CTRN 0 6.4-10 1,3 kB12 G-LPVCGETCVGGTC---NTPGCTCSWPV-CTRN 0 6.9 3 vaby A G-LPVCGETCAGGTC---NTPGCSCSWPICTRN 0 7.1 II

varv F G-VPICGETCTLGTC---YTAGCSCSWPV-CTRN 0 7.1 1 kB72 G-LPVCGETCTLGTC---YTQGCTCSWPI-CKRN 0 29 3

vibi D G-LPTCGETCFGGRC---NTPGCTCSYPI-CTRN +1 >30 6 1 The trypthophane in loop 2 is naturally modified into kynurenine 2 kB1, kB2, kB7, kB13=kalata B1, kalata B2, kalata B7, kalata B13 3 References: 1=(Lindholm et al., 2002), 2=(Herrmann et al., 2006), 3=(Burman et al., 2011a), 4=(Svangård et al., 2004), 5=(Gerlach et al., 2010a), 6=(Herrmann et al., 2008)

As could be seen from the list in Table 4, the cytotoxicity of the bracelet cyclotides is generally higher than that of the Möbius cyclotides. Within the bracelets themselves, the more potent ones are those that carry a higher number of charged residues. Structure-activity relationship studies have shown cytotoxic property of the cyclotides to be a function of charged residues and surface properties (amphipathicity) of the proteins (Burman et al., 2011a; Herrmann et al., 2006); disrupting the hydrophobic patch created by Trp resulted in loss of activity. In case of the Möbius, the most potent ones including vaby D contain an extra hydrophobic residue (Phe in loop 3) and have higher net charge. Even though Vibi D is also a charged Möbius cyclotide, it lacks a Trp in loop 5 which would change its amphipathicity and in turn affect its cytotoxicity. Cellular (Svangård et al., 2007) and membrane model studies (Burman et al., 2011b; Huang et al., 2009; Kamimori et al., 2005; Shenkarev et al., 2006) have demonstrated that the mechanism of action in the cytotoxic effect of cyclotides is associated with disruption of the cell membrane. This is discussed later in this thesis under the section "Multiple mode of action".

Genotoxicity The reports on potential application of the cyclotides in drug design have a gap in that there is a lack of fundamental toxicological studies of cyclotides. In Paper IV, a possible DNA damaging effect of cyclotides was determined using the alkaline version of in vitro comet assay. The comet assay, also known as single cell gel assay (SCG) is a sensitive assay for measuring and analyzing even low levels of damage in individual cells. It is a flexible, fast and an easy method to perform requiring only a small number of cells for each sample. After treatment with test substance, cells are suspended in a thin agarose gel on a microscope slide and lysed. The cells are then electrophoresed and stained with a fluorescent DNA binding dye. The electric current pulls any charged DNA from the nucleus hence relaxed and/or broken DNA fragments would migrate further. The resulting images, which appear as ''comets'' (Figure 11), are used to measure DNA damage (Ostling and Johanson, 1984).

Figure 11. The comet assay. The assay got its name because cells appear as ''comets'' in the image analysis. Note that there is no tail in the control cells (A); in cells exposed to a DNA damaging treatment the DNA migrates further from the head (B). The migrated DNA is referred to as tail DNA and it is measured to estimate DNA damage. One advantage of the comet assay is that each cell in the population is individually analyzed. (Photo courtesy of Prof. Björn Hellman)

Three cyclotides, namely, cyo2, kB1 and vaby D were chosen to determine the possible DNA damaging effect of cyclotides. The same cell line, U937- GTB that was used to determine cytotoxicity of the cyclotides in FMCA, was used in the comet assay. Cyclotide exposure concentrations were chosen based on the IC50-values of each cyclotide in the FMCA. The top concentration (TC) was set to be equivalent to the IC50 value, and since the cell count in the comet assay was six times higher than that in the FMCA, the top concentration tested in the comet assay for each cyclotide was six times its IC50- value in the FMCA. Four additional concentrations, i.e., 1/10, 1/50, 1/100 and 1/200 of TC were also tested. The assay was performed according to the procedure given by Singh et al (Singh et al., 1988). Breifly, after incubating with the cyclotides for 3 h, ells were lysed for 1 h in the dark at 4 °C using a freshly prepared solution (2.5 M NaCl, 100 mM Na2-EDTA, 10 mM Tris; pH adjusted to 10 with NaOH, with 1% Triton X- 100 and 10% DMSO added immediately before use). Following lysis, the slides were placed in a horizontal electrophoresis chamber for 40 min with an alkali solution (0.3 M NaOH, 1 mM Na2-EDTA; pH>13) to unwind DNA. Electrophoresis was then performed at 25 V (0.7 V/cm), 300 mA, for 10 min. The slides were removed from the electrophoresis unit and neutral- ized for 15 min with 0.4 M Trizma buffer (pH 7.5). They were then dried at room temperature in a closed hood and stored in a closed container. Before image analysis, slides were dipped in 0.4 M Trizma buffer and stained with 30 µL ethidium bromide (20 µg/mL). The comet assay was based on four independent electrophoresis runs: from each run, three slides, randomly selected and coded, were prepared for the image analysis. Cells (comets) were analyzed using a semi-automated image analysis system (Autocell, Reppalon AB, Hägersten, Sweden) as pre-

41 viously described by Hellman et al. (1995). The number of comets captured per slide was 50 in all experiments; comets without distinct heads ("clouds") were counted manually and excluded from the image analysis. All slides were still coded when the comets were captured. The migration of nuclear DNA from individual cells was measured by evaluating the percentage of DNA in the tail (TDNA). After pooling the TDNA from all captured cells from each individual experiment and treatment, the mean TDNA was used as the indicator of DNA damage. Using the comets as the unit of measure and TDNA as the dependent variable, data for all exposures were evaluated using one-way ANOVA, and the means of TDNA for each treatment were compared with the vehicle control using the unequal N HSD post-hoc test. Statistical analysis was done using the Statis- tica® software, and the level of statistical significance was set at P<0.05. All three cyclotides induced a substantial DNA damage at the top concentration (Paper IV). But the magnitude of DNA damage could not be determined since most cells were dead and appeared as "clouds" under the microscope. At the highest concentration, cell viability was determined to be close to 0%. It should be noted here that the dose-response curve of cytotoxic effect of cyclotides is very steep (Figure 12); a small difference in peptide concentration is enough to induce cell death. The threshold value is also sensitive to the number of target cells and their condition. The combination of these factors is most likely the reason for why almost no cells survived at a concentration that was supposed to represent the IC50 of the tested cyclotides. On the other hand, almost 100% of the cells were viable and no increase in DNA damage or number of "clouds" was observed at the 1/10 and 1/50 dilutions of the highest concentration. However, a modest but statistically significant increase in DNA damage was observed at the 1/100 dilution for two of the cyclotides, cyO2 and vaby D. The results gave a bell shaped dose- response curve for the DNA-damaging effect, as demonstrated in Figure 12 and Table 5. Cell viability was still ~100% and there was no increase in the number of "clouds" at this concentration. The effect disappeared at a further low concentration (1/200) for both compounds. The third cyclotide tested, kB1, did not show any significant DNA damaging effect at any concentration where it was possible to calculate TDNA. The order of potency of the DNA damaging activity of the three cyclotides was analogous to that observed for their cytotoxic effect, i.e., cyO2>vabyD>kB1. However, since the DNA damage was quite modest, it may not lead to cell death but rather there would be a DNA repair. One noteworthy phenomenon observed was that the cyclotide concentration that caused the maximum increase in DNA damage without any con- comitant increase in cell death and/or increase in number of "clouds" was the concentration that was a 100-fold dilution of the top concentration tested.

42 This pattern was even observed for kB1 where the DNA damage was not statistically significant (Figure 12).

Figure 12. Comparison of dose-response curves of cyO2 (A), vaby D (B) and kB1 (C) for DNA damage in comet assay (Paper IV) and cell survival in the cytotoxicity assay (FMCA). Note that the maximum increase in DNA damage is always observed at the 100 dilutions of the 6 x IC50 concentrations in FMCA. For FMCA data see references ( Burman et al., 2011a ; Lindholm et al., 2002) and (Paper I).

Table 5. Effect of cyO2, vaby D and kB1 on the DNA migration in human lymphoma cells, measured as percentage of DNA in the tail (TDNA), after 3 h of exposure and 10 min of electrophoresis. The analysis was done after pooling data from 10 different experiments. Treatment Conc. TDNA No. of cells Clouds Viability (µM) (mean ± SEM) (%) (%) DMSO1 - 3.64 ± 0.11 1643 <1 99 catechol1 750 17.48 ± 0.29*** 1153 <5 99 cyO2 2 Not analysed2 - >99 0 0.2 3.93 ± 0.22 449 <1 89 0.04 4.41 ± 0.22 495 <1 97 0.02 5.43 ± 0.27*** 449 <1 99 0.01 4.41 ± 0.25 400 <1 100 vaby D 20 Not analyzed2 - > 99 0 2 3.50 ± 0.18 600 <1 97 0.4 2.71 ± 0.15 498 <1 100 0.2 4.72 ± 0.23* 595 <1 100 0.1 2.87 ± 0.18 398 <1 99 kB1 50 Not analyzed2 - > 99 2 5 3.47 ± 0.20 448 <1 97 1 3.40 ± 0.20 500 <1 98 0.5 3.50 ± 0.20 448 <1 98 0.25 3.06 ± 0.20 399 <1 98

* P < 0.05; *** P < 0.001 (One way ANOVA; Post Hoc: Unequal N HSD) 1DMSO and cathecol were used as a negative and a positive control respectively. 2 Since almost all comets obtained after the electrophoresis at the highest concentration tested of the three different cyclotides were "clouds" (comets without distinct heads), no comets were captured for subsequent image analysis at these concentrations

Cell permeation The results from the comet assay of cyclotides lead to a question: what is the cellular effect of cyclotides at different dose levels, particularly at the 1/100 dilution of the top concentration? Microautoradiography (MARG) analysis was hence carried out on CyO2, the most potent cyclotide in cytotoxicity and also in DNA damaging effect (Paper IV). Autoradiography is a technique where radioactive particles (or

44 energy) localized within a solid substance are imaged by putting closer to a detection media (Solon, 2012). MARG is an application of autoradiographic technique at cellular level; it is used to spatially locate drug-derived radioactivity within cells (Solon, 2012). It is particularly useful in that it has predic- tive value for specific drug targeting (Solon et al., 2010).

Radiolabeling A radioactive derivative of cyO2 was made by labeling the protein with 3H- acetic anhydride at the two Lys residues in loop 5 (Blanchfield et al., 2003). Specific acetylation conditions for cyO2 such as acetic anhydride equiva- lence and reaction time, were first optimized using cold acetic anhydride. Two equivalents of acetic anhydride and a reaction time of 24 h were required for acetylating both Lys residues (Paper IV). The reaction was moni- tored by MS, and the shift in molecular mass (+84 Da) confirmed that both lysines had been acetylated. Hot cyO2 was then prepared in an analogous manner; the radiolabeling was carried out such that the final radioactivity in all three concentrations was 1 µCi/mL. Briefly, three vials containing the three different concentrations of cyO2, i.e., 2 µM (TC), 0.2 µM (1/10 TC), and 0.02 µM (1/100 TC) in DMSO were prepared. To the top and middle (1/10 TC) concentrations, 0.02 equiv and 0.2 equiv, respectively, of 3H- acetic anhydride was added; the reaction was allowed to stand overnight before adding unlabeled acetic anhydride (1.98 equiv to the top and 1.8 equiv to the middle). The low (1/100 TC) concentration was acetylated with only 3H- acetic anhydride (2 equiv). All reactions were performed at room temperature. A minimum amount of water was added to each vial after 24 h and a stream of N2 was passed to remove the volatiles.

Microautoradiographic analysis Cells were exposed to 3H-acetyl cyO2 under the same conditions as in the comet assay. A shorter exposure time (30 min) was also used for the TC because it induced about 40% cell death under the normal exposure time (3 h). The decrease in cytotoxicity from 100% to 40% was because the two Lys residues are modified which is in parallel with the report of a previous structure-activity study (Göransson et al., 2009; Herrmann et al., 2006). The human lymphoma cells (1 x 106 cells/mL) were exposed to the radiolabeled cyO2 (1 µCi/mL) for 3 h for each of the three concentrations or 30 min (TC only) and washed tree times with RPMI 1640. The cells were then fixed in 4% phosphate-buffered formaldehyde solution and washed with PBS. 100 µL of 2% low-melting point agarose was carefully added to the pellet, which was then put on ice. The solidified agarose lump was treated with different concentrations of ethanol: 70% (overnight), 95% (4 h) and

45 100% (3 h), and then xylene (30 min) before it was embedded in paraffin (Histovax, Sweden; melting point 52-54°C) overnight. The paraffin block was sectioned in 4 µm thin sections, which were put on microscope slides. The paraffin was removed by putting the slides in xylene, 100% ethanol (, 95% ethanol, 70% ethanol and then finally in water. The slides were covered with Kodak NTB-2 emulsion and exposed for 2 weeks. After photographic development and fixation (Kodak Unifix, France), the slides were stained with Mayers hematoxylin (30 seconds). Cellular uptake of cyO2 could then be studied, as the silver grains in the liquid film are stained dark by the radioactive protein. The slides were coded, and each slide was visually examined under a microscope for quantitative and spatial distribution of dark silver grains. Four different researchers independently scored the radioactivity associated with the cells using the following scale: 1 (cells without visible silver grains); 2 (cells with 1-3 silver grains); 3 (cells with 3-10 silver grains) and 4 (cells with >10 silver grains). During the quantitative analysis of the labeling of the cells, the position of the silver grains in the cell was also noted (Table 6).

Table 6. Independent observations of four researchers on MARG slides of human lymphoma cancer cells incubated with three different concentrations of 3H-acetyl cyO2 (Paper IV). Dose and Re- Number of silver grains Locataion of silver grains, incubation1 searcher 0 1-3 4-10 >10 additional remarks low, 3 h A 17 90 10 0 B 9 73 19 0 C 0 75 0 0 around the cell membrane and D 58 38 4 0 nucleus membrane low, 3 h A 5 48 47 0 B 6 72 22 0 C 1 38 66 2 D 4 53 36 7 all over the cell middle, 3 h A 8 58 34 0 B 5 58 37 0 at the cell membrane, the C 10 77 23 0 nucleus membrane and over D 27 56 17 0 the nucleus middle, 3 h A 1 11 77 14 B 0 28 66 6 C 1 12 59 29 D 8 48 41 3 all over the cell high, 3 h A 10 41 49 0 B 12 73 15 0 close to the cell membrane C 10 65 24 1 and the nucleus. Fragmented D 7 39 52 0 and irregular cells high, 30 min A 14 73 13 0 B 18 78 4 0 C 19 80 1 0 all over the cell and close to D 13 58 27 2 cell membrane. high, 30 min A 2 19 63 16 B 0 30 67 7 C 9 43 43 5 All over the cell. Leaking D 8 43 40 9 cells. High background. 1All slides were coded during the analysis.

Dark grains were observed inside the cell membrane at all concentrations, as demonstrated in Figure 13. Grains were also observed near the nuclear membrane, and in some cases, apparently also inside the nucleus. This observation confirmed that cyO2 could penetrate the cell membrane, and even the nuclear membrane, at both cytotoxic and lower concentrations.

Figure 13. Images of microautoradiography slides after human lymphoma cancer cells (U-937 GTB) were incubated with 3H-acetic anhydride labeled cyO2 for 30 min at a concentration of 2 µM (A and D) and for 3 h at concentrations of 0.2 µM (B and E) and 0.02 µM (C and F). The dark silver grains are associated with the protein; note that dark grains were observed inside the cell and nuclear membranes and also, in some cases, in the nucleus. Kodak NTB-2 emulsion film was used. Im- ages were taken with NIKON (digital camera, DXM 1200F).

So what do the dark grains represent? They do not represent single cyO2 molecules (because of their low specific activity and the sensitivity of the film) instead, they are interperated to be pinosomes, i.e., small internal membrane vesicles filled with peptides that have permeated into the cell through pinocytosis. Cyclotide pinocytosis was first suggested by Greenwood et al. (2007) and has recently been demonstrated using fluorescence labeled derivatives (Cascales et al., 2011; Contreras et al., 2011). In comparison, the radiolabeling method used in Paper IV brings a minimal structural change to the cyclotide. In comparison to the middle and lower concentrations, the top concentration showed a relatively high background. This background was interpreted as the result of peptides bound to the debris from dead cells. In fact, cell viability dropped from 80% at 30 min to 67% at 3 h. The cells in the TC appeared morphologically deformed at both times of incubation, i.e., 30 min and 3 h. Another point to highlight is that since the total radioactivity was the same in all concentrations, the higher number of dark grains at the highest concentration represented an actual increase in uptake (Table 6). This could be due to the compromised cellular membrane that consequently led to

48 higher cellular permeation. Gerlach et al. (2010b) reported such a phenomenon where the membrane interaction of cyO2, at sub-cytotoxic concentration, increased membrane permeability leading to permeation of substances that would not cross the membrane at normal conditions.

49 Discussion

Cyclotide production

The employment of peptides and proteins as drugs dates back to 1920 when bovine insulin was introduced to therapeutics. The use of these macromole- cules in drug research, however, has become popular fairly recently, during the past two decades. High specificity, affinity, molecular recognition and low toxicity profile are some of the virtues that have attracted interest in protein-based drugs (Loffet, 2002; Nelson et al., 2010; Vlieghe et al., 2010). The possibility to isolate and characterize biologically active proteins from natural product extracts has also contributed to the popularity of these molecules by providing novel sequences with potent activities (O'Keefe, 2001). The field of developing proteins or peptides into drugs is highly challenged by the less stable and orally unavailable nature of proteins and also high cost of production (O'Keefe, 2001; Loffet, 2002). Cyclotides are stable proteins, and the cyclotide kB1 is orally available as could be seen back in the way kalata-kalata was prepared and administered (Gran, 1973b). The cyclotide framework is amenable to total chemical synthesis via solid phase peptide synthesis methods (Aboye et al., 2011; Leta Aboye et al., 2008; Gunasekera et al., 2006). Even though the total chemical synthesis is the main technique for cyclotide research, it is not the option for industrial scale production of cyclotides mainly due to low yields and also due to the involvment of harsh solvents. Other methods of cyclotide production that utilize molecular biology and protein engineering techniques have also been described. Examples include the successful production of kB1 and MCoTI-II in engineered E.coli cells (Camarero et al., 2007; Kimura et al., 2006), production of both circular and linear forms of cyclotides in trans- genic species of Arabidopsis thaliana, Nicotiana tabacum and Nicotiana benthamiana (Gillon et al., 2007), and also in plant bioreactors using cul- tured cells of O.affinis (Dornenburg, 2010). Each of the above methods provided an alternative to the production of cyclotides by extraction from their natural sources. However, the methods have not yet proven to be commercially practical ways of making cyclotides, mainly, because of limitations associated with our lack of full understanding of cyclotide biosynthesis. Hence, the natural sources, i.e., cyclotide-bearing plants continue to be the viable suppliers of cyclotides. The big advantage of

50 direct extraction from plants is that the cyclotides obtained have the correct post-translational modifications. Also, as many as 30 to 50 different cyclotides can be produced by extraction only from a single species (Colgrave et al., 2010; Göransson et al., 2003; Trabi and Craik, 2004). Designing an effective and environmental friendly extraction method that can be extended to industrial scale production, however, is necessary. The optimized extraction protocol in Paper I showed that an optimum yield of the cyclotide cocktail can be obtained using low cost and environ- mentaly friendly solvents within few hours of extraction. Moreover, the method used is a good mimic of the traditional way of extraction but with a gentle treatment of the plant material. The traditional method of kalata- kalata preparation, as mentioned in the introduction section, is by boiling the leaves with water, i.e., making a decoction. Extraction by macerating with organic solvents of medium polarity is known to give extracts of the same consistency as extraction by decoction: in some cases, organic extracts may even be found to have better activity than the decoctions (Samuelsson et al., 1985). The small-scale extraction method that has been used to screen cyclotide content of herbaria and field specimens also has advantages not only in the research laboratory but in commerical production of cyclotides as well. It provides an economical way of screening large number of samples since it requires a very small amount of plant material (10 -100 mg) and time to be carried out.

Nature as a source of novel templates

Diverse sequences As has been mentioned in the first chapter, the cyclotides are one example of nature's immense potential to provide chemical entities with novel structures and desirable properties. Their unique topology, stability and potent activities are good illustrations of this fact. Even more, the diversity and plasticity they exhibit within themselves is intriguing. According the Cybase (Mulvenna et al., 2006a; Wang et al., 2008), 200 cyclotides are so far isolated; and except for the six cysteines and the conserved Glu residue, their intercystine loops exhibit great sequence diversity, especially in the bracelets. Figure 14 depicts the frequency hits of amino acid residues in each cyclotide loop.

Figure 14. Sequence logo of Möbius (top) and bracelet (bottom) cyclotides as constructed by the Weblogo (Crooks et al., 2004). Note that apart from the six cysteines and the conserved Glu residue in loop 2, other positions are occupied by different amino acid residues. The bracelets are more diverse than the Möbius, and contain higher number of residues.

The diversity seen today is well expected to grow as the number of cyclotides existing in nature is predicted to exceed 50,000. Violaceae species alone are estimated to express from 5,000 to 25,000 different cyclotides (Gruber et al., 2006; Mulvenna et al., 2006b). Our group has been involved in mapping the cyclotide distribution and sequence variation within the plant kingdom in general and within the cyclotide-rich family of violets in particular (Broussalis et al., 2001; Burman et al., 2010; Göransson et al., 1999; Göransson et al., 2004b; Herrmann et al., 2008; Svangård et al., 2004; Svangård et al., 2003). Violaceae is a cosmopolitan family of 22 genera and approximately 930 species (Hekking, 1984, 1988). In the above-cited studies, Viola and Hyban- thus species growing in Sweden and also Gleospermum species from Pan- ama are covered. The screening of plants from the Ethiopian highlands was an extension of this effort. Since the cyclotides are defense molecules, plants growing in different habitats could express different cyclotide cocktails at different time points. Trabi et al. (2004) adequately demonstrated this in comparing cyclotide expression patterns of Australian and Swedish violets during different seasons. The five novel sequences isolated from V. abyssinica, collected from 3400 m ASL, added to the known sequence diversity of the cyclotides, particularly that of the Möbius ones (Figure 15). The Ala residue in loop 2 was observed for the first time in the vaby cyclotides and reported in Paper II. The finding further corroborated that investigating the cyclotide contents of violets growing in diverse environments is a promising

52 approach to extend our knowledge of both the structural and the biological diversity of cyclotides.

Figure 15. Contribution of the vaby cyclotides isolated from v.abyssinica (A) to the diveristy wheel of the Möbius cyclotides (B). (Paper II).

Lessons from the knot To once again recall a statement from the introduction, the lessons we take from the architecture of natural molecules are endless (Clardy and Walsh, 2004). The topology of the cyclotides, particularly the CCK motif, is already giving scientists insight into how to design stable proteins with desirable pharmacokinetic properties. A very good case in this context is the case of the conotoxins. The conotoxins are venom peptides of the cone snails. They are disulfide-rich (there are also variants that lack disulfides) linear peptides consisting 12-30 amino acids. They have the ability to block specific ion channels and hence have the potential to be used as diagnostic or therapeutic agents in neurological disorders. One member has already been clinicaly approved for neuropathic pain under the name Prialt® (Lewis, 2009; Terlau and Olivera, 2004). How- ever, despite their great potential as drug leads, they are susceptible to degradation; furthermore they are not orally available. Clark et al, a group that also works with cyclotides, were able to make stable and orally available conotoxin derivatives by cyclizing the backbone of the disulfide-rich conotoxins with a linker sequence of six peptides (Clark et al., 2005; 2010). Simi- lar work using the CCK architecture is used to design a peptide-analogue of a drug is the synthesis of an MCoTI (the cyclic trypsin inhibitors with CCK) analogue of a foot-and-mouth virus inhibitor (Thongyoo et al., 2008). Another class of peptides to which the lesson from the architecture of cyclotides is being applied to is the antimicrobial peptides (AMPs) (Rosengren

53 et al., 2004). AMPs are small molecular weight proteins that are components of the innate host-defense immune system (Izadpanah and Gallo, 2005). They have a broad-spectrum antimicrobial activity against bacteria, viruses, and fungi, and there is an interest in developing them into "peptide antibiot- ics" that could be used in treating infections (Izadpanah and Gallo, 2005; Peschel and Sahl, 2006). The limitation with AMPs, however, is their susceptibility to degradation by bacterial proteases (Guina et al., 2000; Peschel and Sahl, 2006; Sieprawska-Lupa et al., 2004). Nature has created a strategy to make AMPs more resistant to proteolysis with the introduction of disulfide bridges; the naturally occurring AMPs with disulfide bridges are more resistant to degradation than those that lack disulfide connectivities (Campopiano et al., 2004; Maemoto et al., 2004). One study has shown thar re-designing of the defensins, one class of naturally occurring AMPs, based on the topology of the cyclotides known as the circulins enhanced the stability of the defensins without affecting their antimicrobial potency (Yu et al., 2000). Cyclization of the linear backbone of AMPs via a disulfide bridge has also shown to be effective in improving the stability of AMPs (Rozek et al., 2003). The other approach is scaffolding of the active sequence from an AMP to the cyclotide framework (Camarero, 2011). But in all cases, it would be good to know the extent of the stability of the framework in a wide array of proteases. The data provided in Paper III, where stability is determined in proteases of S. aureus and P. aeruginosa is one input into the query.

Multiple modes of action The mechanisms and/or targets behind the broad range of activities of the cyclotides are not yet fully understood. Interaction with membranes has been identified as a plausible explanation for many of the activities including uterotonic (Gran, 1973a), antimicrobial (Shenkarev et al., 2006), cytotoxic (Svangård et al., 2007; Burman et al., 2011b), insecticidal (Barbeta et al., 2008) and neurotensin antagonist (Witherup et al., 1994) effects. Studies based on membrane models reported the interaction of cyclotides with membrane components such as cyclotide binding to membrane lipids (Shenkarev et al., 2008; Shenkarev et al., 2006), cyclotide membrane affinity (Kamimori et al., 2005), and formation of multi-pores in the membrane by cyclotides (Huang et al., 2009). Recent reports showed that the cytotoxic effect of the cyclotides is correlated with their ability to adsorb and disrupt model lipid membranes of different charge densities (Burman et al., 2011b; Henriques and Craik, 2012). At cellular level, the membrane-disrupting effect of cyO2 has been established using the liposome assay where the cyclotide caused leakage in the liposomes (Svangård et al., 2007).

54 The findings in Paper IV demonstrated that there is another mode of action of cyclotides apart from membrane interaction. The types of DNA damage that could be detected in the comet assay range from DNA single strand breaks (including incomplete excision repair sites) to increase in alkali labile sites and/or cross-linking (Tice et al., 2000). The presence of any possible damage is, in all cases, measured on viable cells, making it clear that the cytotoxic effect and the DNA damaging effect could not be caused by the same mechanism of action. As such, the DNA damaging effects of cyO2 and vaby D at the 1/100 dilution are the first functional proof of a mechanism of action that is not targeting the cellular membrane per se. The cellular permeation of cyclotides, as confirmed by MARG could confirm the ability of cyO2 to penetrate into the cell and possibly even to the nucleus. Nonetheless, from the results, it was not possible to explain the bell- shaped dose-response curves of the cyclotides in comet assay since cell permeation was observed at all concentrations but with no DNA damage at the TC or the middle (1/10 TC) concentrations. These facts demonstrate that DNA damage must have been caused by another mode of action at the lower concentration (1/100 TC). It is reasonable to hypothesize that the down- stream target organelle such as the lysosome may be affected. Another possible target could be the nuclear membrane or a target inside the nuclei, as suggested by the dark grains observed in those areas of the cells (Figure 13). Hence, even if the internal target remains unknown, the findings of Paper IV have unequivocally demonstrated that these cyclotides have multiple modes of action, and that the type of target is dependent on concentration. However, we are far from knowing the full story about the mode of action of this intriguing family of circular proteins. The fact that they could have more than one mode of action makes them even more attractive molecules in areas like cancer and infectious diseases treatment where drug resistance is a major problem. It would take considerable effort from the cancer cells and/or the microorganisms to develop resistance to more than one mode of action.

55 Concluding remarks and future perspectives

Tuning parameters that could be inputs toward the development of cyclotide- based drugs was the interest of this thesis. An extraction protocol was developed where cyclotide extraction was achievable with mild and inexpensive solvents such as hydro acloholic mix- tures of MeOH or EtOH. The protocol can be practically extended to large- scale production of the cylotides from their natural sources. The cyclotide content of Violaceae and Rubiaceae species from Ethiopian highlands was also mapped. The only Viola species in the area, i.e., Viola abyssinica, was characterized in detail and five novel sequences were determined. From the Rubiaceae samples, it was surprising to see that even the Oldenlandia species was devoid of cyclotides. The results from this study in general, illustrated how cyclotide expression varies with habitat. In the stability study, native cyclotides have demonstrated highly stable profile indicating that they can be employed in the development of antimicrobial drugs that are resistant to degradation by bacterial proteases. Based on the results of the genotoxicity study, it can be concluded that the cyclotides do not induce an alarming DNA damage. In fact one of the cyclotides, kB1, did not show any significant effect in the comet assay. Besides providing the first toxicological data for the proteins, the results provided a new insight about the effect of cyclotides on subcellular targets. The ability of the cyclotides to permeate into cells even at cyctotoxic concentrations was clearly demonsted by the microautoradiography study. The two cellular studies added to our knowledge about cyclotide-cell interaction by clearly demonstrating that these proteins have multiple modes of action. This would make the cyclotides even useful in therapeutic areas where drug- resistance is a major problem. The cyclotides are indeed remarkable proteins with immense potential. It is no surprise that they have become an attractive subject of research ever since their discovery. As the discovery of the cyclotides is relatively recent (at least in the scientific community), there remain a lot of dots to connect to reveal the full picture about their properties. I believe areas covered in this thesis would contribute toward this effort; and accordingly to the realization of a cyclotide-templated drug, the aspiration of the "cyclotide groups" working at different areas of the world. For the future, it is my hope to see the biosynthesis of the cyclotides fully understood so that producing them by means of genetic engineering and/or

56 biotechnology becomes achievable, and also to see the full distribution extent and pattern of these interesting compounds in nature uncovered. It would be even more exciting to establish the modes of cellular and subcellular interactions of cyclotides including their final fate within a cell.

'Look deep into nature, then you will understand everything better' Albert Einstein

57 Popular science summary

We all know what proteins are, the big molecules formed by chains of small building blocks called amino acids. This thesis focused on proteins but very small ones known as cyclotides. Apart from size, two things make the cyclotides different from the common proteins. First, unlike the chain structure of most proteins that has a head and a tail, the cyclotide chain is completely closed like a ring. In fact that's where they got their name from, i.e., cyclo- peptides, cyclotides. The second interesting thing about the cyclotides is that they are highly stable against treatments such as heat, enzymes and chemicals, which would break down other proteins readily. The exceptional stability of the cyclotides always reminds me of a Swedish phrase I saw on a dress for an infant, it read 'liten men tuff'. The cyclotides are not infant proteins yet to grow bigger but indeed they are 'little but tough'. Tough not only in with- standing breakage but also in inducing mny different activities such as uterotonic, antiHIV, antimicrobial, cytotoxic, insecticidal etc. The discovery of teh cyclotides dates back between the mid 1960s and early 1970s. Prof Finn Sandberg from the Pharmacognosy Department of Uppsala University and Lorents Gran a physician in a Red Cross team (now a Professor in Norway) independently reported the use of a plant known as Oldenlandia affinis in Central African Republic and Congo to facilitate child birth. Gran brought a sample of the plant to Norway and identified the active uterotonic constituent to be the first cyclotide that he named kalata B1 (after the traditional name of the drug in Congo, i.e. kalata-kalata). Today, about 200 cyclotides are known and most of them have been isolated from the family of the violets, Violaceae. Rubiaceae, the coffee family, was the origin of the first cyclotide kalata B1; but many of its species including coffee, were not found to contain cyclotides. Plants are the only known natural sources of the cyclotides, and a single cyclotide-bearing plant can produce from 30 to 50 different types of cyclotides at a time. The purpose of the cyclotides in plants that produce them is to provide chemical defese against microbial pathogens and pests. Natural product research laboratories, on the other hand, have seen an immense potential in these molecules to develop peptide-based drugs. The cyclotides, due to the wide range of activities they exhibit, may be employed as drugs by their own right. For example studies in our group have shown the cyclotides to be a novel type of cytotoxic agents against a panel of human cancer cell lines. Another way of using cyclotides in drug development is to use

58 their ultra-stable framework as a scaffold for otherwise labile molecules. There are several puzzles to solve though, from establishing a reliable production means to completing the toxicity profile of the molecules and a lot in between, before the realization of a cyclotide-based drug. Synthesis methods to make cyclotides in laboratories have been established; it was also possible to produce them in genetically engineered bacteria and in cell cultures of O. affinis. None of these, however, have been proven to be viable means of cyclotide production for large-scale supply. Therefore, cyclotide production still relies on extraction of the proteins from their natural sources. An extraction protocol is given in this thesis; the protocol involves the use of mild and inexpensive solvents. Moreover, it was possible to illustrate that by exploring plants growing in different habitats, cyclotides of diverse sequences can be obtained. Viola abyssinica a plant that grows on the Ethiopian highlands was found to contain novel sequences that have not been found in other species growing elsewhere. A stability study was conducted on the cyclotides using proteases (enzymes that breakdown proteins) from common pathogenic bacteria, namely, Pseudomonas aeruginosa and Staphylococcus aureus. These bacteria are known to secrete a cocktail of proteases as a means to enhance their virulence. The cyclotides tested in this thesis were found to be resistant to the action of these proteases. Another part of the thesis covered understanding of the cyclotide-cell interaction. A toxicological study, the first of its kind on the cyclotides, was undertaken to see the possible DNA damaging effect of the proteins. One of the tested cyclotides, the prototype kB1, did not induce any significant effect whereas two other cyclotides induced a mild DNA damage.. The studies on cellular effects were further continued; a cyclotide was radiolabeled by at- taching a radioactive chemical to one of the amino acids within the chain. Cells were incubated with the labeled cyclotide; they were sliced and examined under the microscope. It was clearly seen that the cyclotides could cross the cell membrane and penetrate into the cell. Results from the cellular studies demonstrated that the cyclotides have more than one target within the cell. The findings of this thesis would have their contribution toward the effort to utilize the cyclotides in drug development. With the rate of the cyclotide research at the present speed, it is likely that a lot more gaps will be filled and a cyclotide-based drug would soon appear on a clinical trial.

59 አጭር መግለጫ

ዚህ መግለጫ ይህ መመረቂያ ፅሁፍ የተመሰረተባቸው ጥናቶች አጠር በ ባለና ማንኛውም ሰው ሊረዳቸው በሚችል መልኩ ይዳሰሳሉ፡፡ ይህ ፕሮጀክት የአዲስ አበባ ዩንቨርሲቲ ፋርማሲ ትምህርት ቤት እና ስዊድን አገር የሚገኘው ኡፕሳላ ዩንቨርሲቲ በጋራ ከነበራቸው ለመድሐኒትነት ሊውሉ የሚችሉ እፅዋት ምርምር ፕሮጀክት አንዱ አካል ነው፡፡ ለጥናቶቹ አስፈላጊ የነበሩት ስራዎች የተከናወኑትም በሁለቱ ዩንቨርሲቲዎች በሚገኙት የፋርማኮግኖሲ የትምህርት ክፍሎች ውስጥ ነው፡፡ ፋርማኮግኖሲ ምንድነው? ፋርማኮግኖሲ የሚለው ቃል “ፋርማኮን” እና “ግኖሲስ” ከሚሉ ሁለት የግሪክ ቃላት የተፈጠረ ሲሆን ጥሬ ትርጉሙም “ስለ መድሐኒቶች ማወቅ” ማለት ነው፡፡ የፋርማኮግኖሲ ጥናትና ምርምር በተለይ ከተፈጥሮ በሚገኙ ለመድሐኒትነት ሊውሉ በሚችሉ ንጥረ ነገሮች ላይ ያተኩራል፡፡ በዚህም በእፅዋት፣በእንስሳት፣በማእድናት እንዲሁም በአይን በማይታዩ ረቂቅ ህዋሳት ውስጥ ያለውን ለመድሐኒትነት የመዋል አቅም መመርመር ዋነኛው የፋርማኮግኖሲ ተግባር ነው፡፡ የሰው ልጅ ተፈጥሮን በተለይም እፅዋትን ከብዙ ሺህ ዘመናት ጀምሮ በባህላዊ ህክምና ለመድሐኒትነት ሲጠቀም ኖሯል አሁንም እየተጠቀመባቸው ይገኛል፡፡ ዛሬ በሳይንሳዊ መንገድ ተጠንተው በህክምና ለምንገልገልባቸው መድሐኒቶች መገኘት ተፈጥሮ ያበረከተችው ድርሻም ምንም አይነት መተኪያ ይቅርና ተወዳዳሪም ያልተገኘለት ነው፡፡ እነዚህ በሳይንስ ተጠንተው በህክምና ላይ ከዋሉ መድሐኒቶች አብዛኞቹ በቀጥታ ከተፈጥሮ የተገኙ ወይም ተፈጥሮን እንደ መነሻ በማድረግ በላቦራቶሪ የተሰሩ ናቸው፡፡ እንደዚህም ሆኖ በተፈጥሮ ውስጥ ያለው እምቅ ሀይል ገና ያላወቅነውና ያልተነካ ነው፡፡ ተፈጥሮ አዳዲስ ይዘትና ከለመድናቸው የተለየ ችሎታ ያላቸው ንጥረ ነገሮች ምንጭ መሆኗም ይቀጥላል፡፡ ይህ ጥናት የተካሄደባቸው ሣይክሎታይዶች የተባሉ ፕ ሮቲኖች ለዚህ እውነት አንድ ጥሩ ማሳያ ናቸው፡፡ ሣይኮሎታይዶች ከእፅዋት የሚገኙ ፕሮቲኖች ሲሆኑ ለመጀመርያ ግዜ ስለመኖራቸው የታወቀው ከ50 ዓመታት በፊት ነው፡፡ እንደ አውሮፓውያን

60 አቆጣጠር በ1965 ዓ.ም. ወደ ምስራቅ አፍሪካ ጉብኝት አድርገው የነበሩት የኡፕሳላ ዩንቨርስቲ ፋርማኮግኖሲ ፕ/ር ፊን ሳንድበሪ በአካባቢው ነዋሪዎች ዘንድ “ወተገሬ” እየተባለ የሚጠራ ቅጠል ወሊድን ለማፋጠን ምጥ ለያዛቸው እናቶች ሲሰጥ ማየታቸውን ይፅፋሉ፡፡ ከ 5 ዓመታት በኋላ በ1970 ዓ.ም. በኮንጎ (የቀድሞዋ ዛየር) የነበረውን ችግር ለመርዳት ተጉዞ የነበረው ቀይ መስቀል ቡድን አባላት በኮንጎም “ካላታ-ካላታ” የተባለ ባህላዊ መድሐኒት ለተመሳሳይ ጥቅም መዋሉን ይመለከታሉ፡፡ የቡድኑ አባል የነበሩት ኖርዌያዊው የህክምና ዶ/ር ሎሬንትስ ግራን በኮንጎ ዩንቨርስቲ ውስጥ የመድሐኒቱን ተክል ምንነት ካለስዩ በኋላ ለተጨማሪ ምርምር ወደ ኖርዌይ ዩንቨርስቲ በመውሰድ በተክሉ ውስጥ ያለው የማህፀንን ጡንቻ የማኮማተር ሃይል ያለው ንጥረ ነገር ፕሮቲን መሆኑን ይለያሉ፡፡ ለንጥረ ነገሩም ካላታ-ቢ-ዋን የሚል መጠርያ ይሰጡታል፡፡ በሁለት የተለያዩ ቦታዎች ለተመሳሳይ ጥቅም ሲውል የተገኘውና ካላታ -ቢ-ዋንን ያበረከተው ተክል በሳይንሳዊ አጠራር ኦልዴንላንዲያ አፊኒስ የሚባል ሲሆን በስነ- እፅዋት መሰረት ምድቡ ቡና የሚገኝበት የ እፅዋት ቤተሰብ ውስጥ ነው፡፡ ካላታ-ቢ-ዋን ፕሮቲን መሆኑ ቢታወቅም ሙሉ ይዘቱና ከተለመዱት ፕሮቲኖች የተለየ የሆነው ቅርፁ ተጠንቶ ያለቀው ከ20 ዓመታት በኋላ በ1990ዎቹ መጨረሻ ላይ ነው፡፡ ሣይክሎታይዶች ከለመድናቸው ፕሮቲኖች ከሚለዩባቸው ዋና ዋና ነጥቦች አንዱ ኬሚካላዊ ቅርፃቸው ነው፡፡ እንደሚታወቀው ፕሮቲኖች አሚኖ አሲድ የተባሉ ትንንሽ ንጥረ ነገሮች እንደሰንሰለት በመ ያያዝ የሚፈጥሯቸው ግዙፍ ኬሚካሎች ናቸው፡፡ በተፈጥሮ የሚገኙና ፕሮቲኖችን በመፍጠር የሚታወቁት አሚኖ አሲዶች 20 ብቻ ቢሆኑም በተለያየ ቅደም ተከተል በመደርደርና በተለያየ መልኩ በመደጋገም ለቁጥር አዳጋች የሆነ ብዛት ያላቸውን ልዩ ልዩ ፕሮቲኖችን ይፈጥራሉ፡፡ አንድ ፕሮቲን በብዙ መቶዎች የሚቆጠሩ አሚኖ አሲዶችን ሊይዝ የሚችል ሲሆን የአሚኖ አሲዶቹ ሰንሰለት በተለያየ መልኩ ሊተጣጠፍ ይችላል፡፡ የሰንሠለቱ ሁለት ጫፎችም የፕሮቲኑ “ ራስ” እና “ጅራት” ተብለው ይጠራሉ፡፡ ወደ ሣይክሎታይዶች ስንመለስ በአንድ ሣይክሎታይድ ውስጥ ያሉ አሚኖ አሲዶች ብዛት በአማካኝ 30 ብቻ ሲሆን ሰንሰለታቸውም “ ራስ” እና “ጅራት” የለውም፡፡ ምክኒያቱም የሰንሰለቱ ሁለት ጫፎች ተመልሰው በመጋጠም እንደ ቀለበት የተዘጋ ቅርፅ ስለሚፈጥሩ ነው፡፡ በተጨማሪም አጠቃላይ ቅርፁ በቀለበቱ ውስጥ በተዘረጉ 3 የተቆላለፉ ድልድዮች የተቋጠረ ነው፡፡ እንደዚህ ያሉ ድልድዮች ሲስቲን የተባለው አሚኖ አሲድ ጥንድ ሆኖ ሲገኝ አንዱ ከሌላኛው ጋር በመያያዝ የሚፈጥሯቸው ናቸው፡፡ ስለዚህም በእያንዳንዱ ሣይክሎታይድ ውስጥ 6 ሲስቲኖች ይገኛሉ ማለት ነው፡፡ ይህ የተቆለፈ የሣይክሎታይዶች ቅርፅ “ቀለበታማው የሲስቲን ቋጠሮ” በመባል ይጠራል፡፡ ሌላው ሣይክሎታይዶችን ከተለመዱት ፕሮቲኖች የሚለያቸው ነጥብ እንደ ሙቀት፣ኬሚካልና የመሳሰሉትን ፕሮቲኖችን የሚሰባብሩ ነገሮች መቋቋም መቻላቸው ነው፡፡ የፕሮቲኖች ሰንሰለት (እጅግ ግዙፍ የሆኑትን ጨምሮ) በሙቀት፣ በኬሚካል ወይም ፕሮቲኖችን በሚሰብሩ ኤንዛይሞች

61 በቀላሉ ሊሰበር ወይም ይዘቱን ሊቀይር የሚችል ነው፡፡ ይህ በቀላሉ የመሰባበር ፀባ ይ የፕሮቲኖችን ለመድሐኒትነት የመዋል እምቅ ችሎታ እንዳንጠቀም ፈተና ከሆኑት ነገሮች ውስጥ ዋናው ሲሆን እንደ ሣይክሎታይዶች ያሉ ይህንን ፈተና የተሻገሩ ፕሮቲኖች መገኝት በዘርፉ ላይ ለሚደረገው ምርምር አዲስ ምዕራፍ የከፈተ ነው ማለት ይቻላል፡፡ በተጨማሪም እነዚህ “ትንንሽ ግን ብርቱ” ፕሮቲኖች ባክቴሪያን፣ ፈንገስን፣ ተባዮችን እንዲሁም ኤች አይ ቪ ቫይረስንና የካንሰር ህዋሳትን ለመግታት ያላቸውን አ ቅም ለመፈተሸ የተደረጉ ምርምሮች ተስፋ ሰጪ ውጤቶችን አስመዝግበዋል፡፡ በአሁኑ ወቅት በሣይኮሎታይዶች ላይ የሚደረገው ምርምር ሁለት ዋና መንገዶችን ይዞ እየተካሄደ ነው፡፡ የመጀመርያው “ቀለበታማው የሲስቲን ቋጠሮ” ላይ ያተኮረ ሲሆን ለመድሐኒትነት ተፈላጊ ፀባይ እያላቸው ነገር ግን በቀላሉ በመሰባበራቸው ምክኒያት አገልግሎት ላይ መዋል ያልቻሉ ፕሮቲኖችን ሣክሎታይዶች ላይ በመጫን መሰባበርን ለመቋቋም እንዲችሉ ለማድረግ መሞከር፤ ወይም ተፈጥሮ በሣክሎታይዶች ያስተማረችንን መሰረት በማድረግ በቀላሉ ለሚሰባበሩ ፕሮቲኖች በላቦራቶሪ ውስጥ “ቋጠሮ” ለማበጀት መሞከርን ያካትታል፡፡ ሁለተኛው የምርምር መንገድ ደግሞ ሣይክሎታይዶችን ራሳቸውን ለመድሐኒትነት ለማዋል የሚደረግ ነው፡፡ በዚህ የመመረቂያ ፅሁፍ የተካተቱት ጥናቶች ለእነዚህ የምርምር መንገዶች ግብዓት እንዲሆኑ ታልመው የተደረጉ ናቸው፡፡ ሣይክሎታይዶች ለመድሐኒትነት ኢንዲውሉ ካስፈለገ በብዙ መጠን ሊመረቱበት የሚችሉበትን መንገድ መቅረፅ የግድ ነው፡፡ ሣይክሎታይዶችን በላቦራቶሪ ውስጥ ለመስራት የተቻለ ቢሆንም አንድ ሣይክሎታይድ ለመስራት የሚፈጀው ግዜና ገንዘብ እጅግ ብዙ ከመሆኑም በላይ በመጨረሻ የሚገኘው የሣይክሎታይድ መጠን በጣም አነስተኛ ነው፡፡ በእፅዋት ውስጥ ሣይክሎታይዶች የሚሰሩበትን መንገድ ጄኔቲክ ኢንጅነሪንግ እና ባዮቴክኖሎጂ በተሰኙ ቴክኒኮች ወደ ባክ ቴሪያና ወደ ሌሎች እፅዋት አስተላልፎ ሣይክሎታይዶችን ለማምረት የተደረገው ጥረትም እስካሁን አስደሳች ውጤት አላሳየም፡፡ ስለሆነም እነዚህን ፕሮቲኖች በብዙ መጠን ለማምረት አሁን ያሉን ብቸኛ አማራጮች በተፈጥሮ ሣይክሎታይዶችን ይዘው የ ሚገኙት ተክሎች ናቸው፡፡ በመጀመርያው ጥናት የተከናወነውም ሣይክሎታይዶችን ከተፈጥሮ ምንጮቻቸው በአጥጋቢ ሁኔታ ለመለየት የሚያስጭል ዘዴ በመቀየስ ላይ ነበር፡፡ በውጤቱም አካባቢ ላይ ብዙ ጫና በማያሳድሩና በጣም ውድ ባለሆኑ የውሀና የአልኮል ቅልቅሎችን በመጠቀም ሣይክሎታይዶችን ከእፅዋት የምናገኝበትን መንገድ ለማሳየት ተችሏል፡፡ ሣይክሎታይዶችን በኢንዱስትሪ ደረጃ ለማምረት ቢፈለግ ይህንን መንገድ ምንም ለውጥ ሳያስፈልገው በቀጥታ መጠቀም ይቻላል፡፡ ሁለተኛው ጥናት ያተኮረው በኢትዮጲያ ውስጥ የሚገኙ ዝርያዎች ውስጥ ሣይክሎታይዶችን መፈለግ ላይ ነበር፡፡ የመጀመርያው ሣይክሎታይድ የተገኘው ቡና ከሚገኝበት የእፅዋት ቤተሰብ ቢሆንም በመቀጠል ቡናን ጨምሮ አብዛኞቹ ዝርያዎች ላይ በተደረገው ጥናት ሣክሎታይዶችን

62 ማግኘት አልተቻለም፡፡ ይህም በዚህ የእፅዋት ቤተሰብ ውስጥ ሣይክሎታይዶች በብዛት እንደማይገኙ አመልክቷል፡፡ በተቃራኒው “ቫዮሌሴ” ወይም በተለምዶ “ቫዮሌት” የሚባል ቤተሰብ ውስጥ ሣይክሎታይዶች በሰፊው ተገኝተዋል፡፡ በአገራችን ወደ 10 የሚጠጉ የቫዮሌት ዝርያዎች ሲገኙ በዚህ ጥናት ሁሉም ዝርያዎች ተካተዋል፡፡ ከቫዮሌቶቹ መካከል “ቫዮላ አቢሲኒካ” የተባለውን ተክል ከመስክ መሰብሰብ የተቻለ ስለነበር ተክሉን በዝርዝር በማጥናት ከውስጡ 5 ከዚህ በፊት ያልታወቁ አዳዲስ ሣይክሎታይዶች እና 1 ቀደም ሲል ከሌላ “ቫዮላ” የተገኘ በጠቅላላው 6 ሣክሎታይዶች ተለይተዋል፡፡ አዲስ የተገኙት ሣይክሎታይዶችም የቫቢ (ቫዮላ አቢሲኒካ ከሚለው በመውሰድ) ሣክሎታይዶች የሚል መጠርያ ተሰቷቸዋል፡፡ በቀሪዎቹ ቫዮሌቶች ላይ ጥናት ለማረግ ናሙናዎች የተወሰዱት አዲስ አበባ ዪንቨርስቲ ከሚገኘው ብሔራዊ የእፅዋት ቤተ- መዘክር ነበር፡፡ ናሙናዎቹ ከመስክ ከተሰበሰቡ እስከ 50 ዓመት እድሜ ያስቆጠሩ ቢሆንም ሣክሎታይዶችን በውስጣቸው ይዘው ተገኝተዋል፡፡ ይህም የሣይክሎታይዶችን ሳይበላሹ የመቆየት ጥንካሬ የበለጠ ያሳየ ነው፡፡ በተመሳሳይ ከቡና ቤተሰብ የሆኑ 10 እፅዋት (አንድ ኦልዴንላንድያን ጨምሮ) ላይ በተደረገው ፍተሻ እፅዋቶቹ ውስጥ ሣይክሎታይዶች እንደሌሉ ታይቷል፡፡ በአጠቃላይ ከሁለተኛው ጥናት የተረዳነው በተለያዩ አካባቢዎች የሚበቅሉ እፅዋት የተለያዩ አይነት ሣይክሎታይዶችን እንደሚይዙ ወይም እንደሚያድጉበት አካባቢ ሁኔታ ሣይክሎታይዶችን ላያመርቱ እንደሚችሉ ነው፡፡ እፅዋት ሣይክሎታይዶችን የሚያመርቱበት ዋና ተግባር ራሳቸውን ከበሽታ አምጪ ተህዋስያንና ተባዮች ለመከላከል ነው፡፡ በሶስተኛው ጥናት ሣይክሎታይዶችና ባክቴሪያዎች ተካተዋል፡፡ ለጥናቱ የተመረጡት ባክቴሪያዎች ሲውዶሞናሰ አሬጊኖሳ እና ስታፊሎኮከስ ኦሪየስ ይባላሉ፡፡ ባክቴሪያዎቹ የበርካታ በሽታ ምክኒያቶች ከመሆናቸውም በላይ አሁን በገበያ ላይ ያሉትን አብዛኞቹን መድሐኒቶች ሰለተለማመዱ ለማከም አስቸጋሪ የሆኑ ናቸው፡፡ እነዚህ ባክቴሪያዎች የሰውን በሽታ የመከላከል ሀይል ከሚያዳክሙበት መንገድ አንዱ በሰውነታችን ውስጥ በተፈጥሮ የሚገኙ ፀረ -ባክቴሪያ የሆኑ ፕሮቲኖችን መሰባበር ነው፡፡ ለዚህም እንዲረዳቸው የተለያዩ ኤንዛይሞችን ያመነጫሉ፡፡ በአሁኑ ወቅት እነዚህን የባክቴሪያ ኤንዛይሞች መቋቋም የሚችል ተከላካይ ፕሮቲን ለመፍጠር ብዙ ምርምር በመደረግ ላይ ይገኛል፡፡ ለዚህ ፅሁፍ በቀረበው ሶስተኛ ጥናት ሳይክሎታይዶች የነዚህን ባክቴሪያዎች ኤንዛይሞች ለመቋቋም ያላቸው ሀይል ተለክቷል፡፡ ለኤንዛይሞቹ ከተጋለጡት ሣይክሎታይዶች ላይ በተለያዩ ሠዓታት ናሙና በመውሰድ በተደረገው ክትትል ኤንዛይሞቹ ሣይክሎታይዶቹ ላይ ምንም ለውጥ ለማምጣት አለመቻላቸውን ለማየት ተችሏል፡፡ በተጨማሪም “ቀለበታማው የሲስቲን ቋጠሮ” ለዚህ የሣይክሎታይዶች ጥንካሬ ያለውን አስተዋፆ ለመለካት ሲባል ቋጠሮውን በኬሚካላዊ መንገዶች ላቦራቶሪ ውስጥ በመፍታትና ቋጠሯቸው የተፈታውን ሣይክሎታይዶች ለኤንዛይሞቹ በማጋለጥ ጥናት ተደርጓል፡፡

63 በውጤቱም ቋጠሯቸው የተፈታው ሣይክሎታይዶች በ30 ደቂቃዎች ውስጥ በኤንዛይሞቹ መሰባበራቸው ታይቷል፡፡ አንድ ንጥረ ነገር ምንም ያህል ተፈላጊ ፀባዮች ቢኖሩት በሰው ልጆች ላይ የሚያስከትለው ጉዳት ካለ ለመድሐኒትነት መዋል አይችልም፡፡ ስለዚህም የትኛውንም ንጥረ ነገር ወደ መድሐኒትነት ለመቀየር ከሚደረጉ ምርምሮች ውስጥ ንጥረ ነገሩ በተለያየ ደረጃ ሊኖረው የሚችለውን መርዛማነት መመርመር አንዱ አካል ነው፡፡ አራተኛው የዚህ ፅሁፍ ጥናት በሣይክሎታይዶች ላይ የተደረገ የመጀመርያው “መርዛማነትን የመመርመር” ጥናት ነው፡፡ በጥናቱ ሣይክሎታይዶች ዲ ኤን ኤ የተሰኘውን ዋና የህዋስ ክፍል ሊጎዱ መቻል ወይም አለመቻላቸው ተፈትሿል፡፡ በውጤቱም ካላታ- ቢ-ዋን ምንም ጉዳት እንደሌለው ሲታይ ቫቢ ዲ (ከቫዮላ አቢሲኒካ የተገኘው) እና ሳይክሎቫዮሌሲን-ኦ-ቱ የተባሉት ሣይክሎታይዶች ግን መጠነኛ ጉዳት አድርሰዋል፡፡ የዚህ ጥናት ውጤት አስገራሚ ነበር፡ የመጀመርያው አስገራሚ ክስተት የሣይክሎታይዶቹ የተለያየ ፀባይ ይዞ መገኘት ሲሆን ሁለተኛው ደግሞ ዲ ኤን ኤ ላይ ጉዳት ያሳየው የሁለቱ ሣይክሎታይዶች ልክ (መጠን) ሲጨመርም ሆነ ሲቀነስ ዲ ኤን ኤ ላይ የታየው ጉዳቱ መጥፋቱ ነበር፡፡ የበለጠ ለማብራራት ለምሳሌ ጉዳቱ የታየበትን የሣ ይክሎታድ ልክ (መጠን) “ለ” ብ ንለው የ “ለ” 10 እጥፍም ሆነ የ “ለ” ግማሽ መጠን ዲ ኤን ኤ ላይ ጉዳት አያደርሱም፡፡ ይህ በአብዛኛው ከምናውቀው መጠን ሲጨምር የሚከተለው ውጤትም አብሮ መጨመር ሁኔታ የተለየ ነበር፡፡ ተጨማሪ ከጥናቱ የታየው ሌላ ክስተት የዲ ኤን ኤ ጉዳት የደረሰባቸው ህዋሳት በሕይወት የነበሩ መሆናቸው ነው፡፡ ከዚህ በፊት ሣይክሎታይዶች ህዋሳትን (ሴሎችን) የመግደል ፀባይ እንዳላቸው ተረጋግጧል፡፡ ህዋሳትን የሚገድሉበት መጠን ግን የ “ለ” 100 አጥፍ ነው፡፡ በአጠቃላይ ከዚህ ጥናት ውጤት ሣይክሎታይዶች ከህዋሳት ጋር ስላላቸው ግንኙነት ያልተረዳነው ነገር እንዳለ ለመገንዘብ ተችሏል፡፡ ስለዚህም “ሣክሎታይዶችን በህዋስ ውስጥ መከተል” የሚል አምስተኛ ጥናት ማካሄድ አስፈላጊ ነበር፡፡ የጥናቱ ዋና ዘዴ ማይክሮአውቶራዲዮጊራፊ የተሰኘ የጨረር እና የፎቶግራፍ ጥበብን አንድ ላይ በማቀናጀት የሚሰራ ቴክኒክ ነው፡፡ በጣም በቀላሉ ለመግለጽ፡ በመጀመርያ ሣይክሎታይዱ ላይ ጨረራማ ኬሚካል በመጫን ሣክሎታይዱ ራሱ ወደ ጨረራማነት እንዲቀየር ተደረገ፤ በመቀጠልም ለጥናቱ የተዘጋጁት ህዋሳት በተለያየ መጠን (በ “ለ”፣ በ “ለ” 10 እጥፍና በ “ለ” 100 እጥፍ) ከተዘጋጀ ጨረራማ ሣይክሎታይድ ጋር ተቀላቅለው ቆዩ፤ በመጨረሻም ናሙናዎች ተወስደው እንደ ፎቶግራፍ ፊልም ማጥቆርያ ባለ ፊልም ተሸፈኑ፡፡ ይህ ፊልም እንደተባለው እንደፎቶግራፍ ማጥቆሪያ ቢሆንም የሚጠቁረው ግን ጨረር ሲነካው ነው፡፡ ስለዚህ በዚህ ጥናት ጠቁሮ ሊታይ የሚችለው ጨረራማ ሣይክሎታይድ ያለበት ቦታ ብቻ ነው ማለት ነው፡፡ ከውጤቱም ሣይክሎታይዱ በሁሉም መጠን ወደ ህዋሳቱ ውስጥ ዘልቆ የመግባት ባህሪ እንዳለው ታይቷል፡፡ ሕዋሳትን በሚገድለው መጠን ጭምር ህዋሳቱ ከመሞታቸው በፊት ሣይክሎታይዱ በውስጣቸው ገብቶ እንደነበር ለማየት

64 ተችሏል፡፡ ከዚህ በፊት የምናውቀው ሣይክሎታይዶች ህዋሳትን የሚገድሉት የህዋሳቱን የውጪ ሽፋን በመጉዳት እንደነበር ነው፡፡ አሁን በዚህ ጥናት የተገኘው ውጤት ግን ሣይክሎታይዶች ሌላም ህዋሳትን የመግደል መንገድ እንዳላቸው አሳይቷል፡፡ የዚህን መንገድ ምንነት በትክክል ማወቅ ግን ገና ይቀረናል፡፡ ከአምስቱ ጥናቶች የተገኙት ውጤቶች በሣይክሎታይዶች ምርምር ላይ ጠቃሚ አስተዋፆ እንዳበረከቱ እምነቴ ነው፡፡ በተለይም አራተኛውና አምስተኛው ጥናት በዘርፉ ላለው ምርምር አዳዲስ መልሶችን ሳይሆን አዳዲስ ጥያቄዎችን በመጨመር ለተመራማሪዎች አዲስ አቅጣጫ የጠቆሙ ናቸው፡፡ በመጨረሻ ለማጠቃለል የምፈልገው እነዚህ የብዙ ሳይንቲስቶችን ቀልብ የሳቡና በሚሊዮኖች የሚቆጠር የምርምር ገንዘብ የሚፈስባቸው ሣይክሎታይዶች መነሻቸው ከባህላዊ ህክምና የተገኘ መረጃ መሆኑን በማስታወስ ነው፡፡ በአገራችን የባህል ህክምና ውስጥ የካበተ እውቀት አለ፤ ይህ እውቀት በአግባቡ ጥቅም ላይ ውሎ የምናይበትን ቀን እኔና ፋርማኮግኖሲስት ባልደረቦቼ ተስፋ እናደርጋለን፡፡

ምስጋና፡ ለፕሮጀክቱ አስተባባሪዎች፡ ፕ/ር ጽጌ ገብረማርያም፣ዶ/ር ቃለአብ አስረስ እና ፕ/ር ላሽ ቦሊን ለጥናቱ አማካሪዎች፡ ዶ/ር ኡልፍ ዮራንሶን፣ዶ/ር ቃለአብ አስረስ እና ፕ/ር ላሽ ቦሊን ለአዲስ አበባ ዩንቨርስቲ ብሔራዊ የእፅዋት ቤተ-መዘክር ለፕሮጀክቱ የሚውል በጀት ላበረከተው የስዊድን ሲዳ/ሳሬክ

65 Acknowledgements

'I love you, LORD, my strength. The LORD is my rock, my fortress and my deliverer; my GOD is my rock, in whom I take refuge, my shield and the horn of my salvation, my stronghold' (Psalms 18: 1-2). My foremost gratitude goes to Him alone.

Writing this section of the thesis was probably the most challenging because the list of individuals I would like to thank would simply, as the saying goes, take a village.

This study was part of a collaboration project between the School of Phar- macy, Addis Ababa University (AAU) and the Division of Pharmacognosy, Uppsala University (UU). Funds for the project were kindly provided by the Swedish International Development Cooperation Agency (Sida)/SAREC.

I am indebted to Prof Tsige Gebre-Mariam, Dr Kaleab Asres, from AAU, and Prof Lars Bohlin, from UU, for their efforts in designing this insightful project. I would also like to acknowledge the two senior members of AAU, for they have been, and continue to be, great inspirations for us who got the chance to be their students and colleagues. Prof Lars Bohlin and Dr Kaleab Asres were also involved in my study as deputy supervisors.

My main supervisor, Dr Ulf Göransson, I am sincerely impressed by your calm personality and way of analyzing things. And thank you for believing that this was possible even when I myself thought otherwise.

People who took part in the studies: Dr Robert Burman - my co-author in all of the papers, thanks for being a helpful colleague, Prof Björn Hellman - my mentor on the comet assay and microautoradiography, it was really great to work with you. Lena Norgren - a person who knows what she does and never forgets to smile, I do not have enough words to thank you. Dr Adam Strömstedt - the bacteria and membrane specialist, thank you for the history and culture discussions for a change.

Melaku Wondafrash, from all those field trips we made together I could only see that you are a walking book of the flora of Ethiopia, thanks for finding the violet that was hidden in our sight. Hailemeskel (Gash Haile) -

66 thank you for your invaluable help during the plant collection. My due thanks is also to project students Camilla Eriksson and Hanna Justad.

I am grateful to the kind cooperation of the people at the National Herbar- ium, AAU who understood the purpose of the study and allowed me to take samples from their precious collections.

Many thanks to the staff at the International Science Program, UU who have effectively handled the administrative side of the project and bent the rule that needed to be bent. I would like to thank in particular, Dr Peter Sundin and Dr Linnea Sjöblom, for their effort to extend the collaboration, Hossein Aminaey and Pravina Gajjar, for handling the practical matters and Zsuzsanna Kristófi for her warm hospitality.

To the former and present members of our cyclotide group: Dr Erika Sved- lund - who introduced me to the lab for the first time on my very first day, Dr Anders Herrman - the MS and HPLC guy, Dr Sunithi Gunasekera with whom I had great times in Copenhagen and Stockholm and around the lab as well, Sungkyu Park and Soahib Zafar - the new runners in the relay, Stefan Svahn - my room mate, Delgerbat Boldbaatar - who introduced me to Mongolia, and project students Taj M Khan, Millie Rådjursöga, Erik Jacobsson, I want to say, it has been a pleasure!

I can not forget to mention Dr Teshome Leta Aboye, a fellow colleague all the way back from Addis and his wife Argane Dhaaba, for sharing a lot along the path in ways only people away from home would understand.

I am also grateful to my other colleagues at the Division: Dr Cecilia Als- mark - the British only, Anna-Maria Andersson, Prof Anders Backlund - the botanist, Maj Blad - my recipe guide, Dr Hesham El-seedi, Gunilla Eriksson, Dr Jenny Felth, Dr Sonny Larsson, Dr Erik Hedner, Åke Strese, Eva Strömberg, Elisabet Vikeved- nice to have you around, and Dr Christina Wedén - a great personality. One needs a Mac-expert like Mag- nus Jansson to be able to include a summary written in the 'Geez' alphabet.

The "analytical people" from the neighboring division of Analytical Pharma- ceutical Chemistry: Alexander & Anna, Albert, Annica, Annita (remem- ber that fat TV?), Axel, Curt, Douglas, Henrik, Jacob, Marcus, Matilda, Victoria (waiting to see yours next week fellow soldier) and Ylva, thank you all for enjoyable coffee times.

It want to take this opportunity to express my heartfelt gratitude to the won- derful friends I made in Sweden:

67 Kerstin and Jerry Ståhlberg together with their sons are simply the best. I enjoyed Kerstin's warm companionship from the Division's administrative office to the delivery table at the hospital. Not to mention the countless din- ners and holidays spent with their warm family.

Josefin Rosén, my former roommate, made my landing in Sweden soft. Thank you Josse for being my directory, my translator and most of all for being a good friend.

Barbara and Tobbe, remarkable people indeed! Barbara is my personal "eniro" who not only tells directions but also takes me there, we shared a lot of happy events and like 'once in life time' Volvo adventures.

Anna Székely, I wonder if I find the right words to describe what friends my daughters and I found in you (and Lulu, Zelma and Zozo). But one sure thing is some of this work would not have been possible if you were not there as a standby mother for my girls.

The list will not at all be complete with out mentioning Jenny Nordin, an amazing friend I met at the last hour, and Sussana Strand, my dependable neighbor.

My Abesha friends living in Uppsala: Eyob & Aster were the first ones I came to know and I am grateful for all their kindness. I am highly touched by the love from brothers and sisters at Salem Church. I particularly want to thank Mulu & Adamu, Yoki, Tsehay & Bekele and Lidya & Johannes. And also Geniye, who took care of me like a mother. I recommend her 'Mesob African Restaurant' to any one who wants to experience the fantas- tic Abesha cuisine.

Back home...

I am grateful to senior members of the School of Pharmacy, AAU: Dr Ephrem Engidawork, Dr Teferi Gedif, Dr Araya Hymete and Dr Nigussu Mekonnen (ex-member). It is also a great pleasure to join back my former colleagues Adem, Anteneh, Azmera, Beletishachew, Belaynesh, Bruck, Elias, Fantu, Meseret, Solomon, Shemsu, Yetimgeta, and also the younger recruits who expanded our school.

I want to thank my dear friends with whom I have a lot to catch up with: Meron G - for I found a sister in you, Meron M - for laughing and crying together, Edom - Addis won't be the same without you, Yodit (Yoda) - for I know you will always be there, Zinet & Gelane - for your own definition of

68 keeping in touch, Rediet (kulu) - for being such an incredible personality, and Hamere (Hamiye) - for you know me inside out.

My timeless friends in C Choir, thank you for the encouraging words and prayers, it is such a blessing to have you all! Particularly, Brikti & Feven F (Bibiya too), and Hermon, for your sincere love - love you back! Binyam S, Binyam B, Nathnael A & Nathnael A (Germany), Dagnachew, Joseph G, and Elias A for the 'laugh guaranteed' moments, and of course Johannes Kidane (Johnny), for being a great role model.

The following people were always by my side when I needed them: Aberash - my mother-in-law, Abuti (Enanu S), Amanuel, Alem, Asfaw M (Asfiti), Etye Shewaye, Firye, Kebebush (kebe), Medhanit & Shewaneh, Senait A (Seniye), Yeshareg, and Ato Eshetu. I am also indebted to Kesis Belay M and his family. Special thanks goes to Tewabech F who keeps all in order when I am away.

Last but certainly not least, I should thank members of my family, undoubt- edly starting from my grandmother, for she was my first and ever-great teacher and my confidant; for I knew the world only through her. I never thought I would have to write the word 'late' in dedicating this thesis to you, Abuye, may you rest in peace.

My mother, Sara, I could only find two words to describe you best - "iron lady", which indeed you are. My brother Bruck - the little boy who grew into a brave man in front of my eyes, my sister Hiwot (Mimiye) - who is overfilled with kindness and calm, my little sister Bemnet (Doni or was it BB now?) - in whom I see so much of myself, I love you all!!

Words totally fall inadequate to express what I feel about my husband Hab- tamu(ye), and our two jewels - Fikir and Meti. Thank you for making my life complete in countless, love-filled ways. The days of 'good-byes' are finally over - thank the Lord!

69 References

Aboye, T. L., Clark, R. J., Burman, R., Roig, M. B., Craik, D. J., Göransson, U., 2011. Interlocking disulfides in circular proteins: toward efficient oxidative folding of cyclotides. Antioxid Redox Signal. 14, 77-86. Ahern, T. J., Klibanov, A. M., 1985. The mechanisms of irreversible enzyme inacti- vation at 100 °C. Science. 228, 1280-1284. Arnold, K., Bordoli, L., Kopp, J., Schwede, T., 2006. The SWISS-MODEL work- space: a web-based environment for protein structure homology modelling. Bio- informatics. 22, 195-201. Barbeta, B. L., Marshall, A. T., Gillon, A. D., Craik, D. J., Anderson, M. A., 2008. Plant cyclotides disrupt epithelial cells in the midgut of lepidopteran larvae. Proc Natl Acad Sci U S A. 105(4), 1221-1225 Basse, C. W., 2005. Dissecting defense-related and developmental transcriptional responses of maize during Ustilago maydis infection and subsequent tumor formation. Plant Physiol. 138, 1774-1784. Bischoff, R., Kolbe, H. V., 1994. Deamidation of asparagine and glutamine residues in proteins and peptides: structural determinants and analytical methodology. J Chromatogr B Biomed Appl. 662, 261-278. Blanchfield, J. T., Dutton, J. L., Hogg, R. C., Gallagher, O. P., Craik, D. J., Jones, A., Adams, D. J., Lewis, R. J., Alewood, P. F., Toth, I., 2003. Synthesis, structure elucidation, in vitro biological activity, toxicity, and Caco-2 cell permeability of lipophilic analogues of alpha-conotoxin MII. J Med Chem. 46, 1266-1272. Bokesch, H. R., Pannell, L. K., Cochran, P. K., Sowder, R. C., McKee, T. C., Boyd, M. R., 2001. A novel anti-HIV macrocyclic peptide from Palicourea condensata. J Nat Prod. 64, 249-250. Breidenstein, E. B. M., de la Fuente-Nunez, C., Hancock, R. E. W., 2011. Pseudo- monas aeruginosa: all roads lead to resistance. Trends Microbiol. 19, 419-426. Broussalis, A. M., Göransson, U., Coussio, J. D., Ferraro, G., Martino, V., Claeson, P., 2001. First cyclotide from Hybanthus (Violaceae). Phytochemistry. 58, 47-51. Burman, R., Gruber, C. W., Rizzardi, K., Herrmann, A., Craik, D. J., Gupta, M. P., Göransson, U., 2010. Cyclotide proteins and precursors from the genus Gloeo- spermum: filling a blank spot in the cyclotide map of Violaceae. Phytochemistry. 71, 13-20. Burman, R., Herrmann, A., Tran, R., Kivela, J. E., Lomize, A., Gullbo, J., Görans- son, U., 2011a. Cytotoxic potency of small macrocyclic knot proteins: structure- activity and mechanistic studies of native and chemically modified cyclotides. Org Biomol Chem. 9, 4306-4314.

70 Burman, R., Strömstedt, A. A., Malmsten, M., Göransson, U., 2011b. Cyclotide- membrane interactions: Defining factors of membrane binding, depletion and disruption. Biochim Biophys Acta. 1808, 2665-2673. Camarero, J. A., 2011. Novel peptide-based scaffolds for drug discovery. Curr Pharm Des. 17, 4224-4225. Camarero, J. A., Kimura, R. H., Woo, Y. H., Shekhtman, A., Cantor, J., 2007. Bio- synthesis of a fully functional cyclotide inside living bacterial cells. Chembio- chem. 8, 1363-1366. Campopiano, D. J., Clarke, D. J., Polfer, N. C., Barran, P. E., Langley, R. J., Govan, J. R., Maxwell, A., Dorin, J. R., 2004. Structure-activity relationships in defensin dimers: a novel link between beta-defensin tertiary structure and antimicrobial activity. J Biol Chem. 279, 48671-48679. Canell, J.P.R., 1994. How to approach the isolation of a natural product. In: Natural Products Isolation. Humana Press Inc. New Jersy. Carter, G. T., 2011. Natural products and Pharma 2011: strategic changes spur new opportunities. Nat Prod Rep. 28, 1783-1789. Cascales, L., Henriques, S. T., Kerr, M. C., Huang, Y. H., Sweet, M. J., Daly, N. L., Craik, D. J., 2011. Identification and Characterization of a New Family of Cell- penetrating Peptides: cyclic cell-penetrating peptides. J Biol Chem. 286, 36932- 36943. Chan, L. Y., Gunasekera, S., Henriques, S. T., Worth, N. F., Le, S. J., Clark, R. J., Campbell, J. H., Craik, D. J., Daly, N. L., 2011. Engineering pro-angiogenic peptides using stable, disulfide-rich cyclic scaffolds. Blood. 118, 6709-6717. Chiche, L., Heitz, A., Gelly, J. C., Gracy, J., Chau, P. T., Ha, P. T., Hernandez, J. F., Le-Nguyen, D., 2004. Squash inhibitors: from structural motifs to macrocyclic knottins. Curr Protein Pept Sci. 5, 341-349. Claeson, P., Göransson, U., Johansson, S., Luijendijk, T., Bohlin, L., 1998. Frac- tionation protocol for the isolation of polypeptides from plant biomass. J Nat Prod. 61, 77-81. Clardy, J., Walsh, C., 2004. Lessons from natural molecules. Nature. 432, 829-837. Clark, R. J., Daly, N. L., Craik, D. J., 2006. Structural plasticity of the cyclic- cystine-knot framework: implications for biological activity and drug design. Biochem J. 394, 85-93. Clark, R. J., Fischer, H., Dempster, L., Daly, N. L., Rosengren, K. J., Nevin, S. T., Meunier, F. A., Adams, D. J., Craik, D. J., 2005. Engineering stable peptide toxins by means of backbone cyclization: stabilization of the alpha-conotoxin MII. Proc Natl Acad Sci U S A. 102, 13767-13772. Clark, R. J., Jensen, J., Nevin, S. T., Callaghan, B. P., Adams, D. J., Craik, D. J., 2010. The engineering of an orally active conotoxin for the treatment of neuropathic pain. Angew Chem Int Ed Engl. 49, 6545-6548. Cohen, S. A., Michaud, D. P., 1993. Synthesis of a fluorescent derivatizing reagent, 6-aminoquinolyl-n-hydroxysuccinimidyl carbamate, and its application for the analysis of hydrolysate amino-acids via high-performance liquid- chromatography. Anal Biochem. 211, 279-287.

71 Colgrave, M. L., Craik, D. J., 2004. Thermal, chemical, and enzymatic stability of the cyclotide kalata B1: the importance of the cyclic cystine knot. Biochemistry. 43, 5965-5975. Colgrave, M. L., Kotze, A. C., Kopp, S., McCarthy, J. S., Coleman, G. T., Craik, D. J., 2009. Anthelmintic activity of cyclotides: In vitro studies with canine and human hookworms. Acta Trop. 109, 163-166. Colgrave, M. L., Poth, A. G., Kaas, Q., Craik, D. J., 2010. A new "era" for cyclotide sequencing. Biopolymers. 94, 592-601. Contreras, J., Elnagar, A. Y., Hamm-Alvarez, S. F., Camarero, J. A., 2011. Cellular uptake of cyclotide MCoTI-I follows multiple endocytic pathways. J Control Re- lease. 155, 134-143. Craik, D. J., Conibear, A. C., 2011. The chemistry of cyclotides. J Org Chem. 76, 4805-4817. Craik, D. J., Daly, N. L., Bond, T., Waine, C., 1999. Plant cyclotides: A unique family of cyclic and knotted proteins that defines the cyclic cystine knot structural motif. J Mol Biol. 294, 1327-1336. Creighton, T. E., 1980. Counting integral numbers of amino acid residues per polypeptide chain. Nature. 284, 487-489. Crooks, G. E., Hon, G., Chandonia, J. M., Brenner, S. E., 2004. WebLogo: a sequence logo generator. Genome Res. 14, 1188-1190. Daly, N. L., Clark, R. J., Plan, M. R., Craik, D. J., 2006. Kalata B8, a novel antiviral circular protein, exhibits conformational flexibility in the cystine knot motif. Biochem J. 393, 619-626. Daly, N. L., Rosengren, K. J., Craik, D. J., 2009. Discovery, structure and biological activities of cyclotides. Adv Drug Deliv Rev. 61, 918-930. Domon, B., Aebersold, R., 2006. Review - Mass spectrometry and protein analysis. Science. 312, 212-217. Dornenburg, H., 2010. Cyclotide synthesis and supply: from plant to bioprocess. Biopolymers. 94, 602-610. Dutton, J. L., Renda, R. F., Waine, C., Clark, R. J., Daly, N. L., Jennings, C. V., Anderson, M. A., Craik, D. J., 2004. Conserved Structural and sequence elements implicated in the processing of gene-encoded circular proteins. J Biol Chem. 279, 46858-46867. Felizmenio-Quimio, M. E., Daly, N. L., Craik, D. J., 2001. Circular proteins in plants: solution structure of a novel macrocyclic trypsin inhibitor from Momor- dica cochinchinensis. J Biol Chem. 276, 22875-22882. Gerlach, S. L., Burman, R., Bohlin, L., Mondal, D., Göransson, U., 2010a. Isolation, characterization, and bioactivity of cyclotides from the Micronesian plant Psy- chotria leptothyrsa. J Nat Prod. 73, 1207-1213. Gerlach, S. L., Rathinakumar, R., Chakravarty, G., Göransson, U., Wimley, W. C., Darwin, S. P., Mondal, D., 2010b. Anticancer and chemosensitizing abilities of cycloviolacin 02 from Viola odorata and psyle cyclotides from Psychotria leptothyrsa. Biopolymers. 94, 617-625.

72 Gillon, A. D., Saska, I., Jennings, C. V., Guarino, R. F., Craik, D. J., Anderson, M. A., 2008. Biosynthesis of circular proteins in plants. Plant J. 53, 505-515. Göransson, U., Broussalis, A. M., Claeson, P., 2003. Expression of the Viola cyclotides by LC/MS and MS-MS sequencing of intercysteine loops after introduction of charges and cleavage sites by aminoethylation. Anal Biochem. 318, 107- 117. Göransson, U., Craik, D. J., 2003. Disulfide mapping of the cyclotide kalata B1: chemical proof of the cyclic cystine knot motif. J Biol Chem. 278, 48188-48196. Göransson, U., Herrmann, A., Burman, R., Haugaard-Jonsson, L. M., Rosengren, K. J., 2009. The conserved glu in the cyclotide cycloviolacin O2 has a key structural role. Chembiochem. 10, 2354-2360. Göransson, U., Luijendijk, T., Johansson, S., Bohlin, L., Claeson, P., 1999. Seven novel macrocyclic polypeptides from Viola arvensis. J Nat Prod. 62, 283-286. Göransson, U., Sjögren, M., Svangård, E., Claeson, P., Bohlin, L., 2004a. Reversible antifouling effect of the cyclotide cycloviolacin O2 against barnacles. J Nat Prod. 67, 1287-1290. Göransson, U., Svangård, E., Claeson, P., Bohlin, L., 2004b. Novel strategies for isolation and characterization of cyclotides: the discovery of bioactive macrocyclic plant polypeptides in the violaceae. Curr Protein Pept Sci. 5, 317-329. Gran, L., 1970. Oxitocic principles found in Oldenlandia affinis. Medd Nor Farm Selsk. 32, 173-180. Gran, L., 1973a. Oxytocic principles of Oldenlandia affinis. Lloydia. 36, 174-178. Gran, L., 1973b. On the effect of a polypeptide isolated from "Kalata-kalata" (Old- enlandia affinis DC) on the oestrogen dominated uterus. Acta Pharmacol Toxi- col. 33, 400-408. Gran, L., Sandberg, F., Sletten, K., 2000. Oldenlandia affinis (R&S) DC - A plant containing uteroactive peptides used in African traditional medicine. J Ethno- pharmacol. 70, 197-203. Greenwood, K. P., Daly, N. L., Brown, D. L., Stow, J. L., Craik, D. J., 2007. The cyclic cystine knot miniprotein MCoTI-II is internalized into cells by macropi- nocytosis. Int J Biochem Cell Biol. 39, 2252-2264. Gruber, C. W., Cemazar, M., Clark, R. J., Horibe, T., Renda, R. F., Anderson, M. A., Craik, D. J., 2007. A novel plant protein-disulfide isomerase involved in the oxidative folding of cystine knot defense proteins. J Biol Chem. 282, 20435- 20446. Gruber, C. W., Cemazar, M., Heras, B., Martin, J. L., Craik, D. J., 2006. Protein disulfide isomerase: the structure of oxidative folding. Trends Biochem Sci. 31, 455-464. Gruber, C. W., Elliott, A. G., Ireland, D. C., Delprete, P. G., Dessein, S., Göransson, U., Trabi, M., Wang, C. K., Kinghorn, A. B., Robbrecht, E., Craik, D. J., 2008. Distribution and evolution of circular miniproteins in flowering plants. Plant Cell. 20, 2471-2483. Grundemann, C., Koehbach, J., Huber, R., Gruber, C. W., 2012. Do plant cyclotides have potential as immunosuppressant peptides? J Nat Prod. 75, 167-174.

73 Guex, N., Peitsch, M. C., 1997. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis. 18, 2714-2723. Guina, T., Yi, E. C., Wang, H., Hackett, M., Miller, S. I., 2000. A PhoP-regulated outer membrane protease of Salmonella enterica serovar typhimurium promotes resistance to alpha-helical antimicrobial peptides. J Bacteriol. 182, 4077-4086. Gullbo, J., Lövborg, H., Dhar, S., Lukinius, A., öberg, F., Nilsson, K., Björkling, F., Binderup, L., Nygren, P., Larsson, R., 2004. Development and characterization of two human tumor sublines expressing high-grade resistance to the cyanogua- nidine CHS 828. Anticancer Drugs. 15, 45-54. Gunasekera, S., Daly, N. L., Anderson, M. A., Craik, D. J., 2006. Chemical synthesis and biosynthesis of the cyclotide family of circular proteins. IUBMB Life. 58, 515-524. Gunasekera, S., Foley, F. M., Clark, R. J., Sando, L., Fabri, L. J., Craik, D. J., Daly, N. L., 2008. Engineering stabilized vascular endothelial growth factor-A antago- nists: synthesis, structural characterization, and bioactivity of grafted analogues of cyclotides. J Med Chem. 51, 7697-7704. Gustafson, K. R., Sowder, R. C., II, Henderson, L. E., Parsons, I. C., Kashman, Y., Cardellina, J. H., II, McMahon, J. B., Buckheit, R. W., Pannell, L. K., Boyd, M. R., 1994. Circulins A and B: Novel HIV-inhibitory macrocyclic peptides from the tropical tree Chassalia parvifolia. J Am Chem Soc. 116, 9337-9338. Gustafson, K. R., Walton, L. K., Sowder, R. C., Johnson, D. G., Pannell, L. K., Cardellina, J. H., Boyd, M. R., 2000. New circulin macrocyclic polypeptides from Chassalia parvifolia. J Nat Prod. 63, 176-178. Hallock, Y. F., Sowder, R. C., Pannell, L. K., Hughes, C. B., Johnson, D. G., Gulak- owski, R., Cardellina, J. H., Boyd, M. R., 2000. Cycloviolins A-D, anti-HIV macrocyclic peptides from Leonia cymosa. J Org Chem. 65, 124-128. Hekking, W. H. A., 1984. Studies on neotropical Violaceae the genus fusispermum. P K Ned Akad C Biol. 87, 121-130. Hekking, W. H. A., 1988. Studies on neotropical violaceae 2 arrangement of leaves inflorescences and branchlets in neotropical Rinorea. Flora. 180, 345-376. Hellman, B., Vaghef, H., Bostrom, B., 1995. The concepts of tail moment and tail inertia in the single cell gel electrophoresis assay. Mutat Res. 336, 123-131. Henriques, S. T., Craik, D. J., 2012. Importance of the Cell Membrane on the Mechanism of Action of Cyclotides. ACS Chem Biol. DOI: 10.1021/cb200395f Herrmann, A., Burman, R., Mylne, J. S., Karlsson, G., Gullbo, J., Craik, D. J., Clark, R. J., Göransson, U., 2008. The alpine violet, Viola biflora, is a rich source of cyclotides with potent cytotoxicity. Phytochemistry. 69, 939-952. Herrmann, A., Svangård, E., Claeson, P., Gullbo, J., Bohlin, L., Göransson, U., 2006. Key role of glutamic acid for the cytotoxic activity of the cyclotide cycloviolacin O2. Cell Mol Life Sci. 63, 235-245. Hong, J., 2011. Role of natural product diversity in chemical biology. Curr Opin Chem Biol. 15, 350-354.

74 Huang, Y. H., Colgrave, M. L., Daly, N. L., Keleshian, A., Martinac, B., Craik, D. J., 2009. The biological activity of the prototypic cyclotide kalata B1 is modu- lated by the formation of multimeric pores. J Biol Chem. 284, 20699-20707. Ireland, D. C., Colgrave, M. L., Nguyencong, P., Daly, N. L., Craik, D. J., 2006a. Discovery and characterization of a linear cyclotide from Viola odorata: implications for the processing of circular proteins. J Mol Biol. 357, 1522-1535. Ireland, D. C., Colgrave, M. L., Craik, D. J., 2006b. A novel suite of cyclotides from Viola odorata: sequence variation and the implications for structure, function and stability. Biochem J. 400, 1-12. Izadpanah, A., Gallo, R. L., 2005. Antimicrobial peptides. J Am Acad Dermatol. 52, 381-390. Jennings, C., West, J., Waine, C., Craik, D. J., Anderson, M., 2001. Biosynthesis and insecticidal properties of plant cyclotides: The cyclic knotted proteins from Oldenlandia affinis. Proc Natl Acad Sci U S A. 98, 10614-10619. Kamimori, H., Hall, K., Craik, D., Aguilar, M., 2005. Studies on the membrane interactions of the cyclotides kalata B1 and kalata B6 on model membrane systems by surface plasmon resonance. Anal Biochem. 337, 149-153. Kimura, R. H., Tran, A. T., Camarero, J. A., 2006. Biosynthesis of the cyclotide Kalata B1 by Using Protein Splicing. Angew Chem Int Ed Engl. 45, 973-976. Kinghorn, A. D., Pan, L., Fletcher, J. N., Chai, H., 2011. The relevance of higher plants in lead compound discovery programs. J Nat Prod. 74, 1539-1555. Kingston, D. G., 2011. Modern natural products drug discovery and its relevance to biodiversity conservation. J Nat Prod. 74, 496-511. Koehn, F. E., Carter, G. T., 2005. The evolving role of natural products in drug discovery. Nat Rev Drug Discov. 4, 206-220. Kuang, Z., Hao, Y., Walling, B. E., Jeffries, J. L., Ohman, D. E., Lau, G. W., 2011. Pseudomonas aeruginosa elastase provides an escape from phagocytosis by de- grading the pulmonary surfactant protein-A. PLoS One. 6, e27091. Kuroda, M., Ohta, T., Uchiyama, I., Baba, T., Yuzawa, H., Kobayashi, I., Cui, L., Oguchi, A., Aoki, K., Nagai, Y., Lian, J., Ito, T., Kanamori, M., Matsumaru, H., Maruyama, A., Murakami, H., Hosoyama, A., Mizutani-Ui, Y., Takahashi, N. K., Sawano, T., Inoue, R., Kaito, C., Sekimizu, K., Hirakawa, H., Kuhara, S., Goto, S., Yabuzaki, J., Kanehisa, M., Yamashita, A., Oshima, K., Furuya, K., Yoshino, C., Shiba, T., Hattori, M., Ogasawara, N., Hayashi, H., Hiramatsu, K., 2001. Whole genome sequencing of meticillin-resistant Staphylococcus aureus. Lancet. 357, 1225-1240. Laarman, A. J., Ruyken, M., Malone, C. L., van Strijp, J. A., Horswill, A. R., Rooi- jakkers, S. H., 2011. Staphylococcus aureus metalloprotease aureolysin cleaves complement C3 to mediate immune evasion. J Immunol. 186, 6445-6453. Larsson, R., Nygren, P., 1989. A rapid fluorometric method for semiautomated determination of cytotoxicity and cellular proliferation of human tumor cell lines in microculture. Anticancer Res. 9, 1111-1119.

75 Leta Aboye, T., Clark, R. J., Craik, D. J., Göransson, U., 2008. Ultra-stable peptide scaffolds for protein engineering-synthesis and folding of the circular cystine knotted cyclotide cycloviolacin O2. Chembiochem. 9, 103-113. Lewis, R. J., 2009. Conotoxin venom peptide therapeutics. Adv Exp Med Biol. 655, 44-48. Lindhagen, E., Nygren, P., Larsson, R., 2008. The fluorometric microculture cytotoxicity assay. Nat Protoc. 3, 1364-1369. Lindholm, P., Göransson, U., Johansson, S., Claeson, P., Gullbo, J., Larsson, R., Bohlin, L., Backlund, A., 2002. Cyclotides: A novel type of cytotoxic agents. Mol Cancer Ther. 1, 365-369. Loffet, A., 2002. Peptides as drugs: is there a market? J Pept Sci. 8, 1-7. Maemoto, A., Qu, X., Rosengren, K. J., Tanabe, H., Henschen-Edman, A., Craik, D. J., Ouellette, A. J., 2004. Functional analysis of the alpha-defensin disulfide array in mouse cryptdin-4. J Biol Chem. 279, 44188-44196. Mulvenna, J. P., Sando, L., Craik, D. J., 2005. Processing of a 22 kDa precursor protein to produce the circular protein tricyclon A. Structure 13, 691-701. Mulvenna, J. P., Wang, C., Craik, D. J., 2006a. CyBase: a database of cyclic protein sequence and structure. Nucleic Acids Res. 34, D192-194. Mulvenna, J. P., Mylne, J. S., Bharathi, R., Burton, R. A., Shirley, N. J., Fincher, G. B., Anderson, M. A., Craik, D. J., 2006b. Discovery of cyclotide-like protein sequences in graminaceous crop plants: ancestral precursors of circular proteins? Plant Cell. 18, 2134-2144. Nelson, A. L., Dhimolea, E., Reichert, J. M., 2010. Development trends for human monoclonal antibody therapeutics. Nat Rev Drug Discov. 9, 767-774. Newman, D. J., Cragg, G. M., 2012. Natural Products As Sources of New Drugs over the 30 Years from 1981 to 2010. J Nat Prod. DOI: 10.1021/np200906s. Nguyen, G. K., Zhang, S., Wang, W., Wong, C. T., Nguyen, N. T., Tam, J. P., 2011a. Discovery of a linear cyclotide from the bracelet subfamily and its disulfide mapping by top-down mass spectrometry. J Biol Chem. 286, 44833-44844. Nguyen, G. K., Zhang, S., Nguyen, N. T., Nguyen, P. Q., Chiu, M. S., Hardjojo, A., Tam, J. P., 2011b. Discovery and characterization of novel cyclotides originated from chimeric precursors consisting of albumin-1 chain a and cyclotide domains in the Fabaceae family. J Biol Chem. 286, 24275-24287. O'Keefe, B. R., 2001b. Biologically active proteins from natural product extracts. J Nat Prod. 64, 1373-1381. Ostling, O., Johanson, K. J., 1984. Microelectrophoretic study of radiation-induced DNA damages in individual mammalian cells. Biochem Biophys Res Commun. 123, 291-298. Pawlik, J. R., 2011. The Chemical Ecology of Sponges on Caribbean Reefs: Natural Products Shape Natural Systems. Bioscience. 61, 888-898. Peschel, A., Sahl, H. G., 2006. The co-evolution of host cationic antimicrobial peptides and microbial resistance. Nat Rev Microbiol. 4, 529-536.

76 Plan, M. R., Saska, I., Cagauan, A. G., Craik, D. J., 2008. Backbone cyclised peptides from plants show molluscicidal activity against the rice pest Pomacea ca- naliculata (golden apple snail). J Agric Food Chem. 56, 5237-5241. Poth, A. G., Colgrave, M. L., Philip, R., Kerenga, B., Daly, N. L., Anderson, M. A., Craik, D. J., 2011. Discovery of cyclotides in the fabaceae plant family provides new insights into the cyclization, evolution, and distribution of circular proteins. ACS Chem Biol. 6, 345-355. Pränting, M., Lööv, C., Burman, R., Göransson, U., Andersson, D. I., 2010. The cyclotide cycloviolacin O2 from Viola odorata has potent bactericidal activity against Gram-negative bacteria. J Antimicrob Chemother. 65, 1964-1971. Rosengren, K. J., Daly, N. L., Plan, M. R., Waine, C., Craik, D. J., 2003. Twists, knots, and rings in proteins. Structural definition of the cyclotide framework. J Biol Chem. 278, 8606-8616. Rosengren, K. J., Göransson, U., Otvos, L., Jr., Craik, D. J., 2004. Cyclization of pyrrhocoricin retains structural elements crucial for the antimicrobial activity of the native peptide. Biopolymers. 76, 446-458. Rozek, A., Powers, J. P., Friedrich, C. L., Hancock, R. E., 2003. Structure-based design of an indolicidin peptide analogue with increased protease stability. Bio- chemistry. 42, 14130-14138. Samuelsson, G., Kyerematen, G., Farah, M. H., 1985. Preliminary chemical characterization of pharmacologically active compounds in aqueous plant extracts. J Ethnopharmacol. 14, 193-201. Sandberg, F., 1965. Étude sur les plantes médicinales et toxiques d'Afrique équatori- ale. Étude sur les plantes médicinales et toxiques d'Afrique équatoriale, vol. Tome III, Fascicule 1. Cahièrs de la Maboké. 12 Rue de Buffon, Paris, p. 27. Sasidharan, S., Chen, Y., Saravanan, D., Sundram, K. M., Yoga Latha, L., 2011. Extraction, isolation, and characterization of bioactive compounds from plant extracts. Afr J Tradit Complement Altern Med. 8, 1-10. Saska, I., Gillon, A. D., Hatsugai, N., Dietzgen, R. G., Hara-Nishimura, I., Ander- son, M. A., Craik, D. J., 2007. An asparaginyl endopeptidase mediates in vivo protein backbone cyclization. J Biol Chem. 282, 29721-29728. Schöpke, T., Hasan Agha, M. I., Kraft, R., Otto, A., Hiller, K., 1993. Hämolytisch aktive komponenten aus Viola tricolor L. und Viola arvensis Murray. Sci Pharm. 61, 145-153. Schwede, T., Kopp, J., Guex, N., Peitsch, M. C., 2003. SWISS-MODEL: an automated protein homology-modeling server. Nucleic Acids Res. 31, 3381-3385. Shen, J. H., Xu, X. Y., Cheng, F., Liu, H., Luo, X. M., Shen, J. K., Chen, K. X., Zhao, W. M., Shen, X., Jiang, H. L., 2003. Virtual screening on natural products for discovering active compounds and target information. Curr Med Chem. 10, 2327-2342. Shenkarev, Z. O., Nadezhdin, K. D., Lyukmanova, E. N., Sobol, V. A., Skjeldal, L., Arseniev, A. S., 2008. Divalent cation coordination and mode of membrane interaction in cyclotides: NMR spatial structure of ternary complex Kalata B7/Mn2+/DPC micelle. J Inorg Biochem. 102, 1246-1256.

77 Shenkarev, Z. O., Nadezhdin, K. D., Sobol, V. A., Sobol, A. G., Skjeldal, L., Ar- seniev, A. S., 2006. Conformation and mode of membrane interaction in cyclotides. Spatial structure of kalata B1 bound to a dodecylphosphocholine micelle. Febs J. 273, 2658-2672. Sieprawska-Lupa, M., Mydel, P., Krawczyk, K., Wojcik, K., Puklo, M., Lupa, B., Suder, P., Silberring, J., Reed, M., Pohl, J., Shafer, W., McAleese, F., Foster, T., Travis, J., Potempa, J., 2004. Degradation of human antimicrobial peptide LL-37 by Staphylococcus aureus-derived proteinases. Antimicrob Agents Chemother. 48, 4673-4679. Simonsen, S. M., Sando, L., Ireland, D. C., Colgrave, M. L., Bharathi, R., Görans- son, U., Craik, D. J., 2005. A continent of plant defense peptide diversity: cyclotides in Australian Hybanthus (Violaceae). Plant Cell. 17, 3176-3189. Singh, N. P., McCoy, M. T., Tice, R. R., Schneider, E. L., 1988. A simple technique for quantitation of low levels of DNA damage in individual cells. Exp Cell Res. 175, 184-191. Sletten, K., Gran, L., 1973. Some molecular properties of kalatapeptide B-1. Medd Nor Farm Selsk. 35, 69-82. Smith, A. B., Daly, N. L., Craik, D. J., 2011. Cyclotides: a patent review. Expert Opin Ther Pat. 21, 1657-1672. Solon, E. G., 2012. Use of Radioactive Compounds and Autoradiography to Deter- mine Drug Tissue Distribution. Chem Res Toxicol. DOI: 10.1021/tx200509f. Solon, E. G., Schweitzer, A., Stoeckli, M., Prideaux, B., 2010. Autoradiography, MALDI-MS, and SIMS-MS imaging in pharmaceutical discovery and development. AAPS J. 12, 11-26. Svangård, E., Burman, R., Gunasekera, S., Lövborg, H., Gullbo, J., Göransson, U., 2007. Mechanism of action of cytotoxic cyclotides: cycloviolacin O2 disrupts lipid membranes. J Nat Prod. 70, 643-647. Svangård, E., Göransson, U., Hocaoglu, Z., Gullbo, J., Larsson, R., Claeson, P., Bohlin, L., 2004. Cytotoxic cyclotides from Viola tricolor. J Nat Prod. 67, 144- 147. Svangård, E., Göransson, U., Smith, D., Verma, C., Backlund, A., Bohlin, L., Claeson, P., 2003. Primary and 3-D modelled structures of two cyclotides from Viola odorata. Phytochemistry. 64, 135-142. Tam, J. P., Lu, Y. A., Yang, J. L., Chiu, K. W., 1999. An unusual structural motif of antimicrobial peptides containing end-to-end macrocycle and cystine-knot disulfides. Proc Natl Acad Sci U S A. 96, 8913-8918. Terlau, H., Olivera, B. M., 2004. Conus venoms: a rich source of novel ion channel- targeted peptides. Physiol Rev. 84, 41-68. Thongyoo, P., Roque-Rosell, N., Leatherbarrow, R. J., Tate, E. W., 2008. Chemical and biomimetic total syntheses of natural and engineered MCoTI cyclotides. Org Biomol Chem. 6, 1462-1470. Tice, R. R., Agurell, E., Anderson, D., Burlinson, B., Hartmann, A., Kobayashi, H., Miyamae, Y., Rojas, E., Ryu, J. C., Sasaki, Y. F., 2000. Single cell gel/comet as-

78 say: guidelines for in vitro and in vivo genetic toxicology testing. Environ Mol Mutagen. 35, 206-221. Trabi, M., Craik, D. J., 2002. Circular proteins - no end in sight. Trends Biochem Sci. 27, 132-138. Trabi, M., Craik, D. J., 2004. Tissue-specific expression of head-to-tail cyclized miniproteins in Violaceae and structure determination of the root cyclotide Viola hederacea root cyclotide1. Plant Cell. 16, 2204-2216. Trabi, M., Svangard, E., Herrmann, A., Göransson, U., Claeson, P., Craik, D. J., Bohlin, L., 2004. Variations in cyclotide expression in viola species. J Nat Prod. 67, 806-810. Trauger, S. A., Webb, W., Siuzdak, G., 2002. Peptide and protein analysis with mass spectrometry. Spectrosc-Int J. 16, 15-28. Vlieghe, P., Lisowski, V., Martinez, J., Khrestchatisky, M., 2010. Synthetic therapeutic peptides: science and market. Drug Discov Today. 15, 40-56. Wang, C. K., Hu, S. H., Martin, J. L., Sjogren, T., Hajdu, J., Bohlin, L., Claeson, P., Goransson, U., Rosengren, K. J., Tang, J., Tan, N. H., Craik, D. J., 2009. Com- bined X-ray and NMR analysis of the stability of the cyclotide cystine knot fold that underpins its insecticidal activity and potential use as a drug scaffold. J Biol Chem. 284, 10672-10683. Wang, C. K., Kaas, Q., Chiche, L., Craik, D. J., 2008. CyBase: a database of cyclic protein sequences and structures, with applications in protein discovery and engineering. Nucleic Acids Res. 36, D206-210. Weintraub, S. J., Deverman, B. E., 2007. Chronoregulation by asparagine deamidation. Sci. STKE, re7. Witherup, K. M., Bogusky, M. J., Anderson, P. S., Ramjit, H., Ransom, R. W., Wood, T., Sardana, M., 1994. Cyclopsychotride A, a biologically active, 31- residue cyclic peptide isolated from Psychotria longipes. J Nat Prod. 57, 1619- 1625. Wysocki, V. H., Resing, K. A., Zhang, Q. F., Cheng, G. L., 2005. Mass spectrometry of peptides and proteins. Methods. 35, 211-222. Yu, Q. T., Lehrer, R. I., Tam, J. P., 2000. Engineered salt-insensitive alpha- defensins with end-to-end circularized structures. J Biol Chem. 275, 3943-3949. Zhang, J., Liao, B., Craik, D. J., Li, J. T., Hu, M., Shu, W. S., 2009. Identification of two suites of cyclotide precursor genes from metallophyte Viola baoshanensis: cDNA sequence variation, alternative RNA splicing and potential cyclotide diversity. Gene. 431, 23-32.

79 Appendices

Appendix I List of Rubiaceae spp. collected from the Jimma-Bonga area in Ethiopia and tested for cylotide content. The plants were extracted with buffer solution of medium polarity, the extracts were then analyzed by LC-MS with a gradient of 10 to 60% MeCN in 0.1% HCOOH. There were no late eluting peaks with molecular weight ranging between 2500 and 4000 Da, which indicated that the spp are devoid of cyclotides.

1. Galineria coffeoides Del. 2. Gardenia ternifolia Schumach. 3. Kohoutia coccinea Royle 4. Kohoutia platyphylla Bremek 5. Oldenlandia herbacea (L.) Roxb. 6. Pentas lanceolata subsp. quartiniona 7. Rothmannia urcelliformis (Hiern) Robyns 8. Spermacoce sphaerostigma (A. Rich) Votke 9. Spermacoceae spp. 10. Vangueria apiculata K.Schum.

80 Appendix II Phylogenetic tree of the Angiosperms. Cyclotides have been isolated from the plant families Rubiaceae, Violaceae and Fabaceae. The families belong to phylogeneti- cally distant orders, i.e., Gentianales (Rubiaceae), Malpighiales (Violaceae) and Fabales (Fabaceae) indicated in the phylogenetic tree by **. Cyclotide-like genes have been isolated from Poaceae which belongs to the order Poales marked +in the tree. The phylogenic tree is as redrawn by prof. Anders Backlund after APG II.