|Home | About | Journals | Submit | Contact Us | Français|
Cyclotides are a growing family of large plant-derived backbone-cyclized polypeptides (≈30 amino acids long) that share a disulfide-stabilized core characterized by an unusual knotted structure. Their unique circular backbone topology and knotted arrangement of three disulfide bonds makes them exceptionally stable to thermal, chemical, and enzymatic degradation compared to other peptides of similar size. Currently more than 100 sequences of different cyclotides have been characterized and the number is expected to increase dramatically in the coming years. Considering their stability, biological activities and ability to cross the cell membrane, cyclotides can be exploited to develop new peptide-based drugs with high potential for success. The cyclotide scaffold can be engineered or evolved using molecular evolution to inhibit protein-protein interactions implicated in cancer and other human diseases, or design new antimicrobial. The present review reports the biological diversity and therapeutic potential of natural and engineered cyclotides.
Head-to-tail or backbone-cyclized peptides are present throughout nature from bacteria to animals. Bacteria and fungi express numerous backbone-cyclized peptides that are currently in use as therapeutic agents . For example, cyclosporin A is a fungal peptide with potent immunosuppressive properties and is used to treat organ transplant patients . Daptomycin, is a 13-amino acid cyclic lipopeptide with a decanoyl side chain isolated from Streptomyces roseosporus that has recently been approved to treat infections against Gram-positive organisms, including multi-resistant strains . In animals, the only known circular peptides are θ-defensins, which are expressed in blood leukocytes and bone marrow of Old World monkeys [4, 5]. θ-Defensins are antimicrobial peptides with broad-spectrum activities against bacteria, fungi, and viruses [6-8]. Backbone cyclized peptides have been also found in plants . Sunflower trypsin inhibitor 1 (SFTI-1) for example is a bicyclic 14-residue long peptide found in sunflower seeds. SFTI-1 is the most potent known naturally occurring Bowman-Birk trypsin inhibitor . Cyclotides, a novel family of small globular backbone-cyclized micro-proteins (≈ 30-residues long), are also naturally found in plants. Here, we review the properties of cyclotides and the latest developments in the use of the cyclotide scaffold to design novel peptide-based therapeutics.
Cyclotides are small globular micro-proteins with a unique head-to-tail cyclized backbone, which is stabilized by three disulfide bonds (Fig. 1). Currently, over 140 sequences have been identified in the plant species Rubiaceae, Violaceae, and Cucurbitaceae . Natural cyclotides have various activities including insecticidal [12, 13], uterotonic , anti-HIV , antimicrobial [16, 17], antitumor , antihelminthic [19, 20] and have been reported to cross cell membranes . Their insecticidal and antihelminthic properties suggest that they may function as defense molecules in plants.
Cyclotides share a unique head-to-tail circular knotted topology of three disulfide bridges, with one disulfide penetrating through a macrocycle formed by the two other disulfides and inter-connecting peptide backbones, forming what is called a cyclic cystine knot (CCK) motif (Fig. 1). The CCK topology is responsible for the high stability of cyclotides to enzymatic, thermal and chemical degradation . The cyclotide family is divided into three structurally distinct families, Möbius, bracelet, and trypsin inhibitor subfamilies (Fig. 1). Möbius cyclotides are distinguished from bracelet cyclotides by the presence of a cis-Pro residue in loop 5. Trypsin inhibitor cyclotides have very different primary structures from Möbius and bracelet cyclotides, but retain the conserved cystine knot motif. Trypsin inhibitor cyclotides share a high sequence homology with related cystine-knot trypsin inhibitors found in squash such as EETI-II (Ecballium elaterium trypsin inhibitor II), and in fact can be considered cyclized homologs of these protease inhibitors. Thus, cyclotides can be considered natural combinatorial peptide libraries structurally constrained by the cystine-knot scaffold  and head-to-tail cyclization but are permissive of hypermutation of essentially all residues with the exception of the strictly conserved cysteines that comprise the knot [24-27].
Hence, cyclotides form a unique family of structurally-related peptides that possess remarkable stability due to the cystine knot, a small size making them readily accessible to chemical synthesis, and an excellent tolerance to sequence variations. Moreover, the first cyclotide to be discovered, kalata B1, is an orally effective uterotonic . Intriguingly, the cyclotide MCoTI-II has also been shown to cross cell membranes through macropinocytosis . We have also recently found that MCoTI-I has similar cellular-uptake properties to MCoTI-II (unpublished results). All of these features make cyclotides ideal tools for the development of a total novel class of peptide-based therapeutics.
Cyclotides are ribosomally synthesized as precursor proteins, which consist of an endoplasmic reticulum (ER)-targeting sequence, a pro-region, a highly conserved N-terminal repeat (NTR) region, a mature cyclotide domain, and a C-terminal tail (Fig. 2). The combined NTR-cyclotide segment may contain one copy of the cyclotide sequence or there may be multiple copies of the same or different cyclotide sequences separated by additional NTR sequences. The precursor undergoes post-translational processing to generate a circular peptide by a mechanism that has not been completely elucidated yet [28, 29]. It has been hypothesized that a conserved Asn (or Asp) at the C-terminal cleavage site may be a recognition site by asparaginyl endoproteinase (AEP) for cyclization of the peptide in plants [28, 29]. AEP has been shown to be involved in the post-translational processing of concanavalin A from the jackbean , and therefore, it is possible this enzyme may be involved in the cyclization of other plant peptides. Studies using transgenic plants that express a cyclotide precursor have demonstrated the involvement of AEP and requirement for the asparagine residue in the cyclotide sequence [28, 29]. The authors showed that inhibition of AEP led to a decrease in the amount of cyclic product and an accumulation of linear peptides that were transiently expressed . In addition, a complementary study showed that mutation of the asparagine residue or truncation of the conserved C-terminal tripeptide in transgenic plants resulted in no circular peptide production .
The natural function of cyclotides appears to be in protection of plants against insects [12, 13], nematodes [20, 25], and mollusks . Studies have demonstrated that cyclotides can suppress the growth and development of insect and nematode larvae. Various other studies have also shown cyclotides have antimicrobial, hemolytic, uterotonic, and anti-HIV activities. Much of these activities likely involve interaction of the cyclotide with membranes, although the mechanism of action is not totally well understood.
The first cyclotide discovered, kalata B1, was identified in the plant Oldenlandia affinis in central Africa in the 1960’s . This plant was used by the natives to make a tea extract that was used to accelerate childbirth during labor [14, 33]. The main active ingredient in the tea extract was found to be a peptide that was named kalata B1, after the local name for the native medicine. The uterotonic properties of kalata B1 indicated that the peptide was orally bioavailable. The complete sequence, Cys-knotted arrangement, and cyclic nature of kalata B1 was determined 25 years after its original discovery . Since the discovery of kalata B1, many more related cyclotides have been discovered and found to have various biological activities [35, 36].
Cyclotides were initially hypothesized to have antimicrobial activities based on the presence of hydrophilic and hydrophobic patches, which give an amphipathic character similar to classical antimicrobial peptides. The antimicrobial activities of cyclotides have been reported by two groups with conflicting results on the potency of kalata B1 against Escherichia coli and Staphylococcus aureus. In one study, kalata B1 was active against S. aureus, but not E. coli , and in the second study, the peptide had the reverse effect . This is likely due to the technical differences in the experiments. Although kalata cyclotides are amphipathic, the overall charge is close to zero at neutral pH, making it unlikely that they interact with bacterial membranes electrostatically similar to classical cationic antimicrobial peptides . Further studies are necessary to investigate the mechanisms of antimicrobial action given the growing occurrence of antibiotic resistance by microorganisms.
The anti-HIV properties of cyclotides have been extensively studied [15, 38-40]. They appear to mainly act by inhibiting viral entry into host cells as studies have shown a dose-dependent increase in cytoprotection . This suggests the peptides may block binding or fusion of the virus, but the mechanism remains unclear. In general, cyclotide bioactivities appear to involve interactions with membranes and, therefore, this may be a mechanism for anti-HIV activities, by preventing fusion of the viral and host cell membranes. Studies have shown cyclotides can bind to model lipid membranes by surface plasmon resonance , and that binding occurs mainly through the peptide hydrophobic patches exposed on the surface [42-44]. This suggests membrane binding may be one mode for cyclotide activity against microorganisms.
Cyclotide interactions with membranes have also been suggested as their mechanism for cytotoxic activity. Studies have demonstrated antitumor activities of cyclotides, which were selective against cancer cell lines and solid tumors compared to normal mammalian cells [18, 45, 46]. Cancer cells differ from normal cells in the lipid and glycoprotein composition, which alters the overall net charge. The different potencies between cyclotide cytotoxicity are related to the three-dimensional structure as well as specific amino acid residues within the sequence [46, 47].
In addition to having antimicrobial and antitumor activities, some cyclotides have been found to cause extensive hemolysis of human and rat erythrocytes [16, 40, 48]. The cyclotide kalata B1 has strong hemolytic activity, although this can be eliminated by mutation to Ala of any one of eight residues located in the bioactive face of the molecule . A more recent study has also shown that the hemolytic activity of kalata B1 could also be reduced or completely eliminated by mutation to Lys of any of the residues involved in either the bioactive or hydrophobic faces of kalata B1 . The same study showed that the hemolytic, insecticidal and nematocidal activities of the different kalata B1 mutants were correlated, indicating there may be a common mechanism involving a cyclotide-membrane interaction . The same authors have also shown recently the all D-analogue of kalata had similar nematicidal activity than the native peptide, thus corroborating that the biological activity is mediated by a cellular receptor . In contrast, other cyclotides such as MCoTI-cyclotides have shown little or no hemolytic activity, thus demonstrating the diversity of cyclotide properties.
Another cyclotide with interesting biological activity is cyclopsychotride (Cpt) A. Cpt A is natural cyclotide obtained from the organic extract of the tropical plant Psychotropia longipes that has been reported to have neurotensin inhibition properties . Cpt A was able to inhibit neurotensin binding to its receptor to HT-29 cell membranes with an IC50 ≈ 3 μM and increase intracellular Ca2+ levels in a concentration-dependent manner, which could not blocked by neurotensin antagonists . Cpt A, however, showed a similar activity in two unrelated cell lines that did not express neurotensin receptors indicating that the mechanism of action is unlikely to be mediated through an interaction with the neurotensin receptor .
Cyclotides are small peptides, approximately 30 amino acids long, and therefore can be readily synthesized by chemical methods using solid-phase peptide synthesis . Chemical synthesis using a solid-phase approach has been utilized to generate native cyclotide structures as well as grafted analogues [52-56]. This method uses an intramolecular native chemical ligation , in which the peptide sequence contains an N-terminal cysteine and an α-thioester group at the C-terminus [58-60]. Both tert-butyloxyxarbonyl (Boc)- and 9-fluorenyloxycarbonyl (Fmoc)-based chemistries have been used to incorporate C-terminal thioesters during chain assembly (Boc) [61-63] or using a safety-catch based linkers (Fmoc) [60, 64-67]. Once the peptide is cleaved from the resin, both cyclization and folding are carried out in a single pot reaction.
Cyclotides have also been produced recombinantly in bacteria through intramolecular native chemical ligation (see above) by using a modified protein splicing unit or intein (Fig. 3) (see reference  for a recent review). This method can generate folded cyclotides either in vivo or in vitro using standard bacterial expression systems [26, 69, 70]. Inteins are internal self-processing domains that undergo post-translational processing to splice together flanking external domains (exteins) . The approach uses a modified intein fused to the C-terminus of the cyclotide sequence to allow the formation of an α-thioester at the C-terminus of recombinant polypeptides. To obtain the required N-terminal cysteine for cyclization, the peptide can be expressed with an N-terminal leading peptide signal, which can be cleaved either in vivo or in vitro by proteolysis or auto-proteolysis . The simplest way to accomplish this is to introduce a Cys downstream of the initiating Met residue. Once the translation step is completed, the endogeneous methionyl aminopeptidases (MAP) removes the Met residue, thereby generating in vivo an N-terminal Cys residue [72-76]. The N-terminal Cys can then capture the reactive thioester in an intramolecular fashion to form a backbone-cyclized polypeptide (Fig. 3). Additional methods to generate an N-terminal cysteine have used exogenous proteases to cleave the leading signal after purification or in vivo by co-expressing the protease . For example, the protease Factor Xa has been used to remove an N-terminal recognition sequence prior to a cysteine residue [59, 78]. Other proteases that have been used for this task include ubiquitin C-terminal hydrolase [79, 80], tobacco etch virus (TEV) protease , enterokinase  and thrombin . The N-terminal pelB leader sequence has been used recently to direct newly synthesized fusion proteins to the E. coli periplasmic space where the corresponding endogenous leader peptidases [83, 84] can generate the desired N-terminal cysteine-containing protein fragment . Besides proteases, protein splicing has also been used to produce recombinant N-terminal Cys-containing polypeptides. Some inteins can be modified in such a way that cleavage at the C-terminal splice junction can be accomplished in a pH- and temperature-dependent fashion [86-88].
Intein-mediated backbone cyclization of polypeptides has also been recently used for the biosynthesis of the Bowman-Birk inhibitor SFTI-1 . The biosynthesis of other cyclic peptides such as backbone-cyclized α-defensins and naturally occurring θ-defensins is currently underway in our laboratory.
Another approach to generate cyclic peptides in vivo is by protein trans-splicing. This approach utilizes a self-processing intein that is split into two fragments, an N-intein and a C-intein. This method has not been applied yet for the biosynthesis of cyclotides, but has been used to produce other natural cyclic peptides and genetically-encoded libraries of small cyclic peptides [90, 91]. It should be noted, however, that these systems require the presence of specific amino acid residues at both intein-extein junctions for efficient protein splicing to occur [90, 92, 93].
The unique properties associated with the cyclotide scaffold make them extremely valuable tools in drug discovery. There are several studies that have used the cyclotide molecular scaffold to graft peptide sequences and to generate libraries for the purpose of engineering cyclotides with novel biological functions (Fig. 4).
The plasticity of the cyclotide framework was first demonstrated by substituting hydrophobic residues in loop 5 of kalata B1 with polar and charged residues . The mutated cyclotides retained the native fold of kalata B1, but were no longer hemolytic . This showed that cyclotides were amendable to sequence changes, and interestingly, can be modified to change their biological functions.
The potential of grafted cyclotides was first demonstrated in a study aimed to develop novel anticancer peptide-based therapeutics . In this work a peptide antagonist of angiogenesis was grafted into various loops of the kalata B1 scaffold . The grafted cyclotide containing the vascular endothelial growth factor A (VEGF-A) antagonist sequence in loop 3 was found to adopt a CCK native fold and was biologically active at low micromolar concentrations. Additionally, the grafted cyclotide showed increased resistance to degradation in human serum. This study demonstrates the possibility of using the cyclotide scaffold to stabilize bioactive peptide epitopes, which may normally get degraded.
The utility of the cyclotide scaffold in drug design has also been recently shown by engineering non-native activities into the cyclotide MCoTI-II. MCoTI-II is a naturally occurring trypsin inhibitor (Ki ≈ 20 pM, which is the dissociation constant for inhibitor binding) found in the seeds of Momordica cochinchinensis, a tropical plant from the squash family. Mutation of the P1 residue in the active loop (loop 1, see Fig, 1) of the cyclotide produced several MCoTI-II analogs with different specificities towards different proteases . Interestingly, several analogs showed activity against the foot-and-mouth-disease virus (FMDV) 3C protease, a Cys protease key for viral replication, in the low micromolar range . This is the first reported peptide-based inhibitor for this protease and although the potency was relatively low, this study demonstrates the potential of using MCoTI-based cyclotides for designing novel protease inhibitors .
In a more recent study, the same authors also generated inhibitors of the serine proteases β-tryptase and human leukocyte elastase (HLE) using the backbone of MCoTI-II . β-Tryptase is implicated in allergic and inflammatory disorders, and HLE has been associated with respiratory and pulmonary disorders. Replacing the P1 residue in loop 1 produced several MCoTI-II mutants (K6A and K6V) with activity against HLE with Ki values of 20-30 nM  and Ki values against trypsin above 1 μM. Removal of the SDGG peptide segment in loop 6 yielded a ß-tryptase inhibitor with a Ki ≈ 10 nM without significantly altering the three-dimensional structure as determined by NMR . The authors hypothesized that deletion of the aspartic acid residue in MCoTI-II should improve activity by removal of repulsive electrostatic interactions with β-tryptase thus improving the inhibitory constant against trypsin 160-fold when compared to the wild-type MCoTI-II.
In addition to displaying biological activities, MCoTI-based peptides have also been shown to cross cell membranes in macrophage and breast cancer cell lines through macropinocytosis . We have also found that grafting of a helix region from the molluscum contagiosum virus (MCV) FLICE-inhibitory protein (FLIP) into loop 6 of MCoTI-I yielded a folded cyclotide able to cross cell membranes and trigger apoptosis of virally infected cells (unpublished results), thus indicating that loop 6 may be used for grafting purposes without affecting cellular-uptake.
MCoTI cyclotides share a high sequence homology with related cystine-knot trypsin inhibitors found in squash, and could be considered cyclized homologs of these protease inhibitors. Squash cystine-knot trypsin inhibitors have also successfully been used to graft biological activities. For example, the RGD sequence was grafted into loop 1 of EETI-II yielding an EETI-II analog with platelet inhibitory activity . The engineered proteins were much more potent in inhibiting platelet aggregation than the linear grafted peptides, thus highlighting the importance of grafting an epitope into a stable peptide-scaffold. These highly stable peptides could have clinical use for the treatment of patients with acute coronary syndrome, for example.
Additionally, the cyclotide scaffold can be engineered or evolved through molecular evolution techniches to selectively extracellular receptors such as binding G protein-coupled receptors to promote or block cell signaling . These data demonstrate the versatility of the cyclotide scaffold and highlight the extraordinary pharmacological properties of MCoTI-cyclotides and related linear knottins, thus confirming the potential of Cys-knotted polypeptide scaffolds in peptide-based drug discovery .
The ability to create cyclic polypeptides in vivo opens up the intriguing possibility of generating large libraries of cyclic polypeptides. Thus, libraries of genetically-encoded cyclic polypeptides containing billions of members can be readily generated using standard molecular recombinant tools. This tremendous molecular diversity allows one to perform selection strategies mimicking the evolutionary processes found in nature.
There are several examples where in vivo generated libraries of cyclic peptides have been used for rapid selection of biologically active peptides. For example protein trans-splicing has been used by several groups for the generation of libraries of small cyclic peptides (≤ 8 residues) in bacterial and mammalian cells [90, 92, 93, 99-101]. It should be pointed out, however, that the use of protein trans-splicing requires particular amino acids at the intein-extein junctions for efficient trans-splicing . These requirements usually depend on the type of split-intein used, and seriously limit the diversity of the libraries that can be generated when using small cyclic peptide templates. Our group has recently demonstrated the expression in E. coli of libraries based on the cyclic peptide SFTI-1 (a backbone cyclized Bowman-Birk trypsin inhibitor) using an intein-mediated backbone cyclization approach (see above), which allows the biosynthesis of backbone-cyclized polypeptides without any sequence requirement limitation .
The use of small cyclic peptides, however, which technically can be considered a single closed loop (or 2 loops in the case of SFTI-1), could also limit the potency of the peptides selected, especially when targeting protein-protein interactions involving large binding surfaces. In these cases the use of peptide templates such as cyclotides with multiple variable loops could facilitate the selection of peptides with higher affinities.
The potential for generating cyclotide libraries was first explored by our group using the kalata B1 scaffold . In this work wild-type and several mutants of kalata B1 were biosynthesized using an intramolecular native chemical ligation facilitated by a modified protein splicing unit. In this work, six different linear versions of kalata B1 were generated and expressed in E. coli as fusions to a modified version of the yeast vacuolar membrane ATPase (VMA) intein. Results demonstrated in vitro folding and cyclization of kalata B1 to varying degrees depending on which of the six native cysteine residues was at the N-terminus after cleavage of the initiation methionine by endogenous MAP. Cleavage and efficient cyclization of the different linear precursors did not occur equally, suggesting the amino acid residues near the intein as well as the predisposition to adopt a native fold of the corresponding linear precursor may determine the efficiency of the cleavage/cyclization step . This information was used to express a small library based on the kalata B1 scaffold. This library was cyclized in vitro by incubation with a redox buffer containing reduced glutathione (GSH) as a thiol co-factor, thus mimicking the intracellular conditions, where GSH is the most abundant thiol co-factor. The use of GSH allows the cyclization and folding to happen in one step [69, 89]. Analysis of the cyclization/folding reaction by HPLC and mass spectrometry revealed that all the members of the kalata B1 based library were expressed and processed with similar yields to give the corresponding natively folded cyclotides .
More recently, we have also reported the biosynthesis of a genetically encoded library of MCoTI-I based cyclotides in E. coli cells . The cyclization/folding of the library was performed either in vitro, by incubation with a redox buffer containing glutathione, or by in vivo self-processing of the corresponding precursor proteins. The bacterial gyrase A intein from Mycobacterium xenopus was used in this work . This intein typically express at higher yields than the yeast VMA intein in E. coli expression systems . The peptide libraries were purified and screened for activity using trypsin-immobilized sepharose beads, and then analyzed by HPLC and mass spectrometry. Out of 27 mutations studied, only two mutations, G27P and I22G, negatively affected the folding of the resulting cyclotides. All of the remaining cyclotides were able to fold with similar yields. The K6A mutant, as expected, was not able to bind trypsin. This residue is key for binding to the specificity pocket of trypsin, and can only be replaced by positively charged residues . This mutant was found by NMR to adopt a native cyclotide structure, confirming that the lack of biological activity was due to the mutation and not to the ability to adopt a native fold. It is interesting to note that by modifying the nature of this residue, the specificity of the corresponding MCoTI-cyclotide can be changed to target other proteases [95, 96]. The rest of the MCoTI-based library members were able to bind trypsin, suggesting they were able to adopt a native cyclotide fold and retained biological activity. The affinity of each peptide was analyzed using a competitive trypsin-binding assay. The mutants had a wide range of affinity, some being greater than wild type MCoTI-I. The peptides with less affinity were mostly found in loop 1 and the C-terminal region of loop 6, both well conserved among other squash trypsin inhibitors. Overall, these data describe the structural requirements for correct formation of MCoTI-I and the residues that are key to modulating trypsin binding. To our knowledge, this is the first time that the biosynthesis of a genetically-encoded library of MCoTI-based cyclotides containing a complete suite of amino acid mutants is reported
The chemical synthesis of a complete suite of Ala mutants for kalata B1 has also been recently reported . In this work all the mutants were fully characterized structurally and functionally. The results indicated that only two of the mutations explored (W23A and P24A, both located in loop 5, see Fig. 1) prevented folding . The mutagenesis results obtained in our work with the cyclotide MCoTI-I show similar results highlighting the extreme robustness of the cyclotide scaffold to mutations. These studies show that cyclotides may provide an ideal scaffold for the biosynthesis of large combinatorial libraries inside living bacterial cells. These genetically-encoded libraries can then be screened in-cell for biological activity using high-throughput flow cytometry techniques for the rapid selection of novel biologically active cyclotides [69, 103, 104].
In summary, cyclotides are a novel family of structurally related globular microproteins with a unique head-to-tail cyclized backbone, which is stabilized by three disulfide bonds [27, 105]. The number and positions of cysteine residues are conserved throughout the family, forming what is called cyclic cystine-knot (CCK) motif  that acts as a highly stable and versatile scaffold on which 5 hyper-variable loops are arranged (Fig 1). This CCK framework gives the cyclotides exceptional resistance to thermal and chemical denaturation, and enzymatic degradation. This is particularly important for the development of peptide-based therapeutics with oral bioavailability. In fact, the use of cyclotide-containing plants in indigenous medicine first highlighted the fact that the peptides are resistant to boiling and are apparently orally bioavailable [14, 32, 33]. Some cyclotides have also been shown to cross the cell membrane  thus allowing to target intracellular protein interactions, such as that mediated by viral FLIPs to prevent cell-death in virally infected cells (unpublished results). Cyclotides are also medium-sized polypeptides and therefore can be readily synthesized by standard solid-phase peptide-synthesis using either Boc- or Fmoc-based  methodologies thus allowing the introduction of non-natural amino acids or other chemical modifications for lead optimization. They can also be encoded within standard cloning vectors and readily expressed in bacteria or animal cells [26, 70], thus making them ideal substrates for molecular evolution strategies to enable generation and selection of compounds with optimal binding and inhibitory characteristics using high throughput cell-based assays . All of these characteristics make cyclotides appear as very promising leads or frameworks for development of peptide-based therapeutics and diagnostics [27, 68, 105, 107]
This work was supported by funding from the School of Pharmacy at the University of Southern California and by National Institute of Health award GM090323 to J.A.C.