Abstract
Although cancer belongs to one of the leading causes of death around the world, fortunately studies have shown that tumor cells have various targets that are susceptible to attack. Interestingly, tumor cells are comprised of cellular membranes, which are altered in chemical composition relative to non-neoplastic cells, giving them an increased net negative charge. These altered membranes are ideal targets for antimicrobial peptides (AMPs) shown to have additional tumoricidal properties and, hence, named anticancer peptides (ACPs). Several hundred ACPs have been explored in vitro and in vivo on various types of cancer. Novel anticancer agents are supposed not to cause serious side effects and the formation of multidrug-resistant tumor cells. During the quest for potent ACPs, promising candidates were isolated from skin secretions of amphibians, such as the granular glands of the Chinese brown frog, Rana chensinensis. ACPs have to be selective to cancer cells and should not induce strong immune responses or be cleared from the body rapidly. Several modifications can improve ACPs either by optimizing the primary structure rationally or randomly or even by introducing other chemical modifications.
The genetic variability among tumor cells of malignant melanomas is considered to be the largest hurdle for therapy. If mono-therapies, using only one chemotherapeutic drug were conducted, tumors would be temporarily reduced; however, this kind of selective pressure could permit the development of more rapidly-growing aggressive tumor cells.
Taken into account that tumor heterogeneity is an important factor regarding progression of malignant melanoma, the establishment of therapies that exploit striking similarities among these heterogenic cells which represent specific vulnerable targets should be considered.
Obviously, tumor cells differ in their morphological appearance from non-neoplastic cells, which are dependent on their loss of original function and their different chemical composition of the cellular membrane. Cellular membranes of tumor cells express high levels of anionic molecules, such as phospholipid phosphatidylserine (PS), O-glycosylated mucins, sialylated gangliosides and heparin sulfate, conferring them an overall net negative charge, whereas normal mammalian cell membranes are typically zwitterionic in nature (1, 2). This characteristic is an ideal target for cationic anticancer peptides (ACPs) that have been shown to be able to selectively disrupt the membranes of tumor cells (1, 3, 4).
The discovery of Antimicrobial Peptides
The surface of all multi-cellular organisms is colonized by microorganisms. This includes even invertebrates, e.g. insects, solely equipped with a primitive cellular immune system and also plants without any cellular immune system. In 1987, M. Zasloff discovered the group of magainins (deduced from Hebrew: magen=protective shield) (5). Magainins are peptides constitutionally produced by mucous glands in the skin of frogs with a broad-spectrum of antimicrobial activity against Gram-positive and Gram-negative bacteria and also against fungi and protozoa. The discovery of magainins gave a first hint for the existence of skin-specific chemical protective shields of amphibians mainly based on antimicrobial active peptides. The discovery of magainins ushered in the era of antimicrobial peptides (AMPs). Apart from amphibians, AMPs were detected in a variety of species including bacteria, fungi, plants and animals to which humans also belong. The antimicrobial peptide database (APD; http://aps.unmc.edu/AP/main.php) so far contains over 2,400 entries for antimicrobial peptides (APD was accessed in July 15, 2014) (6, 7). Host-defense peptides belonging to the innate immune system of various species have been studied mostly to target Gram-positive and Gram-negative bacteria but also fungi, protozoa and enveloped viruses, such as the human immunodeficiency virus (HIV) and herpes virus (4). The antiviral peptide lactoferricin was shown to inhibit the binding and uptake of the human papilloma virus (HPV), the human cytomegalovirus (HCMV) and the herpes simplex virus (HSV) into human cells (8-10). The effect of AMPs relies mainly on an interaction with molecules on the surface of the microbes, which are part of the cellular membranes. As bacteria expose negatively charged molecules on their membranes, they can be targeted by AMPs. These specific molecules are lipids, such as phosphatidylglycerol (PG), cardiolipin (CL) or phosphatidylserine (PS) (1). The majority of AMPs is cationic in nature and can interact by electrostatic forces with the latter negatively charged lipids (1). In contrast to Gram-positive bacteria, Gram-negative bacteria are additionally equipped with an outer membrane, which mainly contains anionic lipopolysaccharides (LPS). These are stabilized by divalent cations, such as Ca2+ and Mg2+ and can be easily displaced by cationic AMPs, which in turn allows access to the outer membrane (1).
A major problem of common antibiotics is the enormous emergence of resistant bacterial cells, such as those with extended-spectrum beta-lactamases (ESBLs) (11). In the future, AMPs might be a powerful alternative to common antibiotics because their interaction with membrane lipids shows a rapid non-specific mechanism that decreases the probability for the formation of resistant bacterial cells (4).
Therapies using AMPs are currently tested in pre-clinical, as well as in clinical, trials to promote wound healing and for the treatment of cystic fibrosis, catheter site infections, acne and patients who are undergoing stem cell transplantation (12-14).
On the other hand, many AMPs from various sources, with an additional activity on cancer cells, have been identified (1, 4). Their therapeutic advantage is supposed to be based on their selectivity towards malignant cells, while normal mammalian cells are left untouched.
Anticancer Peptides (ACPs)
More than one hundred AMPs have also been assayed with regard to their tumoricidal properties (1, 4). The antimicrobial peptide database so far lists 166 entries for anticancer/antitumor peptides (APD was accessed in July 15, 2014) (6, 7).
It is believed that AMPs are able to interact with membranes of neoplastic cells because of their altered chemical composition relative to normal mammalian cells. On the one hand, normal mammalian cells consist largely of zwitterionic phospholipids, such as phosphatidylethanolamine (PE), phosphatidylcholine (PC) or sphingomyelin (SM) conferring a neutral net charge and making them less attractive to cationic AMPs (1). On the other hand, neoplastic cells have been shown to have a net negative charge due to higher contents of anionic molecules, such as PS, O-glycosylated mucins, sialylated gangliosides and heparin sulfate (1, 2). These obvious membrane differences contribute largely to the selectivity of AMPs where electrostatic interactions between cationic AMP and anionic membrane components account for the major mechanism of cell death (1). Hence, AMPs capable of disrupting cancer cells were named anticancer peptides (ACPs).
Membrane changes are in fact important for a cell to become cancerous as it influences the ability to grow independently of growth stimulatory signals and to adhere and interact with neighboring cells (1). The cell membrane of cancer cells plays also a role for motility and support of tissue invasion and metastasis (1).
Selectivity and mode of action of AMPs and ACPs seem to be similar in theory because a net negative charge is found on bacterial, as well as on tumor, cell surfaces. Two major groups of ACPs have been proposed: those that are effective against bacterial cells and tumor cells lacking toxicity to normal mammalian cells and, secondly, those being effective against all three types of cells (1, 4).
ACPs' Mode of Action
In general, peptides first bind parallel to the membrane at low concentrations and, while raising the concentration, orientate more and more perpendicular to the membrane (15). Furthermore, at high peptide/lipid ratios, insertion into the bilayer occurs and finally trans-membrane pores are formed (15).
For the interactions between ACPs and cell membranes of cancer cells, several models have been established (1, 3, 15):
The barrel-stave model describes the mechanism of α-helical peptides comprising a minimum of ~20 amino acid residues and β-sheet peptides with a minimum of ~8 amino acid residues as this model requires the spanning of the cellular membrane (1). Once monomeric α-helical peptides have bound, they form bundles within the lipid bilayer with a central lumen similar to a barrel with α-helices being the staves resulting in transient trans-membrane pore formation. Hydrophobic regions of the peptide align with the hydrophobic core of the lipid bilayer, whereas the hydrophilic proportion generates the interior of the pore.
The carpet model is described by the parallel annealing of peptides to the cellular surface in a carpet-like manner. If a certain threshold is exceeded, the membrane gets destabilized because of bending and internal osmotic pressure.
The toroidal or two-state model involves the interaction between a cationic peptide with the polar head groups of membrane lipids. Continuous bending of the lipids leads to a connection between upper and lower leaflets of the membrane and, as a result, transient toroidal pores are formed.
The detergent-like effect model or inverted micelle model explains membrane disruption as a cause of membrane blebbing at areas with high peptide density resulting in the release of micellar structures.
The in-plane-diffusion model is based on peptide diffusion within the membrane subsequently triggering membrane thinning and pore formation. This model is applied to low concentrations of amphipathic α-helical ACPs and short ones that cannot sufficiently span the membrane.
Structure and Activity of ACPs
The amino acid sequence among ACPs is greatly differing and, in turn, has an effect on the secondary structure. Nonetheless, particular similarities have been reported as they are generally cationic in nature (net charge at neutral pH varies between +2 and +9) and amphipathic (1). These properties are important for the interaction with lipid membranes. The length of ACPs is often short and lies between 5 and 40 amino acid residues; however, longer ones have also been reported (1). The overall positive net charge is generally conferred by lysine (Lys) and arginine (Arg) residues, while the hydrophobic content is about 30 % or even more (1).
As there are various modes of action described for ACPs, one should also consider the very different secondary structures of ACPs. They have been described to mainly adopt α-helical structures but β-sheet and linear ACPs without a characteristic structure also exist (1, 4).
Skin secretions of amphibians, such as the African clawed frog Xenopus laevis, have been reported to contain bioactive peptides with α-helical structures and amphipathic faces named magainins (5). Magainin 2 was selective to solid tumor cell lines at concentrations that are 5- to 10-fold lower than concentrations that affected normal human peripheral blood lymphocytes or neutrophils (16). Magainin 2 or synthetic analogues were shown to form ion-conducting channels in cancer cell membranes (17) and there is evidence that lower concentrations destabilize membranes by the carpet model (18). Furthermore, magainins that have gained access to the cytosol may trigger apoptosis via the mitochondrial pathway supported by the lyses of isolated rat liver mitochondria (19). Magainin 1 applied on HL-60 human promyelocytic leukemia cells at concentrations between 10 and 50 nM was shown to induce apoptosis by mechanisms dependent on cytochrome c release and increased proteasomal activity (20). Whether membrane disruption or induction of apoptosis is the major force responsible for tumor cell lysis by magainins is not clear (1).
Well-described β-sheet ACPs were found to be a component of the human innate immune system as they are produced in the granules of neutrophils (21, 22). Cytotoxic effects of these defensins have been studied on several human and mouse tumor cells (1). Defensins are rich in cysteine (at least 6 residues) and arginine and their characteristic structural feature is based on three intra-molecular disulfide bonds between the N-terminal domain and the C-terminal domain of the molecule. In this manner hydrophobic and hydrophilic regions are separated to form an amphiphilic molecule. The lysis mechanism of defensins has been studied on human neutrophil peptides 1, 2 and 3 (HNP-1, -2, -3). Membrane binding is mediated by electrostatic interaction, which results in the collapse of membrane potential and finally the loss in membrane integrity (1). Importantly, HNP-1 and -3 were shown to interfere with neovascularization during tumor development by inhibiting α5β1 integrin-mediated migration and adhesion of endothelial cells to fibronectin via the vascular endothelial growth factor (VEGF) (23). Clinical administration of human defensins to cancer patients may not be possible due to their non-selective mode of action on normal leucocytes, epithelial cells and fibroblasts (1).
Another β-sheet ACP is bovine lactoferricin (LfcinB) isolated from cow‘s milk consisting of 25 amino acid residues where two of them are cysteine residues (1). These residues form disulfide bonds linking the positively charged N-terminal domain and the C-terminal domain. LfcinB adopts a twisted β-sheet conformation in aqueous solutions with amphipathic faces. In vitro studies demonstrated the cytotoxic activity of LfcinB on leukemia cells, fibrosarcoma cells, several carcinoma cells and neuroblastoma cells, whereas normal healthy cells were not substantially harmed using the same concentration range (1). Hints for necrotic and apoptotic killing mechanisms have been found depending on the cell line to be assayed. On the one hand, Eliassen et al. found that neuroblastoma cells exposed to LfcinB underwent a loss of membrane integrity due to the formation of trans-membrane pores followed by LfcinB internalization, which associated with negatively charged mitochondria (24). On the other hand, Mader et al. targeted human leukemia and breast carcinoma cells by LfcinB and could show a mechanism that included sequential generation of reactive oxygen species (ROS), loss of mitochondrial trans-membrane potential and finally the activation of caspases triggering apoptosis (25). A clear disadvantage of LfcinB is its reduced cytotoxic activity at high concentrations of serum but, nonetheless, systemic or intratumoral administrations were shown to inhibit growth and/or metastasis in various tumor mice models (1).
Additionally, some ACPs, isolated from porcine small intestine and porcine neutrophils (26, 27), are known to lack a specific secondary structure, such as PR-39, and show a rather unique mode of action (28). Inside the cytosolic compartment, these ACPs interact with src-homology 3-containing proteins which are involved in key cellular signaling processes. As a consequence, the PR-39 gene is up-regulated and reduces the invasive activity of human myeloma and hepatocellular carcinoma cells (28). This intracellular mechanism is believed to be responsible for PR-39 anticancer properties.
Non-membranolytic Mechanisms of ACPs
Further differences between normal and neoplastic cells are an increase in fluidity and the cell-surface area. The greater fluidity facilitates membrane destabilization by bound ACPs. Cholesterol, which is more abundantly found in the membranes of normal cells (29), influences membrane fluidity and is supposed to be able to protect normal membranes from the insertion of cationic peptides (30). Indeed it was shown for cecropin and its analogues that the rate of membrane insertion is negatively correlated with cholesterol content (31, 32).
The greater cell-surface area is due to microvilli that are more abundant on cancer cells and, thus, more ACPs can access the surface of cancer cells (3). In this way, the cell-surface area is supposed to influence receptor accessibility, cell adhesion and cell-cell communication, which may contribute to selectivity of ACPs (1).
Apart from membranolytic mechanisms, others, based on non-membranolytic mechanisms, have been described, such as induction of apoptosis activation of extrinsic apoptotic pathways and inhibition of angiogenesis (33).
Many ACPs have been shown to induce apoptosis in cancer cells via mitochondrial membrane disruption and subsequent ACP uptake into the cytoplasm (1, 3, 4).
A non-membranolytic mode of action of ACPs was the inhibition of angiogenesis within a solid tumor as it was shown for pentastatin-1, chemokinostatin-1 and properdistatin (34). Synthesized peptides were daily injected intraperitoneally into severe-combined-immuno-deficient (SCID) mice carrying xenografts of MDA-MB-231 human breast cancer cells (34). With this approach, it was possible to significantly suppress tumor growth and microvascular density (34). Angiogenesis inhibition is conferred by blocking receptors on angiogenic endothelial cells (4).
Apart from anti-angiogenic effects, additional alternative pathways have been shown to be affected by ACPs, such as mediated immunity (35), hormonal receptors (36) and DNA synthesis inhibition (37, 38).
Although the development and design of selective ACPs is still challenging and the antitumor activity of ACPs is not yet predictable and clearly deducible from primary sequences, ACPs might in the future be able to circumvent the low specificity of common chemotherapy and radiotherapy. It is also conceivable that ACPs pave the way for common chemotherapeutics in combinatorial therapies making a tumor more accessible.
With this gathered knowledge regarding the relationship between peptide structure and mode of action, the targeted design of ACPs might be a possible break-through in the future where various ACPs might use different targets on the tumor entity.
Selectivity of ACPs to Malignant Cells
The selectivity of temporin-1CEa to human breast cancer cell lines relative to non-tumorigenic cells, such as normal human umbilical vein smooth muscle cells (HUVSMCs) and human erythrocytes, has been demonstrated several times (39-43).
In general, fibroblast cultures were utilized to determine cancer cell selectivity of ACPs with regard to cecropin A and B (45) and to determine if the C-terminal amidation of peptides has an influence on selectivity (46). In the case of cecropins, cytotoxicity against bladder cancer cells was compared to two types of benign fibroblasts; mouse fibroblasts of the 3T6 cell line and primary fibroblasts from human gingival tissue isolated using a standard protocol (45). Suttmann et al. also made use of the WST-1 and lactate dehydrogenase (LDH) release assays and found that murine and human fibroblasts were less or not susceptible to peptide treatments (45). In the case of C-terminally amidated peptides versus non-amidated ones, Dennison et al. found no significant differences in selectivity when WI38 human diploid fibroblasts from fetal lung origin were used as normal mammalian cells (46).
Members of the temporin family, in general, showed high hemolytic activity (47). Therefore, the peptide temporin-1CEa (39-42) and analogous peptides (43) were steadily assayed regarding hemolysis. Temporin-1CEa was always termed as moderately hemolytic considering its use for therapeutical application (39-43). Results indicated that human erythrocytes should be definitely included in selectivity assays.
The hemolytic activity of Lys- and Arg-containing synthetic ACPs was compared and it suggested that the guanidinium group of Arg interacts more electrostatically with zwitterionic phospholipids than the primary amine of Lys (48). Introducing Lys residues rather than Arg residues, when constructing cancer cell selective peptides, would be an alternative. This could direct peptide binding towards the negatively-charged tumor cells and avoid hemolytic events (4).
The Temporin Family of Antimicrobial Peptides
Temporins are AMPs, which are synthesized in granular glands of frogs, all belonging to the genus Rana. Upon stress or injury they are released into skin secretions and protect the frog from a variety of pathogenic microorganisms.
The family of temporins is characterized by a short length of 8-17 amino acid residues with a high proportion of hydrophobic residues and an α-amidated C-terminus (47). Hence, they are among the smallest AMPs found in nature. The smallest AMP found to date is temporin-SHf which is a phenylalanine–rich, hydrophobic octapeptide being active against Gram-positive and Gram-negative bacteria, as well as yeasts, with absence of hemolytic activity (49).
Temporins have been isolated from various ranid frogs found in both Eurasia and North America, such as the European red frog R. temporaria (50), which is responsible for their name. Another source of temporins comes from the North American pickerel frog R. palustris (51). In the electrically-stimulated skin secretions of R. temporaria, 10 structurally related peptides were found and classified as temporins (50). Amphibian peptides evolve from precursors with conserved signal peptide and acidic pro-peptide regions, which, in the case of temporins, are named preprotemporins (52). The strength of conservation is even maintained among different frog families (52). In contrast to this, the adjacent antimicrobial peptide domain is hypervariable. Upon maturation, the signal peptide is cleaved at a conserved cysteine residue (Cys22) forming the pro-peptide, which is subsequently processed at a typical lysine-arginine pro-hormone cleavage site (52). The last step of maturation involves the C-terminal α-amidation by a peptidyl-glycine α-amidating monooxygenase using the penultimate glycine residue as a nitrogen donor (52).
Despite temporins being highly variable in their primary structure, a consensus sequence was proposed based on 119 naturally-occurring peptides by Wade et al.: F- L- P- I- L- G- S- L- L- S- G/K- L-L-NH2 (53).
Temporins often comprise at least one single basic residue, usually lysine (K), resulting in a charge of +2 at the physiological pH. Predominance for hydrophobic amino acid residues and α-amidated C-termini are also common. Additionally, temporins have a tendency to adopt an amphipathic α-helical conformation in a membrane-mimetic solvent, such as 50% trifluoroethanol (TFE)/water, whereas they exist in a random coil conformation in water (54). The all-D enantiomer of temporin A was equally effective against bacteria indicating that the peptide interacts with membrane lipids via non-chiral interactions (54).
The mechanism by which temporins destroy bacterial cells is most likely the “toroid pore” mechanism (formation of transmembrane pores) rather than causing a detergent-like disruption of the cell membrane (“carpet” mechanism) (55).
It is also believed that membranolytic activities are more dependent on hydrophobic interactions than on ionic interactions with bacterial membranes; however, the positively charged amino acid residue was shown to be essential for antibacterial activity except in the case of temporin-1Od which lacks a positively charged residue and was effective against Staphylococcus aureus at a minimum inhibitory concentration (MIC) of 13 μM (56).
Targets of temporins are mainly Gram-positive bacteria. For instance, temporin A has been shown to be active against clinically important methicillin-sensitive and methicillin-resistant Staphylococcus aureus, as well as vancomycin-resistant Enterococcus faecium strains (54, 57).
As an exception, temporin L was active against Gram-negative strains, such as Pseudomonas aeruginosa (ATCC 15692) and Escherichia coli D21 and also cytotoxic to human erythrocytes and various human tumor cell lines (58). Temporins have also shown anti-viral (59), anti-fungal (60) and anti-parasitic (61) activities.
Another feature of temporins was chemotaxis validated in vitro, as well as in vivo for temporin A (62). Chen et al. were able to induce migration of human monocytes, neutrophils and macrophages in vitro and infiltration of neutrophils and monocytes upon injections of temporin A into mice (62). The responsible receptor for attracting phagocytes was identified as the formyl peptide receptor-like 1 (FPRL1) (62).
Regarding all aforementioned facts, temporins are attractive candidates for novel anti-infective agents, especially for the treatment of infections with antibiotic-resistant bacterial strains. Their combination with a conventional antibiotic is also considered, as temporins facilitate the entry of conventional agents.
The hemolytic activity of temporins is supposed to hinder their therapeutic application; however, less hemolytic temporins or temporin analogues have been found and synthesized. As an example, a less hemolytic AMP with antitumor properties, known as temporin-1CEa, has been found to be moderately hemolytic (39-43).
Temporin-1CEa, a Relatively Small Selective Anticancer Peptide, Derives from Skin Secretions of the Chinese Brown Frog
Together with other AMPs, temporin-1CEa was isolated from frog secretions of R. chensinensis, the Chinese brown frog. Apart from its antibacterial activities, temporin-1CEa was shown to be efficiently active against various human tumor cells in vitro with moderate cytotoxicity to normal cells (39-43).
Studies revealed that temporin-1CEa is an ACP composed of 17 amino acid residues shown to adopt a well-defined α-helical structure in 50% TFE/water with a mean hydrophobicity of 11.3 (52% hydrophobic residues) and a net charge of +3 because of three lysine residues (43, 6, 7). When temporin-1CEa was cloned the first time, the mature peptide was specified as FVDLKKIANIINSIF-NH2 (39, 40) and in later studies as FVDLKKIANIINSIFGK-NH2 with two additional residues on the amidated C-terminus (41-43). These differences in the primary sequence have at least a consequence on the net charge at neutral pH because the longer sequence ends with an additional positively charged Lys (K) residue.
The first study, which mentioned temporin-1CEa, dealt with the cloning of the cDNA of its precursor peptide, preprotemporin-1CEa (39). After derivation of the amino acid sequence, Shang et al. chemically synthesized and evaluated the antimicrobial, cytotoxic and hemolytic properties of temporin-1CEa. In comparison to three other synthesized peptides also deduced from prepropeptide sequences, temporin-1CEa was outstandingly toxic to the human breast tumor cell line MCF-7 at a half maximal inhibitory concentration (IC50) of 12 μM, while being less cytotoxic to erythrocytes at an IC50 of 230 μM (39). Its mode of action was supposed to be based on its high degree of α-helical conformation and amphipathicity (39).
A large-scale study, conducted to further characterize the antitumor effects of temporin-1CEa, made use of the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) cell viability and LDH release cytotoxicity assays. For the experiments, testing the effect of synthesized temporin-1CEa on cellular targets, twelve different human breast cancer cell lines and also human umbilical vein smooth muscle cells (HUVSMCs) were selected. All tested tumor cells were sensitive to temporin-1CEa in a dose-dependent manner, where MCF-7 cells were the most sensitive at an IC50 of 31.1 μM after a 24-hour treatment (40). The hemolytic effects were moderate in extent (half-lethal dosage (LD50) at 99.08 μM) and cytotoxicity could not be detected on normal HUVSMCs below 100 μM. In addition, the peptide was clearly grouped into the temporin family regarding its primary structure (FVDLKKIANIINSIF-NH2) with a homology from 50 to 56 % (40). Wang et al. made use of circular dichroism (CD) spectroscopy to determine the α-helical structure of temporin-1CEa dissolved in both sodium phosphate buffer and 50 % TFE/water. The CD spectra indicated that temporin-1CEa has an unordered structure in water, whereas in 50 % TFE adopts an α-helical structure. In this manner, temporin-1CEa was thought to interact with anionic membranes (40).
In order to investigate the mechanisms of cell death conferred by temporin-1CEa in detail, two different studies were undertaken making use of the same strategy but using different human breast cancer cell lines as experimental cellular targets; MDA-MB-231,MCF-7 (41) and Bcap-37 (42). Temporin-1CEa induced cell death rapidly after 1 hour in a concentration-dependent manner on the three tested human breast cancer cell lines. The visible results were striking morphological changes evaluated by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) at higher temporin-1CEa concentrations just after 1 h exposure (41, 42). FITC-labeled-temporin-1CEa was traced by laser scanning confocal microscopy and uptake of the latter was induced at higher concentrations, whereas at a lower concentration (20 μM) the labeled peptide was excluded from the inside. It was also shown that FITC-labeled-temporin-1CEa could not internalize at a concentration of 20 μM. At the same concentration, other intracellular events were triggered, such as ROS production, calcium ion release into the cytosol, depolarization of the transmembrane potential and a loss in mitochondrial membrane potential (ΔΦm) (41, 42).
At higher concentrations (30-40 μM) of temporin-1CEa, a direct disruption of the cell membrane was more likely to occur. This was shown by concomitant staining of cancer cells with FITC-coupled annexin-V and propidium iodide (PI) and subsequent analysis by flow cytometry where signals from FITC-annexin V indicated increased PS exposure and PI signals indicating increased loss in membrane integrity relative to non-treated cancer cells (41, 42).
Trans-membrane potential measurements via a membrane potential sensitive dye indicated an immediate depolarization upon treatment with temporin-1CEa (41, 42).
The final disruption of the cell membrane was detected at higher temporin-1CEa concentrations via phosphatidylserine (PS) exposure, membrane permeabilization and leakage of intracellular content. AMP targets within a cell are believed to be situated on mitochondrial membranes which are very similar to bacterial membranes. Overall, these results strengthened the hypothesis that this was also the case for temporin-1CEa. In general, both mechanistic studies were able to link anticancer activity of temporin-1CEa to induction of significant membrane disruption as well as Ca2+-release and ROS over-production (41, 42).
Yang et al. conducted a structure-activity relationship (SAR) study in order to enhance the tumoricidal effects of temporin-1CEa-based analogues and minder the effect on red blood cells ,which was expressed via a better therapeutic index (ratio of LC50 for erythrocytes to the IC50 for tumor cells) (43). Apart from temporin-1CEa, the authors synthesized 6 analogues that had been rationally altered in a primary sequence. The idea was to maintain amphipathicity and an α-helical secondary structure while changing the positive charge and hydrophobicity. With this strategy, enhancement of anticancer activity and therapeutic index was shown to be possible. Their results also indicated that increased hydrophobicity was positively linked to hemolytic and antitumor activities. In general, Yang et al. observed that maintaining the naturally occurring hydrophobicity of α-helical AMPs, while changing its cationicity, is a useful procedure to optimize cytotoxicity to tumor cells and at the same time minder the effects on healthy cells (43).
Conclusion
Considering that ACPs harbor unique modes of action when attacking malignant cells, these proteins could be definitely considered as a solution to fight cancer in the future since they have the power to prevent the formation of multidrug-resistant tumor cells.
Understanding how ACPs really work may help to facilitate their design in the future.
The most important feature of ACPs is their selectivity towards tumorigenic cells, which is usually the result of their altered membrane composition relative to normal healthy cells. The amphibian-derived peptide temporin-1CEa was cytotoxic to breast cancer cells at millimolar concentrations while being moderately toxic to red blood cells. Determining the selectivity of ACPs is definitely mandatory and realizable via hemolysis assays.
The development of an ACP as a potent drug against tumors needs further support from in vivo validations. With a validated ACP, it is conceivable that it could pave the way for common chemotherapeutics in combinatorial therapies making a tumor more accessible. The therapeutic potential of ACPs should not be neglected and, therefore, extended research should be undertaken against cancer‘s high morbidity and mortality rates.
Footnotes
-
↵* These Authors contributed equally to this work.
- Received October 15, 2014.
- Revision received November 7, 2014.
- Accepted November 11, 2014.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved