Review
10.1517/13543784.15.8.933 © 2006 Informa UK Ltd  ISSN 1354-3784
933
Oncologic
Cationic antimicrobial peptides as
novel cytotoxic agents for cancer treatment
Jamie S Mader & David W Hoskin †
†Dalhousie University, Departments of Pathology and Microbiology & Immunology, Faculty of
Medicine, Sir Charles Tupper Medical Building, 5850 College Street, Halifax, Nova Scotia, B3H 1X5, Canada
Cancer treatment by conventional ch emoth erapy is h indered by toxic side effects and th e frequent development of multi-drug resistance by cancer cells. Cationic antimicrobial peptides (CAPs) are a pr
omising new class of nat-ural-source drugs that may avoid the shortcomings of conventional chemo-th erapy because certain CAPs exh ibit selective cytotoxicity against a broad spectrum of human cancer cells, including neoplastic cells that have acquired a multi-drug-resistant phenotype. Tumour cell killing by CAPs is usually by a cell membrane-lytic effect, although some CAPs can trigger apoptosis in can-cer cells via mitochondrial membrane disruption. Furthermore, certain CAPs are potent inhibitors of blood vessel development (angiogenesis) that is asso-ciated with tumour progression. This article reviews the mechanisms by which CAPs exert anticancer activity and discusses th e potential application of selected CAPs as therapeutic agents for the treatment of human cancers.
Keywords: angiogenesis, apoptosis, cancer therapeutics, cationic antimicrobial peptides, membrane lysis
Expert Opin. Investig. Drugs (2006) 15(8):933-946
1. Introduction
Although chemotherapeutic drugs play a pivotal role in the treatment of many dif-ferent cancers, the use of these anticancer agents is not without pitfalls. Because chemotherapeutic agents typically targ
et rapidly dividing cells, both neoplastic and healthy proliferating cells are subject to the cytotoxic effects of chemotherapy. This results in undesirable side effects that range from nausea and vomiting to myelo-suppression and thrombocytopoenia [1-3]. In addition, dormant cancer cells or tumours with a slow rate of growth respond poorly to chemotherapeutic agents that are only active against proliferating cells [4]. F urthermore, cancer cells frequently develop resistance to many anticancer drugs, which greatly reduces their therapeutic usefulness [5,6]. Multi-drug resistance can result from a number of cellular changes,including overexpression of drug transporters such as P-glycoprotein and multi-drug resistance protein-1, induction of various drug-detoxifying enzymes, and defects in apoptotic pathways such as aberrant expression of the antiapoptotic protein Bcl-2and mutation or inactivation of p53. There is clearly an urgent need to develop new approaches to cancer therapy that have a higher degree of selectivity for neoplastic cells as well as avoiding the problem of chemoresistance. A promising development has been in the area of therapeutic agents that restrict or abrogate the blood supply to solid tumours [7]; however, emerging evidence that tumours can develop resist-ance to the current generation of angiogenesis inhibitors [8] underscores the need for additional novel approaches to cancer treatment.
1. Introduction
2. CAP interactions with cancer
cells
3. α-Helical CAPs with anticancer
activity
4. β-Sheet CAPs with anticancer
activity
5. Hybrid CAPs and synthetic lytic
peptides
6. Targeting and delivery of
therapeutic CAPs for cancer treatment 7. Conclusions 8. Expert opinion
E x p e r t  O p i n . I n v e s t i g . D r u g s  D o w n l o a d e d  f r o m  i n f o r m a h e a l t h c a r e .c o m  b y  Y a l e  U n i v e r s i t y  o n  09/20/11
F o r  p e r s o n a l  u s e  o n l y .
Cationic antimicrobial peptides as novel cytotoxic agents for cancer treatment
934
Expert Opin. Investig. Drugs  (2006) 15(8)
Cationic antimicrobial peptides (CAPs) are natural-source drugs that show exciting potential as a new class of anticancer agents. CAPs are expressed in many diverse species (e.g.,insects, fish, amphibians and mammals) and have an important function in host innate immunity to microbial pathogens [9-12].These antimicrobial peptides typically consist of < 40 amino acids and are able to kill a wide range of Gram-negative and -positive bacteria, as well as some fungi, enveloped viruses and protozoa. In addition, many CAPs are potent immuno-modulators, eliciting both pro- and anti-inflammatory activi-ties by the host immune system. Of importance is that certain CAPs (e.g., bovine lactoferricin [LfcinB], cecropins, defensins and magainin 2) also exhibit direct cytotoxic activity against many different types of human cancer cells [13-16]. The cyto-toxic effect of CAPs on microorganisms and neoplastic cells is largely believed to be a function of the cationic nature and sec-ondary structure of these peptides [17,18]. Most of the CAPs are either linear amphipathic peptid
es that assume an α-helical conformation in the presence of biological membranes or β-sheet peptides that are generally stabilised by disulfide bonds and may also contain minor α-helical segments. Linear CAPs with a high proline and/or glycine content that lack secondary structure, as well as CAPs with a characteristic loop structure imparted by a single disulfide, amide or isopeptide bond have also been described.
CAPs offer several important advantages over conventional chemotherapeutic drugs that are currently used in the treat-ment of human malignancies. Many CAPs that kill human cancer cells do not show significant cytotoxicity against untransformed proliferating cells at peptide concentrations that are able to kill cancer cells [13,19-21], suggesting that these CAPs may be administered in vivo  with minimal nonspecific toxicity. However, it is important to screen putative anti-cancer CAPs for haemolytic activity because high concentra-tions of certain CAPs (e.g., melittin and tachyplesin) are known to cause red blood cell lysis [22,23]. Dormant or slowly growing malignancies and cancer cells with a chemoresistant phenotype are predicted to be sensitive to CAPs that kill by direct membrane lysis as this cytolytic process does not depend on the proliferative status of the target cell and should bypass most multi-drug resistance mechanisms. Indeed, cer-tain CAPs (e.g., cecropins and bovine myeloid antimicrobial peptide of 28 residues [BMAP-28]) have already been shown to kill
cancer cells that are resistant to conventional antican-cer drugs [14,20]. Interestingly, although some CAPs have been found to promote blood vessel development [24,25], other CAPs have the ability to interfere with tumour-associated angiogenesis [26,27]. CAPs with antiangiogenic activity that are also directly cytotoxic for human cancer cells offer the excit-ing possibility of attacking solid tumours at two distinct, yet complementary, levels. Although the topic is beyond the scope of this review, it is important to note that certain CAPs may also indirectly interfere with tumour growth via their ability to modulate the host immune response. The interested reader is referred to several recent reviews that provide a
comprehensive overview of CAP interactions with the host immune system [12,28,29].
The large number of CAPs that exist in nature constitute a rich source of potentially novel anticancer agents. This review aims to focus on selected CAPs (Table 1) that exhibit well-doc-umented direct cytotoxic activity against human cancer cells,summarise the available knowledge regarding their mecha-nism of action and discuss their possible therapeutic utility in cancer treatment.
2. CAP interactions with cancer cells
Similar to bacteria, many cancer cells carry a net negative charge due to their elevated expression, r
elative to non-transformed cells, of anionic molecules such as phosphatidyl-serine and O -glycosylated mucins on the outer membrane leaflet [30-33]. This net negative charge allows electrostatic interactions to occur between CAPs and the surface of many cancer cells. In contrast, the high content of zwitterionic phosphatidylcholine in the outer membrane leaflet of healthy eukaryotic cells confers an overall neutral charge on these cells that results in a greatly reduced capacity for electrostatic interactions with CAPs. In addition, membrane fluidity is typically increased in cancer cells relative to their healthy counterparts [34,35], which may facilitate cancer cell membrane destabilisation by membrane-bound CAPs. Furthermore, can-cer cells tend to have more abundant microvilli on their cell surface in comparison with nontransformed cells [36,37]. In combination with the net negative charge carried by many neoplastic cells, this increase in overall surface area may facili-tate CAP-mediated cytotoxicity by allowing a greater number of CAP molecules to interact with the surface of the cancer cell. Collectively, these properties of neoplastic cells may account for the ability of certain CAPs to kill cancer cells without harming healthy eukaryotic cells.
Several models have been developed to explain the inter-actions between CAPs and cell membranes that result in cell death. These include the barrel-stave model, the carpet model, the toro
idal or two-state model, the detergent-like effect model and the in-plane-diffusion model [18,38]. In the barrel-stave model (Figure 1), monomers of α-helical CAPs (such as cecropin B and melittin) bind to the cell membrane and aggregate as α-helical bundles that insert into the hydrophobic core of the membrane bilayer, forming trans-ient transmembrane pores [39,40]. It is important to note that the barrel-stave model cannot account for cytolytic activity by CAPs that consist of < 23 amino acids because these pep-tides do not have sufficient length to span the cell mem-brane. In the carpet model (Figure 1), CAPs that have bound anionic membrane components become aligned in parallel to the cell surface in a carpet-like fashion, which (at some critical threshold) results in membrane destabilisation and disintegration due to curvature stress and internal osmotic pressure without the actual insertion of peptides into the hydrophobic core of the lipid bilayer. At low concentrations,
E x p e r t  O p i n . I n v e s t i g . D r u g s  D o w n l o a d e d  f r o m  i n f o r m a h e a l t h c a r e .c o m  b y  Y a l e  U n i v e r s i t y  o n  09/20/11
F o r  p e r s o n a l  u s e  o n l y .
Mader & Hoskin
Expert Opin. Investig. Drugs  (2006) 15(8)
935
magainin has been suggested to function according to the carpet model but magainin is believed to mediate cytolysis via the barrel stave model at higher concentrations [41]; how-ever, magainin has also been proposed to cause pore forma-tion by the toroidal (two-state) model, which incorporates elements of the barrel-stave and carpet models to account for the formation of transient toroidal pores as a conse-quence of peptide-mediated disruption of the curvature of the target cell membrane [42]. In the detergent-like effect model, membrane disruption is caused by the release of micellar structures (observed as membrane blebbing) from areas of the membrane with high peptide density; however,there is little evidence at present to suggest that CAPs medi-ate cytotoxicity via the detergent-like effect model. The in-plane-diffusion model accounts for cytolytic activity by low concentrations of amphipathic α-helical CAPs, as well as by CAPs that are too short to span the cell membrane. In this model, the in-plane insertion of CAPs disturbs the packing of the membrane bilayer, leading to membrane thinning and formation of transient pores in response to peptide diffusion within the membrane bilayer. Models such as these, which describe cytotoxic peptide–membrane inter-actions, may allow one to predict the effectiveness of individual CAPs as anticancer agents.It is believed that CAPs associate with bacteria via inter-actions with negatively charged mole
cules such as lipopolysaccharide present in the outer membrane of Gram-negative bacteria. Subsequent destabilisation of the outer bacterial membrane allows CAPs to gain access to the cytoplasmic membrane, which is subsequently disrupted by hydrophobic interactions between amphipathic CAPs and the membrane lipid bilayer [10,12,17]. As with bacteria, the mito-chondria of eukaryotic cells are negatively charged and have a highly negative transmembrane potential. The net negative charge on the surface of mitochondria results from the relative abundance of the anionic lipid cardiolipin in mitochondrial membranes [43]. Interestingly, prokaryotic cells and the mito-chondria of eukaryotic cells are believed to share a common ancestry [44]; although many α-helical CAPs with anticancer activity appear to cause cancer cells to die by necrosis as a result of CAP-mediated damage to cell membranes [39,40], it is likely that some CAPs gain access to the cytosolic compart-ment of cancer cells via cell membrane destabilisation or toroidal pore formation. Figure 2 illustrates the interactions of some naturally occurring CAPs with neoplastic eukaryotic cells. Once inside the cell, these CAPs may perturb the integ-rity of negatively charged mitochondrial membranes,resulting in the release of several different mitochondrial
Table 1. Amino acid sequences of cationic antimicrobial peptides with anticancer activity.
Peptide Amino acid sequence
Ref.
BMAP-28GGL R SLG RK IL R AW KK YGPIIVPII R I
[123]Cecropin A K W K LF KK IE K VGQNI R DGII K AGPAVAVVGQATQIA K [124]Cecropin B K W K VF KK IE K MG R NI R NGIV K AGPAIAVLGEA K AL [124]SB-37MP K W K VF KK IE K VG R NI R NGIV K AGPAIAVLGEA K ALG [125]Shiva
MP R W R LF RR ID R VG K QI K QGIL R AGPAIALVGDA R AVG [125]
Defensins
Human neutrophil peptide-1ACYC R IPACIAGE RR YGTCIYQG R LWAFCC [126]Human neutrophil peptide-2
CYC R IPACIAGE RR YGTCIYQG R LWAFCC [126]Human neutrophil peptide-3DCYC R IPACIAGE RR YGTCIYQG R LWAFCC [126]Human neutrophil peptide-4VCSC R LVFC RR TEL R VGNCLIGGVSFTYCCT R V [127]hBD-1DHYNCVSSGGQCLYSACPIFT K IQGTCY R G K A K CC K [128]hBD-2GIGDPVTCL K SGAICHPVFCP RR Y K QIGTCGLPGT K [129]
LfcinB F K C RR WQW R M KK LGAPSITCV RR AF [130]Magainin 2GIG K FLHSA KK FG K AFVGEIMNS [131]MSI-136GIG K FL KK A KK FA K AFV K MNN [64]MSI-238D-GIG K FL KK A KK FA K AFV K MNN [64]Mellitin GIGAVL K VLTTGLPALISWI KRKR QQ [132]P18 hybrid K W K LF KK IP K FLHLA KK F [92]Tachyplesin
K WCF R VCY R GICY RR C R
[86]
Bold text denotes amino acids that are positively charged at neutral pH.
A: Alanine; B: Asparagine or aspartic acid; C: Cysteine; D: Aspartic acid; E: Glutamic acid; F: Phenylalanine; G; Glycine; H: Histidine; I: Isoleucine; K: Lysine; L: Leucine; M: Methionine; N: Asparagine; P: Proline; Q: Glutamine; R: Arginine; S: Serine; T: Threonine; V: Valine; W: Tryptophan; Y: Tyrosine; Z: Glutamine or glutamic acid.
E x p e r t  O p i n . I n v e s t i g . D r u g s  D o w n l o a d e d  f r o m  i n f o r m a h e a l t h c a r e .c o m  b y  Y a l e  U n i v e r s i t y  o n  09/20/11
F o r  p e r s o n a l  u s e  o n l y .
Cationic antimicrobial peptides as novel cytotoxic agents for cancer treatment
936
Expert Opin. Investig. Drugs  (2006) 15(8)
proteins (including cytochrome c) that are potent stimulators of apoptosis [45]. Indeed, LfcinB and BMAP-28 have both been shown to trigger apoptosis in cancer cells via depolarisation of mitochondrial membranes [13,46,47].
3. α-Helical CAPs with anticancer activity
3.1 Cathelicidins
BMAP-28 is an α-helical member of the cathelicidin peptide family that also includes human LL-37/human CAP (hCAP)-18 and porcine PR-39 [11]. BMAP-28 exhibits cyto-toxic activity against a variety of human tumour cell lines,including U937 myelogenous leukaemia and multi-drug-resistant CEM-VLB lymphoblastic leukaemia cells [20]. BMAP-28 treatment of leukaemia cells results in membrane permeabilisation and the influx of Ca 2+ into the cytosol. Although necrosis due to early membrane damage is a major cause of BMAP-28-mediated cytotoxicity,
modulateinternucleosomal DNA fragmentation (which is indicative of apoptosis) is also observed following longer exposure of leukaemia cells to BMAP-28. Indeed, submicromolar concen-trations of BMAP-28 have been shown to destabilise mito-chondrial membranes in K562 chronic myelogenous leukaemia cells, leading to mitochondrial membrane depolari-sation and the release of cytochrome c [46]. In the cytosolic compartment, cytochrome c complexes with caspase-9 and apoptosis-inducing factor-1 to form the apoptosome, which promotes the activation of caspase-9 and downstream proteins involved in apoptosis [45]. However, BMAP-28 has some limi-tations as a possible anticancer agent because
BMAP-281.5–6 µM (concentrations that are cytotoxic for leukaemia cells) also causes membrane permeabilisation and apoptosis in activated human lymphocytes [20]. In contrast, at concentra-tions that are cytotoxic for transformed cells, BMAP-28 does not substantially affect the viability of resting lymphocytes.The selective cytotoxicity of BMAP-28 for proliferating cells suggests that dormant or slowly growing cancer cells may be refractory to the cytotoxic activity of this peptide.
Other cathelicidin family members also show potential as anticancer agents. A 27 amino acid peptide (hCAP18109-135)derived from the C-terminal domain of α-helical hCAP-18has recently been shown to trigger apoptotis via a caspase-independent pathway in SAS-H1 human oral squamous carcinoma cells, without affecting the viability of untransformed human gingival fibroblasts or keratinocytes [48]. Similar to BMAP-28 [46], hCAP18109-135treatment causes mitochondrial transmembrane potential to be lost in SAS-H1 cells, suggesting peptide entry into the can-cer cells following insertion into the cell membrane and a sub-sequent interaction with negatively charged mitochondrial membranes [48]. In addition, human hepatocellular carcinoma cells engineered to express porcine PR-39 (which is a pro-line-rich rather than a α-helical cathelicidin) exhibit altered actin structure and reduced invasive activity, suggesting gene therapy with PR-39 as a possible approach to inhibiting tumour cell metastasis [49]. A recent study [50] indicates that PR-39 gene transfection inhibits the proliferation of
ras-trans-formed cells by PR-39 binding to the PI3K p85α subunit and subsequent inhibition of PI3K activity. These findings are of particular interest because PR-39 is reported to enter eukary-otic cells following its interaction with the cell membrane [51].The 37 amino acid peptide (LL-37) that is released from hCAP-18 is also taken up into mammalian cells, where it localises to the nuclear compartment without apparent growth-inhibitory or cytotoxic effects [52]. The significance of LL-37 uptake and nuclear localisation in terms of possible anticancer activity is not yet known.
3.2 Cecropins
Cecropins are CAPs that are present in mammals and many insects, including the giant silk moth Hyalophora  cecropia  [53,54].These cationic peptides consist of 34 – 39 amino acids and can be divided into cecropin A and B peptide families. All of the
Figure 1. Schematic representation of the barrel-stave and c arpet models of pore formation or membrane destabilisation by CAPs. A. Positively charged peptides bind to cancer cells through electrostatic interactions with negatively charged membrane components and initially align in a planar fashion with the cell membrane. B. In the carpet model,increasing concentrations of peptide cover the cell surface in a carpet-like manner. On reaching a critical concentration,hydrophobic portio
ns of the CAP permeate the cell membrane,causing curvature stress that leads to membrane destabilisation.The peptides remain in contact with anionic lipid headgroups during this process. In an extension of the carpet model known as the toroidal (two-state) model, curvature stress caused by CAPs that are long enough to span the cell membrane results in the formation of transient toroidal pores. C. In the barrel-stave model,amphiphathic  α-helical CAPs that are of sufficient length to span the cell membrane insert into the hydrophobic core of the membrane bilayer and aggregate as  α-helical bundles that form transient transmembrane pores.
CAP: Cationic antimicrobial peptide.
E x p e r t  O p i n . I n v e s t i g . D r u g s  D o w n l o a d e d  f r o m  i n f o r m a h e a l t h c a r e .c o m  b y  Y a l e  U n i v e r s i t y  o n  09/20/11
F o r  p e r s o n a l  u s e  o n l y .
Mader & Hoskin
Expert Opin. Investig. Drugs  (2006) 15(8)
937
cecropins show substantial sequence homology and display a general secondary structure comprised of an α-helical, amphi-pathic N terminus that contains many basic residues and a hydroph
obic C terminus [53]. Both cecropin A and B family members show cytolytic activity against several different human cancer cell lines, including leukaemia and lymphoma cells, but do not lyse normal fibroblasts, lymphocytes or mam-malian erythrocytes at peptide concentrations that are lethal to transformed cells [14,55-57]. Importantly, breast and ovarian carcinoma cell lines with a multi-drug-resistant phenotype are sensitive to cecropin B [14]. Patch clamp analysis of cecropin B and its analogue cecropin B3 reveals that these CAPs form tran-sient channel-like transmembrane pores in malignant cells [39].However, studies conducted with bacteria indicate that cecro-pin B forms toroidal transmembrane pores due to the positive curvature strain imposed on the cytoplasmic membrane by interactions with the amphipathic and hydrophobic segments of the peptide [58]. Interestingly, the combination of cecropin A and the chemotherapeutic agents 5-fluorouracil or cytarabine shows a supra-additive cytotoxic effect against CCRF -SB lymphoblastic leukaemia cells [57], suggesting that combination therapy with other α-helical CAPs and conventional chemo-therapeutic drugs should be examined as a potentially useful approach to cancer treatment.
3.3 Magainin 2
Magainin 2 is a CAP isolated from the skin of the African clawed frog Xenopus  laevis  [59]. This α-helical peptide, which
consists of 23 amino acid residues and is particularly rich in arginine residues, displays selective cytotoxic activity against neoplastic cells, killing several different cancer cell lines at pep-tide concentrations that are < 5 – 10 times lower than those nec-essary to kill untransformed human cells [16]. Growth inhibition of a panel of human small cell lung cancer cell lines has also been achieved with two magainin analogues, magainin A and G, which have enhanced α-helical composition and reduced haemolytic activity in comparison with naturally occurring magainin 2 [60], which has been shown to permeabilise and cross the cell membrane of HeLa human cervical carcinoma cells in an energy- and receptor-independent fashion [61]. This mode of action is consistent with model membrane studies showing that magainin 2 permeabilises and crosses the cyto-plasmic membrane of bacteria  to gain access to the cytosolic compartment [62]. Moreover, an analysis of the interactions between magainins and isolated rat mitochondria suggests that magainins dissipate mitochondrial transmembrane potential by forming channels in mitochondrial membranes [63]. Perme-abilisation of cancer cell mitochondria by magainin 2 may trig-ger apoptosis as a consequence of cytochrome c release and the subsequent activation of caspase-9 [45]. In vivo  studies indicate that magainin 2 and its derivatives show promise as therapeutic agents; for example, administration of magainin 2 or its syn-thetic analogues MSI-136 and MSI-238, which are designed to be resistant to proteolytic degradation, decreases tumour cell invasion into soft tissue and increases the survival of mice bearing P388 murine leukaemia cells [64].
Figure 2. Interac tions of some naturally oc c urring CAPs with c anc er c ells. Although all of the CAPs initially interact with the cytoplasmic membrane of cancer cells where they initiate membrane lytic events, certain CAPs are also able to target intracellular structures and molecules such as mitochondria (BMAP-28, hCAP18109-135, LfcinB and magainin 2), the nucleus (LL-37) and PI3K (PR-39).
CAP: Cationic antimicrobial peptide.
E x p e r t  O p i n . I n v e s t i g . D r u g s  D o w n l o a d e d  f r o m  i n f o r m a h e a l t h c a r e .c o m  b y  Y a l e  U n i v e r s i t y  o n  09/20/11
F o r  p e r s o n a l  u s e  o n l y .

版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。