Background
A subgroup of cationic antimicrobial peptides (CAPs) constitutes a promising group of novel anticancer agents with a new and unique mode of action and a broad spectrum of anticancer activity. CAPs induce cell death by increasing the membrane permeability of the target cells and are therefore unaffected by multidrug resistance mechanisms seen with conventional chemotherapeutic drugs [
1‐
5]. Moreover, several CAPs display a higher specificity for cancer cells versus normal cells in comparison to conventional chemotherapy [
6,
7]. Their potential as anticancer agents has been further established by
in vivo studies, as these peptides have been shown to induce regression of primary tumors [
8,
9] and prevent metastases [
10‐
13]. Recently we reported that intratumoral injection of a short CAP, LTX-302, derived from the naturally occurring CAP bovine lactoferricin (LfcinB), leads to a local inflammation followed by a complete regression of the tumor. Interestingly, local treatment with LTX-302 also elicited immunization against the tumor, resulting in protection against recurrence and metastasis [
14]. LTX-302 displayed a selective disruptive effect on the tumor plasma membrane, leading to necrosis of the tumor cells. However, it is not known what kind of cell surface molecules determines the specificity of this peptide.
LTX-302 consists of an idealized amphiphatic α-helical structure, which facilitates interactions with anionic surfaces. The cell surface of many cancer cells has an increased net negative charge due to an elevated expression of anionic molecules, such as phosphatidylserine in the outer membrane leaflet [
15‐
18], and terminal sialic acids on the cell surface, such as N-linked glycans and O-linked glycans [
19,
20], compared to non-malignant cells.
Several types of cancer cells such as carcinoma cells [
21‐
23], melanoma cells [
24], lymphoma and leukemia cells (Uhlin-Hansen, L. Manuscript in preparation) have different patterns of cell surface proteoglycan expression compared to their normal counterparts. The negatively charged glycosaminoglycans (GAGs) attached to the core protein of cell surface proteoglycans consist of repeating disaccharides and are highly sulfated [
25,
26]. Two major classes of GAGs are heparan sulfate (HS) and chondriotin sulfate (CS). The GAGs are part of the anionic glycoconjugate cell coat that surrounds the cells, and are therefore potential interaction partners for CAPs. The two main families of membrane bound proteoglycans, syndecans and glypicans, have HS chains attached to their core proteins, although CS can also be present on the syndecans [
27,
28].
We have previously shown that the cytotoxic activity of the two peptides, LfcinB and KW5, was inhibited by the presence of HS on the cell surface [
29]. An interaction with different GAG molecules has also been reported for the naturally occurring CAPs α-defensin, LL-37, magainin and melittin [
30‐
32]. The structural diversity in these CAPs and their different net positive charge, ranging from +3 in human α-defensin to +9 for the KW5 peptide, indicate that various structural properties can be involved in binding to GAGs.
The LTX-302 peptide is part of a new generation of small lytic peptides consisting of only 9 amino acids. This new generation of CAPs is based on structure-activity studies performed on LfcinB, in which we have identified structural parameters important for its antitumor activity. By optimizing these critical structural parameters we have designed peptides with a higher antitumor activity than the naturally occurring CAPs [
33‐
36]. The observation that the use of large, bulky, non-coded amino acids enhanced antitumor activity, and could also compensate for a decreased number of aromatic acids [
34,
36], enabled us to design much shorter CAPs than previously reported. The size of the peptides may be an important factor in developing them peptides into potential anticancer drugs, since smaller chemically modified peptides are expected to have increased bioavailability and stability, as well as a reduced immunogenicity. Another hypothesis is that smaller CAPs might slip more easily through the cell coat to the phospholipid bilayer, resulting in an increased cytotoxic effect for the peptide.
In this study the role of GAGs in the cytotoxic activity of LTX-302 and two other 9-mer peptides, LTX-315 and LTX-318, was studied. The three peptides with a net positive charge of +6 were amidated in their carboxy terminal, and included a non-coded aromatic acid, but differed in their primary structure and cytotoxic activity against cancer cells and normal cells. In contrast to our previous study [
29], this study revealed that the cytotoxic activity of these smaller CAPs is either enhanced or not affected by GAGs expressed on the cell surface.
Methods
Reagents
All Fmoc-amino acids, Fmoc-resins and chemicals used during peptide synthesis, cleavage and precipitation were purchased from PerSeptive (Hertford, UK), Fluka (Buchs, Switzerland) and Sigma-Aldrich (St. Louis, MO). Fetal bovine serum (FBS) was obtained from Biochrom KG (Berlin, Germany), and L-glutamine from Gibco (Paisley, Scotland). MTT (3-(4, 5-dimethylthiazol-2-yl)-2.5-diphenyl tetrazolium bromide) was obtained from Sigma-Aldrich (Oslo, Norway). Chondroitinase ABC (EC 4.2.2.4) was purchased from Seikagaku Corporation (Chuo-ku, Tokyo, Japan). Chondroitin sulfate (C-4384) and heparan sulfate (H-7640) were obtained from Sigma-Aldrich (Oslo, Norway). [35S]Sulfate (code SJS-1) was purchased from Amersham Biosciences (Buckinghamshire, UK).
The lymphoma cell lines KMS-5, KMM-1 and Sudhl-4 were a kind gift from Mark Raffeld, Hematophathology Section, Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, MD. Jeffery D. Esko, Department of Cellular and Molecular Medicine, University of California, San Diego, USA, kindly provided us with the mutant Chinese hamster ovary cell line pgsA-745, which does not express GAGs at the cell surface, as well as the wild-type CHO-K1 that expresses normal amounts of GAGs [
37,
38]. The lymphoma cell lines U-266, Ramos, the colon carcinoma cell line HT-29, the breast carcinoma cell line MT-1 and the neuroblastoma cell line Kelly were obtained from the American Type Culture Collection. Human umbilical vein endothelial cells (HUVEC) were obtained from MedProbe, Lonza.
Peptide synthesis, purification and analysis
The peptides LTX-302, LTX-315 and LTX-318 were synthesized by solid-phase methods using standard Fmoc chemistry on a Pioneer Peptide synthesizer (Applied Biosystems, Foster City, CA). Crude peptides were purified by preparative RP-HPLC (Waters, Milford, MA) using a C18 column (Delta-Pak™ C18, 100Å, 15 μm, 25-100 mm), and analysed on an analytical C18 HPLC column (Delta-Pak™ C18, 100Å, 5 μm, 3.9 × 150 mm) (Waters, Milford, MA). The purity of the peptides was found to be >95%. Peptide characterization was done by positive ion electrospray ionization mass spectrometry on a VG quattro quadrupole mass spectrometer (VG Instruments Inc., Altringham, UK).
Cell cultures
The HT-29, MT-1, Kelly, HUVEC and MRC-5 cells were maintained as monolayer cultures. The HT-29, MT-1 and Kelly cells were maintained in RPMI-1640 (R8758, Sigma-Aldrich, Oslo, Norway) supplemented with 10% (v/v) FBS. MRC-5 cells were maintained in MEM (M4655, Sigma-Aldrich, Oslo, Norway) supplemented with 10% (v/v) FBS. The HUVEC cells were maintained in Endothelial Cell Growth Medium-2 BulletKit obtained from MedProbe, Lonza. The CHO-K1 and pgsA-745 cell lines were maintained as monolayer cultures in HAM`s-F12 (E15-817, PAA Laboratories, Oslo, Norway) supplemented with 10% (v/v) FBS. All the lymphoma cell lines were grown in suspension in RPMI-1640 medium supplemented with 10% (v/v) FBS. All cells were grown in tissue culture flasks in a humidified atmosphere of 95% air and 5% CO2 at 37°C.
Cytotoxicity assay
The colorimetric MTT viability assay was used to investigate the cytotoxic effect of the peptides. The HT-29, MT-1, Kelly, HUVEC and MRC-5 cells were seeded at a concentration of 2 × 105 cells/ml, 1.5 × 105 cells/ml, 2 × 105 cells/ml, 1 × 105 cells/ml and 1 × 105 cells/ml in a volume of 0.1 ml in 96-well plates, respectively. CHO-K1 and pgsA-745 cells were seeded at a concentration of 1 × 105 cells/ml. The cells were allowed to adhere overnight in complete medium. Before adding different concentrations of the peptides (10-500 μg/ml) to the cells, the culture medium was removed and the cells were washed twice in serum-free culture medium. The non-adherent lymphoma cell lines were seeded at a density of 4 × 105 cells/ml using serum-free medium. After incubating the cells with peptides for 30 minutes at 37°C, 0.5 mg MTT-solution was added to each well and the incubation was continued for 2 hours. A volume of 70 μl or 130 μl per well was removed from the non-adherent and adherent cells, respectively. In order to dissolve the formazan crystals, 100 μl of 0.04 M HCl in isopropanol was added and the plates were shaken for 1 hour on a Thermolyne Roto Mix (Dubuque, IA) at room temperature. The optical density was measured on a microplate reader (VERSAmax™ Molecular Devices, CA). Cells treated with 1% Triton X-100 in serum-free medium was used as positive control for 100% cell death, whereas cells in serum free medium were used as negative control. Cell survival was determined from the ΔA590 nm relative to the negative control (100% living cells) and expressed as 50% inhibitory concentration (IC50).
Hemolytic assay
The hemolytic activity of the peptides was determined using freshly isolated human red blood cells (RBC) as previously described [
39]. Briefly, venous blood was collected, transferred to a tube containing heparin (10 U/ml) and centrifuged at 1500 rpm for 10 minutes in order to isolate the red blood cells. The red blood cells were washed three times with PBS (35 mM phosphate buffer with 150 mM NaCl, pH 7.4) by centrifugation at 1500 rpm for 10 minutes, and adjusted to 10% hematocrit with PBS. Peptide solutions were added to yield a final concentration range of the peptide from 1000 μg/ml to 1 μg/ml and a red blood cell concentration of 1%. The resulting suspension was incubated with agitation for 1 hour at 37°C. After incubation the suspension was centrifuged at 4000 rpm for 5 minutes, and the released hemoglobin were monitored by measuring the absorbance of the supernatant at 405 nm on a microplate reader (VERSAmax™ Molecular Devices, CA). PBS and 1% Triton X-100 were used as negative and positive controls, respectively. Peptide concentrations corresponding to 50% hemolysis (EC
50) were determined from dose-response curves.
Radiolabeling and isolation of 35S-labeled macromolecules
CHO-K1 cells and the lymphoma cells were radiolabeled for 20 hours by adding [
35S]sulfate to a final concentration of 50 μCi/ml at the time of cell plating. After this incubation time, the lymphoma cells were harvested by centrifugation and washed twice with serum-free medium before a 4 M guanidinium chloride solution containing 2% triton X-100 was added to the cells. The plasma membrane-associated
35S-labeled macromolecules on the CHO-K1 cells were harvested by washing the cells twice with serum free-medium and then incubating them for 15 minutes at 37°C in the presence of 10 μg/ml of trypsin [
40]. Free [
35S]sulfate was removed by gel filtration on Sephadex G50 Fine columns (bed volume 4 ml, equilibrated with 0.5 M Tris/HCl, pH 8.0 and 0.15 M NaCl and eluted with distilled H
2O). Aliquots from the membrane fractions were analysed for radioactivity in a scintillation counter after the addition of Ultima Gold XR scintillation fluid. The rest of the material was immediately frozen and stored until further analysis.
Alkali treatment and gel chromatography
The 35S-labeled macromolecules were subjected to alkali treatment (0.5 M NaOH overnight at 45°C, followed by neutralization with 0.5 M HCl), resulting in liberation of free 35S-labeled GAG chains. The 35S-labeled macromolecules were subjected to Superose 6 gel chromatography both before and after alkali treatment. Markers for void (Vo) and total volume (Vt) were blue dextran and [35S]sulfate, respectively. The columns were run in 4 M guanidine-HCl with 0.05 M sodium acetate, pH 5.8. Fractions were collected and the radioactivity counted in a scintillation counter.
Selective PG degradation
The
35S-labeled macromolecules were subjected to enzymatic treatment with chondroitinase ABC (cABC), which depolymerizes CS. Incubations with cABC were performed at 37°C overnight with 0.01 U enzyme per sample in 0.05 M Tris/HCl, 0.05 M sodium acetate, pH 8.0. The samples were analyzed on Sephadex G-50 Fine columns (bed volume 4 ml, equilibrated and eluted with the Tris/HCl buffer). Parallel samples were subjected to HNO
2 treatment at pH 1.5 in order to degrade the HS chains [
41]. The samples were analyzed by Sephadex G-50 Fine columns (bed volume 4 ml, equilibrated and eluted with dH
2O). Aliquots from the collected fractions were analyzed for radioactivity in a scintillation counter after the addition of Ultima Gold XR scintillation fluid.
Affinity assay
Two different Sepharose affinity columns were prepared, using HS and CS as ligands. The ligands were mixed with swollen CNBr-activated Sepharose 4B. Non-reactive groups were blocked with 0.1 M Tris-HCl, pH 8.0 and the gel was washed before packing. The peptides were dissolved in 5 mM phosphate-buffer, pH 7.4 at a concentration of 0.5 mg/ml, and 50 μl samples were applied. A gradient of NaCl was used to elute the different peptides from the columns using a GradiFrac from Amersham Pharmacia Biotech (Uppsala, Sweden) at a flow rate of 1.0 ml/min. The peptides were detected using a monitor UV-1 from Amershan Pharmacia Biotech (Uppsala, Sweden).
Discussion
Several carcinoma, melanoma, lymphoma and leukemia cell lines are more sensitive to CAPs compared to normal cells [
46‐
49]. It is believed that this selectivity is due to a more negatively charged cell surface of the cancer cells. However, we have previously shown that negatively charged GAGs on the cell surface inhibit the cytotoxic activity of CAPs, probably by sequestering the peptides away from the phospholipid bilayer. In the present study three peptides consisting of only 9 amino acids, and with a net positive charge of +6, were tested for their antitumor activity and selectivity. Compared to LfcinB, the 9-mers were more active and killed cancer cells more effectively, showing that the 9-mers are more optimized for antitumor activity than LfcinB. By examining the role of cell surface GAGs on the cytotoxic effect of the 9-mers, we found that cell surface GAGs had a different effect on the cytotoxic activity of this new generation of shorter peptides compared to what we previously reported for the longer naturally occurring LfcinB (25-mer) peptide and the KW5 (21-mer) peptide.
All the three 9-mer peptides displayed a higher activity towards the lymphoma, carcinoma and neuroblastoma cell lines compared to normal endothelial cells, fibroblasts and red blood cells. One exception was the lower activity of LTX-315 against the carcinoma cell lines compared to the endothelial cells.
LTX-315 killed the various tumor cells more efficiently than LTX-302 and LTX-318. However, the LTX-318 and LTX-302 peptides displayed a higher specificity for the tumor cells versus the non-tumor endothelial and fibroblast cells than did LTX-315. These findings are in agreement with our earlier findings that enhanced antitumor activity may result in reduced tumor cell specificity [
50,
51].
The relatively higher cytotoxic activity against the lymphoma and neuroblastoma cells compared to the endothelial cells, the fibroblast cells and the red blood cells suggest that differences at the cellular membrane level decide their vulnerability to the peptides. Differences in cell membrane composition, fluidity [
52] and surface area [
53,
54] between cancer cells and normal cells may be factors that make the former cells more susceptible to the peptides.
The lack of correlation between the cytotoxic activity of the peptides and the expression of HS on the cell surface of the lymphoma cells indicates that membrane components other than HS affect the susceptibility of the lymphoma cells against the 9-mers. The cell lines that displayed the highest sensitivity against the peptides also had the highest amount of cell-associated CS. It can therefore be speculated if CS is involved in the cytotoxic effect of the peptides. However, the correlation between cell-associated CS and cytotoxicity was not significant.
Both the expression of sialic acids, which is another component of the anionic glycoconjugate cell coat that surrounds cells, and the expression of PS in the outer membrane leaflet have been shown to affect the CAPs interactions with the lipid bilayer [
11,
55]. Moreover, the membrane fluidity has been demonstrated to be an important determinant for the selective permeabilization of membranes [
56‐
58].
In order to study the possible contribution of HS to the cytotoxic activity of the 9-mers more directly, the peptides cytotoxic activity was tested against CHO wild-type cells expressing HS on the cell surface and its mutant lacking HS on the cell surface. CHO cells have been widely used to study the role of cell surface GAGs in various processes such as viral infection, growth factor signaling and cell adhesion [
59]. The pgsA-745 cells have defective xylosyltransferase, an enzyme necessary for biosynthesis of HS and CS [
37]. Although CHO cells are derived from normal tissue, both CHO-K1 and pgsA-745 induce solid tumors when injected into immunodeficient mice [
60,
61]. By examining the expression pattern of GAGs on the cell surface of the CHO-K1 cells, we found that the cell surface PGs primarily contained HS chains. This expression profile, in which HS is the dominant type of cell surface GAGs, is common among most cell types [
27]. Our experiments with CHO cells clearly indicate that cell surface GAGs increase the cytotoxic effect of LTX-302 and LTX-318. However, the cytotoxic effect of LTX-315, which lysed the cells more efficiently, was not influenced by cell surface GAGs.
We found that soluble CS and HS inhibited the cytotoxic activity of LTX-302 and LTX-315 against the CHO cells. The stronger inhibition of the cytotoxic activity obtained by HS compared to CS indicates that the peptides bound more strongly to HS than to CS. This was confirmed by affinity chromatography, which exhibited a higher affinity of the peptides to HS compared to CS. The difference in the affinity could be explained by the higher conformational flexibility in HS compared to the more rigid CS [
62], as the peptides may require a high flexibility in the molecules they bind to. The cytotoxic activity of LTX-318 was not affected by the presence of soluble CS or HS. Considering the low activity that LTX-318 displayed against the CHO-K1 cells, the cytotoxic concentration of the peptide might be too high in order for the amount of exogenous CS and HS to affect the activity. The finding that the peptides interact more strongly with HS, together with the higher amount of HS chains attached to syndecans and glypicans compared to CS [
27,
28], strongly indicates that HS and not CS is the major interaction site for the 9-mers.
Despite having the same net positive charge, LTX-315 and LTX-318 showed a higher affinity for HS in comparison to LTX-302. The difference in affinity to HS may be due to the position of the basic residues in the peptides. In addition to cationic residues, the CAPs include lipophilic residues, which are important for interactions with the lipid layer of the cell membrane leading to an irreversible membrane destabilizing effect. The relative positions of the lipophilic and cationic residues affect the flexibility of CAPs, which permit the transition from its solution conformation to its membrane-interacting conformation [
63,
64]. Both the position of the cationic residues and the relative flexibility of the three 9-mers can therefore affect their interaction with cell surface GAGs.
The ability of the 9-mer peptides and LfcinB to interact with GAG chains will increase the cell surface concentration of the peptides. However, the finding that cell surface HS can act as a facilitator for small CAPs is in contrast to our recent report which shows that the longer lytic peptides LfcinB and KW5 displayed a higher cytotoxic activity against the GAG-deficient cell line [
29]. The inhibitory effect of GAGs on the cytotoxic activity of LfcinB could be due to the higher affinity for HS compared to the 9-mer peptides. The LfcinB peptide has a higher net positive charge (+8) than the 9-mers, which may explain its higher affinity for HS. However, it has been documented that the affinity for HS is only partly correlated with the net charge of the peptides [
45,
65]. Several studies have demonstrated that peptide analogues with arginine residues bind more tightly to heparin-like molecules than comparable analogues substituted with lysine [
65‐
68]. The 9-mers have no arginine residues in their sequences, while the LfcinB peptide contains five arginine residues. It is believed that the tighter interaction observed for arginine is due to a strong hydrogen bond formation between the guanidine group of arginine and sulfate. The presence of arginine residues in LfcinB might therefore also contribute to the higher affinity for HS compared to the 9-mers. The difference in the size of the 9-mers and LfcinB peptides could also affect the affinity for HS due to differences in the flexibility of the secondary structure. Whereas the LfcinB peptide forms a stabilized amphiphatic β-sheet, the smaller peptides might have a higher plasticity of their secondary structure, thus leading to a less defined binding domain for GAGs.
Hence, the difference in HS affinity between LfcinB and the 9-mers seems to affect the mechanism of action of LfcinB and the 9-mers differently. We therefore propose a mode of action model in which both the LfcinB peptide and the 9-mers are attracted to the anionic glycoconjugate cell coat that surrounds cells. This anionic cell coat consists of both GAGs and sialic acids. The repeating disaccharide structures of HS containing multiple sulfate groups are larger and more negatively charged than sialic acids, which is a monosaccharide with a carboxylic acid group. A stronger electrostatic interaction is therefore expected to occur between CAPs and HS in comparison to sialic acids. In order for the CAPs to exert their permeabilization effect leading to cell death, they have to navigate through this anionic cell coat to reach the phospholipid bilayer. The inhibitory effect of HS on the cytotoxic activity of LfcinB shows that the anionic cell coat may play a limiting role in the cytotoxic activity of LfcinB, in which HS at the cell surface of target cells hinders LfcinB from reaching the phospholipid bilayer. Furthermore, LfcinB that complex with cell surface HS may not be in close enough proximity to the cell surface to destabilize the membrane. The cytotoxic activity of the 9-mers is not inhibited by cell surface HS, thus suggesting that the 9-mers are attracted to HS without being captured. A higher amount of the 9-mer peptides will therefore reach the phospholipid bilayer compared to LfcinB.
Competing interests
ØR is director of Oncology Research for Lytix Biopharma.
Authors' contributions
BF carried out the chromatography studies, participated in the cytotoxicity studies and wrote the first draft of the manuscript. IL carried out the peptide synthesis, participated in the cell culturing and cytotoxicity studies. LUH and ØR designed the study and helped to draft the manuscript. All authors contributed to the discussion and interpretation of the results. All authors read and approved the final manuscript.