Anti-proliferative effect on a colon adenocarcinoma cell line exerted by a membrane disrupting antimicrobial peptide KL15

Abstract

The antimicrobial and anticancer activities of an antimicrobial peptide (AMP) KL15 obtained through in silico modification on the sequences of 2 previously identified bacteriocins m2163 and m2386 from Lactobacillus casei ATCC 334 by us have been studied. While significant bactericidal effect on the pathogenic bacteria Listeria, Escherichia, Bacillus, Staphylococcus, Enterococcus is exerted by KL15, the AMP can also kill 2 human adenocarcinoma cells SW480 and Caco-2 with measured IC₅₀ as 50 μg/ml or 26.3 μM. However, the IC₅₀ determined for KL15 on killing the normal human mammary epithelial cell H184B5F5/M10 is 150 μg/ml. The conformation of KL15 dissolved in 50% 2,2,2-trifluroroethanol or in 2 large unilamellar vesicle systems determined by circular dichroism spectroscopy appears to be helical. Further, the cell membrane permeability of treated SW480 cells by KL15 appears to be significantly enhanced as studied by both flow cytometry and confocal microscopy. As observed under a scanning electron microscope, the morphology of treated SW480 cells is also significantly changed as treating time by 80 μg/ml KL15 is increased. KL15 appears to be able to pierce the cell membrane of treated SW480 cells so that numerous porous structures are generated and observable. Therefore, KL15 is likely to kill the treated SW480 cells through the necrotic pathway similar to some recently identified AMPs by others.

Keywords:

• antimicrobial
• antiproliferation
• apoptotic
• anticancer
• membrane permeability
• necrotic
• peptide

Abbreviations

ABC	=	ATB-binding cassette
HPK	=	histidine protein kinase
HNSCC	=	head and neck squamous cell carcinoma
HFFF	=	human foetal foreskin fibroblast
PS	=	phospholipid phosphatidylyserine
TFE	=	2,2,2-trifluoroethanol
MTT	=	3-(4 5-cimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide

Introduction

The antimicrobial peptides (AMPs) are important defense substances produced by microorganisms to kill the invading bacteria for competing with the nutrients.¹ Despite their high sequence diversity, AMPs share some fundamental structural features such as short sized, carrying positive net charge, and possessing membrane permeability or amphiphilic nature which could facilitate them to insert into the cell membrane. The molecular mechanism of this last action is still not fully understood though the antimicrobial effect usually depends on it for causing the cell lysis or other intracellular effects.² The inhibition of protein synthesis is one of such other effects that has been reported. The AMPs produced by bacteria are commonly referred as bacteriocins which often exhibiting higher target specificity and stronger potency over those produced by the eukaryotes.^2,3 Due to their amphiphilic nature, the bacteriocins often trigger their initial interaction with the negatively charged bacterial membranes.

Besides the bactericidal effect, some AMPs are also known to kill the eukaryotic cells for instances for some cancerous cells. For examples, nisin has been used as a potential therapeutic for treating the head and neck squamous cell carcinoma (HNSCC) since it can induce the apoptosis, cell cycle arrest, and reduce the cell proliferation and tumorigenesis in vivo.⁸ Pyocin S2 produced by Pseudomonas aeruginosa is known to inhibit the growth of tumor cell lines HepG2 and Im9, while exerting no inhibitory effect on the normal cell line human foetal foreskin fibroblast (HFFF).⁹ Plantaricin A (PlnA) produced by Lactobacillus plantarum C11 is found to permeabilize the cell membrane of cancerous rat pituitary cells (GH(4) cells), whereas exhibiting no deleterious effect on the normal rat anterior pituitary cells.² Not only natural AMPs, some synthetic AMPs are also known to possess antiproliferation effects on some cancerous cells. For examples, D-K6L9 is a synthetic peptide that can bind with phosphatidylserine (PS) and then induce necrosis in some cancer cells. It can also inhibit the growth of murine melanoma tumors B16-F10 though terminating the treatment would result in tumor relapse. A more complete elimination of tumor cells in animal models has been achieved¹⁰ through a combined therapy involving D-K6L9, glycyrrhizin (an inhibitor of HMGB1 protein), BP1 peptide, and interleukin-12.

Unlike the conventional treatment by chemotherapy which is usually sabotaged by toxicity and frequent development of drug resistance effects, certain AMPs are found to avoid these shortcomings and emerging as a new promising class of natural-source drugs. Indeed, some AMPs exhibit selective cytotoxicity against a broad spectrum of cancer cells rather than the normal mammalian cells and erythrocytes. This may be ascribed to the fact that the cell membranes of cancer cells are often negatively charged while those of normal cells are often zwitterionic. The negative charges are due to the increased expression of membrane phospholipid PS and the glycoprotein O-glycosylated mucines.¹¹ The anticancer activity of AMPs is thought to be similar in mechanism to their antimicrobial activity which is often induced through membrane permeabilization.¹² Though some AMPs can trigger apoptosis in cancer cells via disrupting the mitochondrial membrane, most AMPs kill the tumor cells through a cell membrane lytic process. However, some AMPs are also known to be potent inhibitors of the blood vessel development (angiogenesis) associated with the tumor progression process.¹³ In the past 2 decades, numerous researches have been focused on how AMPs selectively permeabilize the tumor cells and how synthetic peptides may be designed to optimize their antimicrobial and antitumor activities and therapeutic capabilities.^2,14 Since the structure, cationic and hydrophobic characteristics of an AMP may determine its mode of action, direct modification of these features allows the rational design of potent AMPs.¹⁵

Recently, we have identified several novel bacteriocin loci from the genomic sequences of a completely sequenced lactic acid bacteria (LAB) Lactobacillus (L.) casei ATCC 334 (GI number: 116103724) which is otherwise not known as a bacteriocin producer.¹⁶ The DNA sequences of some of these putative AMPs such as m2163 and m2386 were cloned and heterologously expressed. We have found that both peptides exhibited antimicrobial activity against some lactobacilli and several Listeria species.¹⁶ In this work, we have treated peptides m2163 and m2386 as templates for gradually modifying their sequences with amino acid (aa) residues of higher hydrophobicity and helix-forming propensity to find a synthetic AMP KL15 that can exert strong antiproliferation effects on some human colon adenocarcinoma cell lines such as SW480 and Caco-2. The IC₅₀ determined by the MTT assay for the AMP treated cancer cells is around 50 μg/ml or 26.3 μM. We then perform a series of flow cytometry and Western blot analyses to investigate whether the apoptosis pathway in the treated SW40 cells is induced by KL15 or not. A study by circular dichroism (CD) spectroscopy is also performed to find that the conformation of KL15 is helical in a membrane mimicking solution containing 50% 2,2,2-trifluoroethanol (TFE) or in some large unilamellar vesicle (LUV) made by sodium dodecyl sulfate (SDS) or dodecylphosphocholine (DPC). Moreover, the morphology of treated SW40 cells appears to be significantly altered even accompanied by the formation of numerous porous structures on the cell membranes as examined by a scanning electron microscopy (SEM). This result is confirmed by further studies using a confocal microscopy and flow cytometry that KL15 is able to damage the cell membrane and causing the leakage or enhancement in membrane permeability and eventually the cell death.

Results

In silico modification of peptide m2163 and m2386 to find AMP KL15

Without the associated immunity and ABC transporter and also lacking the characteristics of class IIa, b, c sequences as according to the classification scheme given by Nissen-Meyer et al.³ peptide m2163 can only be recognized as a class IId like bacteriocin. Further, peptide m2386 only possesses partial characteristics of a pheromone peptide since its associated HPK and ABC sequences are incomplete. However, the sequences of these 2 peptides were gradually modified and characterized for the positive net charge, hydrophobicity, and predicted secondary structures using some online available tools. As shown in Tables 1 and S1, there are 6 such modified peptides designated as m1 (m2163–4), m2 (m2163–6), m3 (m2386–2163–4(7)), m4 (m2386-KL15–1(8)), m5 (m2386-KL15–2(9)), and KL15 being chosen for experimental testing. These six peptides are synthesized and purified and their antiproliferation activity against SW480 cells are measured using the MTT assay. The IC₅₀ determined for each of these peptides are 60, 70, 60, 110, 100, and 50 µg/ml, respectively. Peptide KL15 is then chosen for further studies since its antiproliferation activity is among the best determined. To examine whether peptide KL15 can selectively kill the cancerous but not normal cells, we have treated a normal human mammary epithelial cell line H184B5F5/M10 plus an another cancerous cell line Caco-2 with the peptide and then employing the MTT assay on each of them. In these assays, the survival rate corresponding to the negative control treated only by doubly deionized water (ddw) is set as 100% while that treated by triton X-100 is set as positive control and is set as 11%. As shown in Figure 1A, both treated cancerous cells show a more significant decrease in the survival rate than the normal cells after they both being treated with peptide KL15 for 24 h. This result also shows that the effect of KL15 on Caco-2 is close to that on SW480 cells (Fig. 1A). The IC₅₀ determined for both Caco-2 and SW480 cells are around 50 while that determined for H184B5F5/M10 cells is around 150 μg/ml (Fig. 1A). The IC₅₀ for peptide KL15 is determined again using the trypan blue assay on the SW480 cells. As shown in Figure 1B, both MTT and trypan blue assays agree with each other on the IC₅₀ measured which is around 50 μg/ml or 26.3 μM though lower survival rate is observed for the former.

Figure 1. (A) The MTT assay performed for human colon adenocarcinoma cell lines SW480 and Caco-2 and a normal humman mammary epithelial cell line H184B5F5/M10. Each of these cells are treated with increasing concentration of KL15 from 20 to 100 μg/ml for 24 h. (B) Comparing the MTT and trypan blue assay results for the treated SW480 cells by increasing amount (0, 12.5, 25, and 50 µg/ml) of KL15 used. The statistical significance of the data was determined by the independent-samples t-test with ′*′, ′**′, and ′***′ denoting p < 0.05, p < 0.01, and p < 0.001, respectively.

Table 1. Peptide sequences, net charge, pI, hydrophobicity, hydrophobic moment, and secondary structure content estimated for peptides m1, m2, m3, m3, m4, m5, and KL15

peptide ID and sequences	net charge	pI	hydrophobicity	Hydrophobic moment	secondary structure content shown in % of -helix/-sheet/coil
m1 RLYRLYFRLYRLWRRLTL	+6	11.72	0.714	0.616	80.8%/19.2%/0.0%
m2 RLWRLAWRLYRLWRLTL	+5	12.18	0.881	0.543	63.2%/36.8%/0.0%
KL15 KRKLYKWFAHLIKGL	+5	10.58	0.491	0.667	100.0%/0.0%/0.0%
m3 KGIRDWLRGFNKLRKWFLRLIKGL	+7	12.02	0.450	0.751	60.3%/39.1%/0.6%
m4 LRRLDRWLRLFNRLRRWFLRLLRGL	+8	12.54	0.517	0.694	100.0%/0.0%/0.0%
m5 LRLLDRWLRLFNRLLRWLRLLRGL	+6	12.40	0.690	0.570	84.0%/16.0%/0.0%

Determining the secondary structure by CD for AMP KL15

It is known that the conformation of some anticancer AMPs are helical.^17,18 Here, the conformation of KL15 and the other 5 peptides are determined in 3 different solvent systems by CD since the content of secondary structures can be conventionally estimated from a CD spectra. The presence of α-helix or β-sheet is usually characterized by a huge negative peak at wavelength 222 or 217 nm, while the presence of a weak positive peak at 220–230 nm is indicative of random coil. However, the existence of some aromatic residues on a peptide sequence may enhance the aromatic rotation and leading to the change in peak features.¹⁹ As shown in Figure 2A–2F, the conformation detected for KL15 and all the other 5 peptides dissolved in either ddw or PBS buffer appear to be random coil while that dissolved in 50% TFE appears to be helical. The CD spectra for KL15 dissolved in 2 LUVs made from SDS or DPC are also determined. As shown in Figure 2G, the conformation of KL15 in these 2 LUVs also appears to be helical.

Figure 2. The CD spectra determined for peptide (A) m1, (B) m2, (C) m3, (D) m4, (E) m5, and (F) KL15 dissolved in aqueous solution (represented by solid circle ), PBS buffer (represented by open circle ), and 50% TFE solution (represented by solid triangle). The concentration of each peptide solution is fixed as 60 µM. (G) The CD spectra for KL15 in 2 LUVs prepared from SDS or DPC.

Cell cycle analyses by flow cytometry

To quantify the subpopulation of SW480 cells that may be damaged by KL15 treatment, the flow cytometry is used to analyze the treated cells stained by 20 µg/ml fluorescence dye propidium iodide (PI). Through this analysis, a DNA profile representing cells in G1, S, and G2/M phase can be observed with the damaged cells being represented by a sub G₀/G₁ population. We define the ratio between the sub and normal G₀/G₁ peak as the % sub-G1 population to characterize the amount of damaged cells. As shown in Figure 3A, the % sub-G1 population detected gradually increases from 1.49 to 5.05% as the treatment time is increased from 1 min to 24 h. However, when the treatment time is fixed at 24 h, the % sub-G1 population detected gradually increases from 3.74 to 4.40% as the concentration of KL15 is increased from 40 to 120 µg/ml (Fig. 3B).

Figure 3. The cell cycle of treated SW480 cells by (A) 80 µg/ml KL15 with increasing time from 0 min to 24 h and (B) with increasing concentration from 40 to 120 µg/ml for 24 h are analyzed by flow cytometry using fluorescence dye PI. The sub G1 regions (marked by M3 bar) where DNA in the treated cells is significantly damaged are quantified by the bar chart shown on the right.

Flow cytometry analyses using fluorescence dyes PI and AnnexinV-FITC

it is possible to distinguish the early (early apoptosis) from the late (apoptosis-necrosis) phase of damaged cells by KL15 through a combined use of both fluorescence dyes PI and Annexin V- fluorescein isothiocyanate (FITC). The fluorescence dye Annexin V-FITC can bind with an early apoptosis marker PS which will flip from inside the cell membrane to the outside surface during the early stage of apoptosis,²⁰ while fluorescence dye PI can penetrate the membrane and bind to DNA in the late stage of apoptosis.²¹ The SW480 cells are treated with an increasing amount of KL15 from 40 to 120 μg/ml for 24 h. The cells are then harvested and stained with both Annexin V-FITC and PI for analyzing by the FACS flow cytometry. The dot plots obtained correspond to the distribution of cell population in 4 different phases namely late apoptotic or necrotic (quadrant UR where Annexin V-positive/PI-positive signals are exhibited), primary necrotic (quadrant UL where Annexin V-negative/PI-positive signals are exhibited), viable (quadrant LL where Annexin V-negative/PI-negative signals are exhibited), and early apoptotic (quadrant LR where Annexin V-positive/PI-negative signals are exhibited).²² The cell numbers in each quadrant are counted and then divided by the total number of cells counted to give the % of cells in each quadrant. As shown in Fig. 4B–4D, as the concentration of KL15 is increased from 40 to 120 µg/ml, the % of cells in quadrant UR which stands for late apoptotic or necrotic is also increased. This would imply that treatment by KL15 will render the treated cells entering into the apoptotic or necrotic phase. However, some significant amount of cells (4.2%) are detected in quadrant UL which stands for primary necrotic when the treatment time for SW480 cells by 50 µg/ml is shortened to 2 h. (Fig. 4E). This indicates that treatment by KL15 would immediately cause some treated cells entering into the necrotic phase.

Figure 4. The flow cytometry analyses performed for the SW480 cells treated with (B) 40, (C) 80, and (D) 120 µg/ml KL15 for 24 h using 2 fluorescence dyes PI (represented by FL2 axis) and Annexin V-FITC (represented by FL1 axis). (A) is the control without any treatment by KL15. (E) is the treatment by 50 µg/ml KL15 for 2 h. The number or % of primary necrotic cells, the late apoptotic or secondary necrotic cells, the viable or live cells, and the cells undergoing early apoptosis are characterized by dots counted in quadrant UL, UR, LL, and LR, respectively.

It is known that some AMPs are able to damage the cell membrane and increase the cell membrane permeability leading to the cell death.²³ This phenomenon can be studied using fluorescence dye PI alone since it can intrude into the cell when the cell membrane integrity is lost. The SW480 cells are treated with 3 different dosages 40 to 120 µg/ml of KL15 for different time lengths from 0 to 24 h. With 40 µg/ml KL15 used, the fluorescence intensity detected is barely changed in all the time lengths treated as compared with that of the control (Fig. S1A). However, when the treatment dosage for KL15 is increased up to 80 and 120 µg/ml, there is an apparent right shift in fluorescence intensity detected in all the time lengths treated as shown in Fig. S1B and S1C. The results imply that under these treatments the cell membrane is damaged to allow significant amount of PI entering into the cells. However, this effect does not sustain long since no further increase in fluorescence intensity is observed for longer treatments for 12 or 24 h for both higher KL15 concentrations used as shown in Fig. S1B and S1C.

Cell morphology examined by SEM

The cell morphology of treated SW480 cells by 80 μg/ml KL15 for different time lengths from 0 to 24 h is examined by SEM. As shown in Figure 5A, the cell morphology of untreated SW480 cells appear to be intact and plenty of cilia can be seen around the cell. On the contrary as shown in Figure 5B to 5D, the morphology of treated SW480 cells becomes significantly changed as treating time by 80 μg/ml KL15 is increased. The treated SW480 cells for longer time by 80 μg/ml KL15 appear to be pierced to generate numerous porous structures and also somewhat shattered with loss of some cilia around the cells (Fig. 5B–D).

Figure 5. The morphology of treated SW480 cells by 80 µg/ml KL15 for different time lengths from (A) 0 h, (B) 1 h, (C) 12 h, and (D) 24 h was examined by SEM.

Confocal microscopy examination

The distribution of KL15 in the treated SW480 cells is also investigated using the confocal microscopy. To conduct this study, AMP KL15 is labeled with a green fluorescence dye N-hydroxysuccinimide (NHS)-fluorescein, the cell membrane is labeled with a red fluorescence dye Di-8-ANEPPS, and the cell nucleus is labeled with a blue fluorescence dye DAPI. The control group or background fluorescence is made by green fluorescence dye FITC alone since the NHS-ester moiety can be easily hydrolyzed and become inactive. The image of cell border can be approximately revealed by the red fluorescence. A change in fluorescence color is resulted when up to 2 fluorescence signals with similar intensity are coincided at the same location which is designated as the co-localization. The location of KL15 in the treated SW480 cells may be identified through the co-localization process. As shown in Figure 6A for the untreated cells, the background green fluorescence is well coincided with the red fluorescence used to label the cell membrane. The cell membrane appears to be intact (Fig. 6A). However, for the treated cells shown in Figure 6B, except that there are significant amount of KL15 entering into the cells, the cell morphology is also dramatically altered. Moreover, some red fluorescence is even co-localized with DAPI which is the blue fluorescence used for staining the nucleus. This would agree with that observed by SEM (Fig. 5) that KL15 can damage the treated cells through penetrating the cell membrane and intruding into the cells. Note that the cell-killing ability of fluorescein-labeled KL15 is nearly the same as that of KL15 alone as shown in Fig. S2 where the MTT assay results for the treated cells by 25 and 50 µg/ml either peptide are compared.

Figure 6. The confocal laser scanning microscopy images of (A) the untreated and (B) treated SW480 cells by 80 µg/ml KL15 for 12 h. The images of differential interference contrast (DIC) and merge are also separately shown. The fluorescence colors are defined as: green, NHS-fluorescein for peptide KL15; green background, FITC; red, Di-8-ANEPPS for the cell membrane; and blue, DAPI for the cell nucleus. The co-localization is observed when up to 2 fluorescence signals are overlapped at the same place such that the colors of corresponding fluorescence signals are changed.

Western blot analyses

The expression of some apoptotic (caspase 3, caspase 6, and caspase 8) and necrotic related (IRE1α, HIF1α, RIP3, CHOP, HMGB1 and Cyclophilin A) proteins possibly induced in the treated SW480 cells with 40, 80 and 120 μg/ml of KL15 for 24 h are analyzed by Western blot and presented in Figure 7. The expression of β-actin is used as an internal control in this study. As shown in Figure 7A, except the expression of RIP3 or IRE1α and CHOP is slightly increased or decreased, no apparent change in expression of other proteins is detected when the treatment concentration by KL15 is increased from 40 to 120 μg/ml (Fig. 7A and B)

Figure 7. The expression of some (A) apoptotic (caspase 3, caspase 6, and caspase 8) and (B) necrotic (IRE1α, HIF1α, RIP3, CHOP, HMGB1 and Cyclophilin A) related proteins possibly induced in the treated SW480 cells with 40, 80 and 120 μg/ml of KL15 for 24 h is analyzed by Western blot. The expression of β-actin is used as an internal control.

DISCUSSION

In this study, we have found that AMP KL15 obtained from gradual modification of the sequence of peptide m2163 from L. casei 334 exhibits selective antiproliferation activity on 2 human colon adenocarcinoma SW480 and Caco-2 but not the normal human mammary epithelial H184B5F5 cells. While increasing the net positive charge may enhance the interaction between the modified AMP with the anionic cell membrane, placing more cationic characteristic may not be the right choice.²⁴ The net charges for about 2000 bacteriocins collected in the APD2 database²⁵ are only ranging from +2 to +4. Further, the net charges of peptide m2163 and m2386 previously identified by us¹⁶ are +3 and +2, respectively. The interaction between an AMP and cell membrane is also affected by hydrophobicity or hydrophobic moment of the peptide. While the predicted hydrophobic moment 0.667 for KL15 is rather high, the predicted secondary structure for the peptide is also 100% helical (Table 1). In fact, the secondary structure measured by CD for all the 6 peptides studied (Fig. 2) is helical in either 50% TFE or 2 LUVs. These indicate that KL15 may behave like an amphipathic AMP for penetrating into the lipid bilayer of cell membranes.²⁶ Since higher content of anionic phospholipids and mucins are existing in the cell membranes of prokaryotic and cancer cells, the transmembrane potentials of these 2 cell species are usually high. On the contrary, the normal eukaryotic cells possess lower transmembrane potential due to that numerous zwitterionic phospholipids and cholesterols are existing in the membranes. Higher cholesterol content in normal eukaryotic cell membranes may prevent the peptide-induced membrane disruption through increasing cohesion and stiffness of the lipid bilayer.²⁶ Therefore, most AMPs can preferentially disrupt the prokaryotic and cancer rather than the normal eukaryotic cell membranes.²⁷

Through examining by SEM, a significant morphological change in the treated SW480 cells by KL15 is observed (Fig. 5) which is similar to that exerted by AMP magainin II on the bladder cancer cells.²⁸ In this latter study, the adherent smooth membrane surface is observed for the untreated while for the treated cells the disrupted and porous membrane is observed.²⁸ The membrane disruption effect on the treated cells by a hybrid AMP NS has also been reported.²⁹ The membrane permeability measured by flow cytometry for the treated SW480 cells by KL15 is also increased when both treatment dosage and time are increased (Fig. S1), indicating that the treated cells may undergo severe change in cytoskeleton (Fig. 5), loss in ion homeostasis, and malfunctioning in organelles.³⁰ The amphiphilic and helical nature of AMPs may enable them to target the nonpolar lipid cell membrane such that some ion-dissipating channels are formed which would lead to the depolarization and irreversible cytolysis of treated cancer cells.²⁸ Indeed, It has been reported that some AMPs can kill cancer cells through cell membrane-lytic process though some AMPs may trigger apoptosis via disrupting the mitochondrial membrane of cancer cells.¹³

Based on the analyses by PI-Annexin V double staining (Fig. 4), the treated SW480 cells by KL15 may proceed to the necrotic process since the detected signal for the first quadrant UR is increased with the increase of treatment dosage, while that detected for the fourth quadrant LR is low and rather unchanged (Fig. 4). Moreover, as shown in Figure 4E, about 4.2% of the total cells would undergo the necrosis process when the treatment time is only 2 h. This indicates that there are not many cells undergo the early apoptotic process despite that the treatment dosage has been gradually increased. Recently, a non-apoptotic cell death pathway or regulated necrosis model³¹ has been proposed involving the following 7 protein species: (i) pathophysiological stimuli induced by hyperactivation of poly(ADP-ribose) polymerase (PARP-1), (ii) mitochondrial permeability transition, (iii) mitochondrial complex I, (iv) Cys/Glu antiporter, (v) necrosome, (vi) NADPH oxidases, and (vii) inflammasome. To characterize the cell death pathway induced in the treated SW480 cells by KL15, we employed several commercially available antibodies of early apoptosis (caspase 3, caspase 6, and caspase 8) and regulated necrosis markers (IRE1α, HIF1α, RIP3, CHOP, HMGB1 and Cyclophilin A) for performing the Western blot analyses. However, no apparent change in expression is detected for most of the protein species studied except RIP3, IRE1α, and CHOP is up or down regulated somewhat when the treatment concentration by KL15 is increased from 40 to 120 μg/ml (Fig. 7). However, a 89 kDa cleaved fragment of PARP-1 was detected by Western blot analysis on the treated SW480 cells by 50 μg/ml KL15 for 0, 6, 12, to 18 h as shown in Fig. 8A. Since no apparent increasing in expression of caspases is detected (Fig. 7B), the cleavage on PARP-1 could be due to the proteases released by lysosomes during the necrosis process. Moreover, Fig. 8B shows that the release of lactate dehydrogenase (LDH) by treated SW480 cells through only 2 h treatment with various amount of KL15 can also be detected. The % cytotoxicity measured increases from 18 to 62% when the treatment dosage is increased from 12.5 to 100 μg/ml (Fig. 8B). Therefore, we believe that KL15 treatment can cause the treated SW480 cells entering into the necrosis process. A peptide ABP-CM4 isolated from the Chinese silkworm has been found to kill the leukemia cells K562, U937, and THP-1 through only disrupting the cell membrane and without penetrating into the cells.³² However, a hybrid peptide P18 is known to kill the human melanoma cells not only through disrupting the cell membrane but also inducing the necrosis process.³³

Figure 8. (A) Detection of a 89 kDa cleaved fragment of poly(ADP-ribose) polymerase (PARP-1) by Western blot. The SW480 cells were gradually treated by 50 μg/ml KL15 for 0, 6, 12, to 18 h before being subjected to the Western blot analysis. (B) Detection of the release of LDH by treated SW480 cells through 2 h treatment with 12.5, 25, 50, and 100 μg/ml KL15 by a LDH detection kit.

Conclusion

We have shown in this work that AMP KL15 modified from the sequence of a previously isolated peptide m2163 from L. casei 334 exhibits selective antiproliferation activity on 2 human colon adenocarcinoma SW480 and Caco-2 but not the normal human mammary epithelial H184B5F5/M10 cells. Comparing with some small molecule drugs already in market for treating the colorectal cancer, there are certain favorable characteristics for the peptide drugs developed due to their better safety, good selectivity, and lower drug resistance encountered. However, there are still numerous associated challenges namely low oral bioavailability, low product stability, and short half-life ahead need to be overcome. We think AMP KL15 found in this work might be a good start to develop a more potent and safer peptide drug to treat the colorectal cancer. It is known that substituting some L- by D-form aa residues may reduce the hemolytic whereas cause little effect on the anticancer activity. Although better proteolytic stability may be obtained by creating sequences with mixed D,L-form aa for the AMP, it is still necessary to study how the AMP binds with the cancer cell and what is the underline mechanism of killing.

Methods

In silico modification of peptide m2163 and m2386

How similar were the sequences of peptide m2163 and m2386 previously identified by us¹⁶ to the known anti-cancer AMPs listed in database BACTIBASE³⁴ was examined first using the alignment tools such as FASTA, BLAST, and SSearch. The sequences of m2163 and m2386 were then set as templates for gradually replacing with aa residues of higher hydrophobicity and better helix forming propensity. The modification process was also assisted by the following web available tools HeliQuest,³⁵ ProtParam,³⁶ and SSCP³⁷ for monitoring the resultant hydrophobicity, pI, and secondary structure predicted for the modified sequences.

CD Spectroscopy

AMP KL15 (>95% purity) was synthesized by solid-phase method using resin 4-methylbenzhydrylamine by Kelowna International Scientific (Taiwan). All the other synthetic peptides (>85% purity) m1, m2, m3, m4, and m5 were purchased from MDBio incorporation (Taiwan). The composition and molecular weight were determined by HPLC and mass spectroscopy analyses. The concentration of each peptide solution prepared was measured by NanoDrop ND-1000. Peptide stock solutions were kept in −20°C while working peptide solutions were dissolved in sterile and double distilled water (ddw) and stored in 4°C. The secondary structure of 60 μM KL15 dissolved in ddw, PBS, or 50% TFE and all the other 5 peptides m1, m2, m3, m4, and m5 treated in the same manner were measured by Far-UV (190–260 nm) CD spectroscopy on an AVIV 202 spectropolarmeter. The light path length used was 1 mm and the measuring temperature was fixed at 25°C. Each spectrum automatically recorded by the spectropolarmeter was an average of 3 scans taken by a step size of 0.5 nm and presented as molar ellipticity against wave length. The LUV systems were prepared by suspending SDS (Sigma) or DPC (Sigma) in 10 mM Tris and 20 mM NaCl buffer solution (pH 7.3) to produce 5 mM dispersions and then subjected to 3 freeze/thaw cycles. The homogeneous dispersions were then extruded 10 times through an Avanti Mini-Extruder (Alabaster) using 0.4 μm (polycarbonate) filters (Alabaster), followed by 10 times through 2 stacked 0.1 μm filters to produce LUVs of 100 nm diameter. Then, some appropriate volumes of KL15 and LUVs were mixed to produce 300 μL samples with lipid to peptide molar ratio fixed as 30:1.³⁸ All the samples were allowed to equilibrate at 25°C prior to CD measurement.

MTT, trypan blue, and LDH releasing assays

The human colon adenocarcinoma cell line SW480 (BCRC60249, Taiwan) was cultured in Leibovitz's L-15 Medium (GIBCO BRL) containing 10% fetal bovine serum (FBS) and 10% antibiotics in 37°C and 5% CO_2. The same culturing conditions were used for culturing the human colon adenocarcinoma cell line Caco-2 (BCRC67001, Taiwan) except that the medium used was Minimum Essential Medium (GIBCO BRL) containing 10% FBS. The human mammary epithelial cell line H184B5F5/M10 (BCRC60197, Taiwan) was cultured in Dulbecco's Modified Eagle Medium (GIBCO BRL) containing 15% FBS, 1.5 g/L sodium bicarbonate, and 10% antibiotics, in 37°C and 5% CO₂. The cell viability for cells being treated with KL15 or other peptide solutions was determined using the MTT (3-(4, 5-cimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay. There were different peptide concentrations added into the cell culture plates with each carrying 5 × 10⁴ to 10⁵ cells and incubated at 37°C for 24 h. Then, the culture medium was removed and PBS was added for washing the cells before they were incubated with a MTT solution (0.5 mg/ml) (Sigma) for 4 h. Thereafter, the MTT solution was removed and 100 μl DMSO was added for a further incubation for 20 min. An ELISA reader (TECAN, Sunrise™) was used to measure the OD value at 565 nm. The experiments were performed in quartet for eliminating the extremum values. The cell viability was determined as % cell viability = OD of experimental group /OD of control group. To proceed with trypan blue assay, about 5 × 10⁵ cells were seeded in a 6-well plate and incubated overnight. The cells were then treated with different concentrations of KL15 or other peptide solutions. After 24 h of incubation, the cells were harvested and the same volume of trypan blue solution was added. The viable cells were counted using a hemocytometer through the formula: viable cell count (cells/ml) = average cell amount (cells/grid) × dilution factor × hemocytometer factor. The Cytotoxicity Detection Kit^plus (Roche) was also used to detect if there was any LDH released³⁹ from the cytosol of treated SW480 cells by KL15. About 5 × 10⁵ SW480 cells were seeded in each well of a 96-well plate. The grown cells were then treated with 12.5, 25, 50, and 100 μg/ml KL15 dissolved in a non-serum containing medium for 2 h. Cells without treatment by KL15 or treated by a Lysis Reagent provided by the kit were served as low or high control. To determine the LDH released, 100 μl freshly prepared Reaction Mixture was added into each well and incubated 10 min at 25 °C under dark. 50 μl Stop Solution was then added into each well, thereafter the plate was shaken for 10 sec. The absorbance at 492 nm was measured by an ELISA reader (TECAN, Sunrise™). The % cytotoxicity was calculated from the average of triplicate samples and controls as follows: % cytotoxicity =(OD of samples—OD of low control)/(OD of high control—OD of low control).

Flow cytometry analyses by fluorescence dye PI

After treating 2 × 10⁶ SW480 cells with different concentrations 40, 80 or 120 μg/ml of KL15 solutions for 24 h, the cells were trypsinized and washed with PBS and then centrifuged at 2000 g for 5 min. The supernatant was discarded while the pellet was suspended in 70% ethanol and mixed well at 4°C for more than 12 h. The solution was then centrifuged at 2000 g for 5 min to remove the supernatant. The cell DNA was stained with 20 μg/ml Propidium iodide (PI) solution containing 50 μg/ml RNase in PBS and reacted in dark for 30 min at 25°C. A FACS Calibur Flow Cytometry (BD) was used to measure the cell fluorescence intensity. To perform the cell membrane permeability tests, about 2 × 10⁶ SW480 cells were treated in different time lengths (1 min, 1 h, 12 h, and 24 h) with different doses 40, 80, 120 μg/ml of KL15 solutions using H₂O₂ as a positive control. The treated cells were incubated with 20 μg/ml PI for 30 min. Then, the cells were harvested for analyzing the released fluorescence using the FACS Calibur Flow Cytometry (BD).

Flow cytometry analyses by fluorescence dyes PI and AnnexinV-FITC

For proceeding with Annexin V and PI double staining experiments, 2 × 10⁶ SW480 cells were treated with different concentrations 40, 80, and 120 μg/ml of KL15 solutions for 24 h. The cells were trypsinized and mixed with 20 μg/ml Annexin-V-FITC and 5 μg/ml PI and then incubated at 25°C for 15 min. The untreated cells were used as negative while cells treated with an apoptosis inducer camptothecin (Selleckchem) were used as positive controls. The released fluorescence was analyzed by the FACS Calibur Flow Cytometry (BD). The number of necrotic, living, early apoptotic, and late apoptotic or dead cells detected were characterized by dot plots where the number of dots counted in 4 different quadrants, the upper left (UL), the upper right (UR), the lower left (LL), and the lower right (LR), represent the primary necrotic cells, the late apoptotic or secondary necrotic cells, the viable or live cells, and the cells undergoing early apoptosis, respectively.²²

SEM examination

About 4 × 10⁴ cells were grown on several 0.12–0.17 mm thick glass coverslips in a 12-well plate. The cells were fixed with 2.5% glutaraldehyde dissolved in a 100 mM phosphate buffer (pH 7.2). This was followed by osmication with 1% osmium tetroxide dissolved in the same buffer for 30 min. After rinsing twice with the same buffer for 15 min, the cells were dehydrated in increasing concentration (50%, 70%, 90%, and 100%) of ethanol before being dried at 80°C overnight. The coverslips were taped on a specimen holder and then sputter-coated with 20 nm of gold in an ion coater and examined by a HITACHI S-4700 scanning electron microscope.

Confocal microscopy examination

AMP KL15 was labeled at its primary amino groups in pH 7–9 buffers with 1 mg NHS-fluorescein dissolved in 100 μL DMSO at room temperature for 1h. There were 2 × 10⁵ cells grown on several glass coverslips in a 6-well plate. The grown cells were then treated with 40 μg/ml KL15 dissolved in a non-serum containing medium for 12 h. The medium was removed and the cells were rinsed 3 times with PBS. The cells were fixed with some formalin dissolved in PBS for 15 min. After washing with PBS 3 times, 2.5 μM di-8-ANEPPS was added and incubated in dark for 15 min. The cells on the coverslips were then sealed with a mounting gel containing DAPI for staining the cell nuclear. The cell fluorescence images were observed using a confocal microscope LSM 780 (Carl Zeiss). AMP KL15 was also labeled with FITC through conjugation with a functional group 6-aminocaproic acid attached at its N-terminus (Yao-Hong BioTech Inc.., Taiwan). The FITC labeled peptides were purified by high performance liquid chromatography with a purity of 96.54% obtained.

Western blot analyses

About 5 × 10⁶ SW480 treated cells by 40, 80, or 120 μg/ml of KL15 solutions for 24 h were lysised with 300 μl ProteinSeeker Mammalian Cell Lysis Solution (GenDEPOT) and 3 μl 100× protease inhibitors (Amresco). The samples were briefly centrifuged to collect the extracted proteins in supernatant which were then analyzed by SDS-PAGE and transferred onto a polyvinylidene fluoride (Pall) membrane with 250 mA for 90 min. The transferred membrane was soaked into 20 ml TBS (10 mM Tris pH 7.5, 2.5 mM EDTA pH 8.0, 50 mM NaCl, and 1% Tween-20) buffer containing 5% skim milk for 90 min to block the nonspecific binding. Different specific primary antibodies properly diluted in TBS buffer were added and incubated at 4 °C for more than 12 h. The samples were washed 3 times with 20 ml TBS buffer and followed by adding the secondary peroxidase-conjugated anti-rabbit IgG antibody (Jackson) diluted in 1:10000 with TBS for reaction with shaking at 75 rpm for 2 h. After removing the buffer, 1 ml ECL (Advansta) mix A and B agents were added onto the transferred membrane and incubated at 25°C for 2 min. The signals were detected and analyzed using a luminescent image analyzer ImageQuant™ LAS 4000 mini (GE Healthcare).

Data analysis

All the measured data were presented as mean ± SD using the SPSS software package (SPSS Standard version 16.0, SPSS Inc.). The independent-samples t-test implemented in the package was used to evaluate the significance of difference between the treatment and control groups. P value < 0.05 was considered as statistically significant.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

YC Chen has performed most of the cell works including all the flow cytometry and Western blot analyses, TL Tsai has performed SEM and confocal microscopy analyses, XH Ye has conducted the in silico modification, CD characterization, and antimicrobial activity tests, TH Lin has organized and written the manuscript.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.

Funding

This work is supported in parts by a grant NSC102–2313-B007–001-MY3 from the National Science Council, Taiwan ROC.