Tuning of Peptide Cytotoxicity with Cell Penetrating Motif Activatable by Matrix Metalloproteinase-2.

Although diverse cell penetrating motifs not only from naturally occurring proteins but also from synthetic peptides have been discovered and developed, the selectivity of cargo delivery connected to these motifs into the desired target cells is generally low. Here, we demonstrate the selective cytotoxicity tuning of an anticancer KLA peptide with a cell penetrating motif activatable by matrix metalloproteinase-2 (MMP2). The anionic masking sequence introduced at the end of the KLA peptide through an MMP2-cleavable linker is selectively cleaved by MMP2 and the cationic cell penetrating motif is activated. Upon treatment of the peptide to H1299 cells (high MMP2 level), it is selectively internalized into the cells by MMP2, which consequently induces membrane disruption and cell death. In contrast, the peptide shows negligible cytotoxicity toward A549 cancer cells with low MMP2 levels. Furthermore, the selective therapeutic efficacy of the peptide induced by MMP2 is also corroborated using in vivo study.

Introduction

Various peptides have been widely employed to prepare a myriad of therapeutic agents due to their unique properties such as cell or tissue penetration, responsiveness toward specific biomolecules, and selective binding toward targeted cellular constituents.[1−5] Furthermore, protease-mediated activation is widely utilized for construction of therapeutic peptides.[6−9] In particular, cell penetration capability is one of the most essential features of peptides for utilization as therapeutic agents, which significantly improves the efficacy of the therapeutic agents possessing peptides by enhancing the intracellular uptake.[3−5] In general, cell penetrating peptides possess numerous lysine and/or arginine residues, of which positive charge can electrostatically interact with the negatively charged cell membranes. Upon membrane binding, the penetration of peptides could occur via endocytosis or direct translocation. Furthermore, cell penetrating peptides can deliver cargo covalently bound to the peptide into the cytoplasm of cells.[3,4]

Since the discovery of the penetrating capability of human immunodeficiency virus-1 Tat protein through mammalian cell membranes,[10] numerous cell penetrating peptides have been discovered and developed including penetratin, transportan, and polyarginine sequences.[11] However, using these cell penetrating peptides, the specificity of cargo delivery connected to the peptides into the desired target cells is generally low. To overcome this limitation, numerous efforts have been made to enhance the selectivity of cell penetrating peptides while retaining their penetrating ability such as development of activatable or cell-type-specific cell penetrating sequence and combination with tumor targeting or transducible agents.[3,12−17] In particular, activatable cell penetrating peptides consisting of cationic polyarginine and anionic poly(glutamic acid) units have been used for selective cargo delivery.[18] The two domains were conjugated with the PLGLAG sequence for selective cleavage by matrix metalloproteinase-2 (MMP2). Therefore, the negatively charged domain could effectively prevent the positive charge of the polyarginine unit from binding to the cellular membranes for penetration. Upon cleavage of the PLGLAG sequence by MMP2, the masking domain was removed from the peptide and the cell penetrating capability was recovered. Consequently, several small cargos, such as a fluorescence dye and MRI contrast agent connected to the end of the cell penetrating sequence, could be selectively delivered into the cytoplasm of cells with upregulated MMP2 expression.[18−20] However, more extensive research on the selective delivery of large cargos such as therapeutic peptides by employing this type of activatable cell penetrating peptide is required for diverse applications.

The KLA peptide with the (KLAKLAK)2 sequence is an antimicrobial peptide designed de novo from melittin, a major element of honey bee venom.[21] The KLA peptide has an amphiphilic chemical structure that forms an α-helix on negatively charged cellular membranes.[22,23] Therefore, this peptide could form a helical pore on the mitochondrial or plasma membrane of prokaryotic microbial cells, leading to apoptotic or necrotic death of the microbe.[23−27] Although the KLA peptide could be used as an anticancer agent for several cancer cells with highly negatively charged lipid membranes,[28−30] targeting or cell penetrating motifs are typically required for the successful utilization of the KLA peptide as an effective anticancer agent owing to its low endocytic capability.[22,23,30−33]

In this study, we adopted the concept of an activatable cell penetrating peptide for triggered intracellular uptake and selective cytotoxicity of the KLA peptide by the MMP2 enzyme, as shown in Figure . To the end of the KLA peptide, a cationic cell penetrating sequence (polyarginine, R7) was conjugated. The anionic masking sequence [poly(aspartic acid), D7] was introduced at the other end of the KLA peptide using an MMP2-cleavable linker (PLGLAG sequence). The cell penetrating capability of the resulting peptide [D-KLA-R, D7GGPLGLAG(KLAKLAK)2R7 sequence] would be low without MMP2 owing to the masking effect of the cell penetrating sequence by the poly(aspartic acid) unit. MMPs are metal ion-dependent enzymes such as calcium and zinc to perform their biological functions.[34] MMPs can degrade the histological barrier of cancer cell invasion by degrading the extracellular matrix and basement membrane and play an important role in tumor growth, differentiation, angiogenesis, invasion, metastasis, and immune surveillance.[34,35] MMP2, one of the members of MMPs,[34] is related to malignant tumors and promotes the motility, proliferation, and metastasis of cancer cells.[34−37] Because MMP2 is abundantly secreted by various cancer cells, it is considered an effective target for developing cancer therapeutics.[38] Therefore, the cell penetrating capability of the D-KLA-R peptide could be selectively activated in the cancer niche with overexpressed MMP2. Furthermore, the activated peptide can efficiently penetrate cancer cells and exhibit selective cytotoxicity.

Figure 1

Schematic representation for selective cytotoxicity of the D-KLA-R peptide with cell penetrating motif activatable by MMP2.

Experimental Section

Preparation of D-KLA-R Peptide

D-KLA-R (D7GGPLGLAG(KLAKLAK)2R7) peptide was purchased from GL Biochem Ltd. The peptide purity was confirmed by high-performance liquid chromatography (HPLC) (Figure S3) and LC–mass spectrometry (Figure S4). D-KLA-R: m/z calculated for [M + 3H]3+ 1363.12, found 1363.12, m/z calculated for [M + 4H]4+ 1022.59, found 1022.60, m/z calcd for [M + 5H]5+, 818.28; found, 818.27.

Synthesis of D-KLA-R-FITC Peptide

The D-KLA-R peptide (3 mg, 0.73 μmol) was allowed to react with FITC (0.3 mg, 0.73 μmol) in DMF (200 μL) with DIPEA (1.6 μL, 9.5 μmol) overnight at RT. The crude peptide was purified by reverse-phase HPLC (YL9100, Yonung Lin Instruments, Korea) using a C18 column (Sunfire C18, 4.6 × 150 mm) as the stationary phase. Buffer A (water with 0.1% v/v TFA) and buffer B (acetonitrile with 0.1% v/v TFA) were used as the mobile phase. The gradient conditions of the mobile phase were as follows: 3 min at 100% Buffer A, followed by a linear gradient of 100–0% Buffer A over 25 min. After lyophilization of the collected fraction, a yellow powder of purified peptide was obtained. The successful synthesis of D-KLA-R-FITC was confirmed by HPLC and mass spectrometry (Figures S5 and S6). D-KLA-R-FITC: m/z calculated for [M + 4H]4+, 1119.85; found, 1119.85, m/z calcd for [M + TFA + 4H]4+, 1148.35; found, 1148.35, m/z calcd for [M + 5H]5+, 896.08; found, 896.08, m/z calcd for [M + TFA + 5H]5+, 918.88; found, 918.88.

Peptide Penetration Assay

H1299/shCont and H1299/shMMP2 cells were seeded at a density of 5 × 104 cells per well in 24-well plates and treated with 3 μM peptides (4:1 ratio mixture of D-KLA-R-FITC and D-KLA-R). After 1 h of incubation, cells were observed under a fluorescence microscope (Olympus Life Science, Tokyo, Japan).

Stable Cell Lines

To construct stable cell lines, cells were seeded at a density of 5 × 104 cells per well in 24-well plates and transfected with a 50 μL mixture containing 1 μg pshCont (control) or pshMMP2 (Qiagen) and TurboFect in vitro transfection reagent (Fermentas). Transfected cells were selected with 1 mg/mL G418 (Duchefa Biochemie) for one week and maintained in RPMI-1640 containing 0.5 mg/mL G418 during the experiments.

Tumor Xenografts in Nude Mice

All animal experiments were conducted in accordance with the protocols approved by the Institutional Animal Care and Use Committee of the Korea Institute of Radiological & Medical Sciences (IACUC approval no. KIRAMS2020-0009). Female athymic BALB/c nu/nu mice were purchased from Orient Bio Inc. (Seoul, Korea) and housed under specific pathogen-free conditions, supplied with standard rodent feed, and tap water ad libitum. A single cell suspension of H1299/shCont cells or H1299/shMMP2 cells (9 × 106 cells/100 μL HBSS) with a viability of 95% was subcutaneously injected into the hind legs of mice. Each group consisted of three mice, and the tumor volumes were determined according to the formula (L × l2)/2 by measuring the tumor length (L) and width (l) with a caliper. When the tumor reached a minimal volume of 150–200 mm3, intravenous injection of vehicle phosphate-buffered saline (PBS) or D-KLA-R peptide (3 mg/kg in 100 μL PBS) were administered to the mice every three days, and the tumor volume was measured daily. Experiments were terminated once the tumors in the vehicle-treated group reached 200 mm3 in volume.

Results and Discussion

Because the cytotoxicity of α-helical antimicrobial peptides including the KLA peptide is typically proportional to helicity of the peptides,[28,29] we investigated the helicity of the peptides in the aqueous phase. As shown in Figure , the circular dichroism (CD) spectrum of D-KLA-R exhibited weak negative peaks at 204 and 222 nm and a positive peak at 194 nm in PBS buffer (pH 7.4, 3 mM), which is the typical CD pattern of α-helical peptides. To quantitatively investigate the CD spectra, the percent helicity of the peptide was calculated from the mean residue ellipticity at 222 nm ([θ]222) of the CD spectrum (see Supporting Information for details).[39,40] The calculated percent helicity of the D-KLA-R peptide in PBS was approximately 18%. This result indicates that D-KLA-R adopted a weak helical structure in PBS buffer.

Figure 2

CD data of D-KLA-R peptide with and without SDS in PBS and PBS/TFE (1:1, v/v).

In general, the helicity of amphiphilic α-helical antimicrobial peptides including the KLA peptide in PBS buffer is relatively low compared to that on the surface of lipid membranes due to hydrogen bonding with water and non-specific interactions between hydrophobic side chains such as those in alanine and leucine residues.[41,42] To investigate the ressonable conformation of D-KLA-R peptide similar to that on the cell membranes, trifluoroethanol (TFE) or sodium dodecyl sulfate (SDS) was added to the PBS buffer. The addition of TFE generally enhances the helical structure of amphiphilic peptides by inducing higher intramolecular hydrogen bonding and fewer non-specific hydrophobic interaction.[42−44] The addition of SDS generally provides an environment similar to that of the negatively charged surface of the cellular membranes. After adding SDS or TFE to PBS buffer, the intensities of the CD spectra of D-KLA-R were enhanced as shown in Figure . The calculated percent helicities of D-KLA-R with SDS and TFE were about 34 and 44%, respectively. The helicity of D-KLA-R is similar to that of the original KLA peptide without any modifications.[21] These results indicate that the α-helical conformation of D-KLA-R on the cellular membranes is sufficient to induce cell death after intracellular uptake.

We investigated whether MMP2 induced cell death in the MMP2-expressing cancer cells by inducing structural changes in the D-KLA-R peptide. A previous study demonstrated that A549 and H1299 cells express and secrete low and high levels of MMP2, respectively.[45] Consistent with the previous study using gelatin zymography, we found that the expression and secretion of MMP2 increased in H1299 cells but not in A549 cells (Figure a). Whether D-KLA-R peptide induces cancer cell death via MMP2 was investigated using the ATP-Glo cell viability assay in A549 and H1299 cells. D-KLA-R peptide induced cancer cell death in H1299 cells expressing high levels of MMP2 in a concentration-dependent manner but did not induce cancer cell death toward A549 cells expressing low levels of MMP2 (Figure b). The viability of A549 cells was slightly reduced upon treatment with 5 μM D-KLA-R peptide. To confirm that this peptide induces apoptosis in H1299 cells in an MMP2-dependent manner, we transfected H1299 cells with siCont or siMMP2 (Figure c), treated with 5 μM D-KLA-R or left untreated, and analyzed cancer cell death using the ATP-Glo cell viability assay. Inhibiting the MMP2 expression can reduce the proliferation, invasion, and metastasis of cancer cells and promote their apoptosis.[36,37] Consistent with this, siMMP2 slightly reduced the viability of H1299 cells (Figure d). However, suppressing MMP2 expression effectively inhibited D-KLA-R-induced cancer cell death in H1299 cells (Figure d). Tung et al. reported that the D-form KLA peptide conjugated with R7 unit (kla-r7) having similar structure to KLA-R peptide after MMP2-induced cleavage of D-KLA-R exhibited IC50 value of 3.17 μM in LL/2 (LLC1) lung carcinoma,[46] which is comparable with the cytotoxicity of D-KLA-R peptide toward H1299 cells with MMP2 expression. These results suggest that D-KLA-R selectively induces cell death in MMP2-expressing cancer cells.

Figure 3

Analysis of D-KLA-R-induced cell death in MMP2-expressing cancer cells. (A) Gelatin zymography for MMP2 activity in A549 and H1299 cells. (B) D-KLA-R (0–5 μM)-induced cell death in A549 and H1299 cells. (C) Gelatin zymography for MMP2 activity in H1299 cells transfected with siCont or siMMP2. (D) Effect of MMP2 on the D-KLA-R-induced cell death in H1299 cells. siCont- or siMMP2-transfected H1299 cells were treated with 5 μM D-KLA-R or left untreated. After 24 h of treatment, the cell viability was analyzed using the ATP-Glo cell viability assay.

To this end, it is necessary to elucidate the mechanism by which the D-KLA-R peptide induces MMP2-expressing cancer cell death. H1299 cells were transfected with siCont or siMMP2 (Figure c) and treated with 5 μM D-KLA-R or left untreated. Then, annexin V/propidium iodide (PI) staining was performed to determine whether D-KLA-R induces apoptotic or necrotic cell death. In H1299 cells, treatment with D-KLA-R increased the number of annexin V/PI-positive cells in siCont-transfected cells, and siMMP2 pretreatment effectively inhibited D-KLA-R-induced cancer cell death (Figure ).

Figure 4

Effect of D-KLA-R on necrotic cell death in H1299 cells. siCont- or siMMP2- transfected H1299 cells were treated with D-KLA-R or left untreated and stained with annexin V/PI. Cancer cell death was measured using flow cytometry. (a) Lower-left quadrant represents the viable cells (PI– and annexin V–); upper-left quadrant represents the necrotic cells (PI+ and annexin V–); lower-right quadrant represents the apoptotic cells (PI– and annexin V+); upper-right quadrant represents the apoptotic and necrotic cells (PI+ and annexin V+). (b) Percentage of live, early apoptotic, necrotic, and apoptotic/necrotic cells treated and untreated with D-KLA-R are summarized.

Next, double staining with Hoechst33342 and PI was performed to determine whether D-KLA-R induced apoptosis or necrotic cell death. Hoechst33342 stains the nuclei of apoptotic cells while weakly staining the nuclei of non-apoptotic cells.[47,48] In addition, PI stains the nuclei of necrotic cells, but it cannot cross the cell membrane, and therefore, cannot stain the nuclei of non-necrotic cells.[47,48] We found that D-KLA-R strongly stained the nuclei of siCont-transfected H1299 cells with PI, but not in siMMP2-transfected cells (Figure ). We also observed that Hoeshcst33342-stained cells were found only in siCont-transfected H1299 cells treated with D-KLA-R, but the numbers were very few (Figure ). Therefore, these results suggest that the D-KLA-R peptide induces cancer cell death via necrosis.

Figure 5

Effect of D-KLA-R on necrotic cell death in H1299 cells assessed by Hoechst 33342 and PI double staining.

To confirm whether D-KLA-R-induced cancer cell death was dependent on apoptosis, H1299 cells were treated with D-KLA-R, and immunoblot analysis was performed to observe cleaved PARP, a marker of apoptosis. Doxorubicin (DOX) (3 μM) was used as a positive control. DOX activated PARP cleavage, whereas D-KLA-R did not (Figure S1). These results suggest that D-KLA-R induces necrotic cell death.

The increase in necrotic cell death of cancer cells by D-KLA-R is only possible when D-KLA-R is rapidly introduced into MMP2-expressing cells compared to non-MMP2-expressing cells. To test this, we constructed a stable cell line (Figure S2) and treated the cells with 3 μM peptides (FITC-conjugated D-KLA-R (D-KLA-R-FITC) and D-KLA-R in a 4:1 ratio). As shown in Figure , D-KLA-R was rapidly introduced into H1299/shCont cells at 1 h but not into H1299/shMMP2 cells. At concentrations of D-KLA-R higher than 3 μM, no influx of D-KLA-R into cells was observed due to the rapid necrotic cell death of cancer cells.

Figure 6

Penetration of FITC-labeled D-KLA-R (D-KLA-R-FITC) into MMP2-expressing (H1299/shCont) and non-expressing (H1299/shMMP2) cancer cells. Scale bar = 50 μm.

A previous report demonstrated that the cationic (KLAKLAK)2 peptide disrupts cell membranes, reducing the potential of the mitochondrial and plasma membranes.[27] Consequently, cationic KLA peptide-induced cell death is independent of caspase-dependent apoptosis.[27,30] To investigate whether D-KLA-R induces cell death by reducing the potential of mitochondrial and plasma membranes, we transfected H1299 cells with siCont or siMMP2 and treated them with 5 μM D-KLA-R or left them untreated. As shown in Figure , D-KLA-R reduced plasma and mitochondrial membrane potentials, and siMMP2 effectively inhibited these D-KLA-R-induced loss of membrane potentials in H1299 cells. Collectively, the chemical structure of the peptide was changed by MMP2 expressed in cancer cells, and the changed peptide induced cancer cell death through the destruction of plasma and mitochondrial membranes.

Figure 7

Effect of D-KLA-R on reducing mitochondrial (a) and plasma membrane potentials (b) in H1299 cells.

To verify the in vitro results, we performed in vivo studies using xenograft model. Mouse xenograft data revealed that when the tumor reached a minimal volume of 200 mm3, the xenograft H1299/shCont tumors treated twice with the D-KLA-R peptide showed a decrease in the tumor volume (Figure a). Two of the three mice administered with D-KLA-R peptide showed dramatically inhibited the growth of H1299/shCont tumor cells; no tumor mass was observed at the end of the experiment (Figure a). However, the application of D-KLA-R peptide to xenograft H1299/shMMP2 tumors did not induce any tumor growth inhibition compared to the vehicle-treated group. The peptide injection into mice showed no difference in the body weights of all groups (Figure b). To investigate whether D-KLA-R caused necrosis-induced cancer cell death in H1299/shCont tumors, we performed immunohistochemistry and H&E staining. As shown in Figure , D-KLA induced necrosis in H1299/shCont tumors, but not in H1299/shMMP2 tumors.

Figure 8

(a) D-KLA-R inhibits the growth of tumor xenografts using cancer cells expressing MMP2 (n = 3 mice/group). (b) Body weights of tumor-bearing mice in each group were measured after sacrificing the animals (n = 3 mice/group).

Figure 9

Immunohistochemical analysis (a) and H&E staining (b) of H1299/shCont and H1299/shMMP2 xenograft tumors with or without D-KLA-R injection. The sections were stained for anti-MMP2 antibody using 3,3′-DAB. Scale bar = 200 μm. Necrotic cells (N); viable cells (V).

Conclusion

In conclusion, we have developed an antimicrobial KLA peptide (D-KLA-R) with an activatable cell penetrating motif for selective cytotoxicity triggered by MMP2. The D-KLA-R peptide exhibited an α-helical structure sufficient to induce cytotoxicity. Upon treatment of H1299 cells (high MMP2 levels) with D-KLA-R, the peptide was selectively internalized into the cells by MMP2, consequently inducing cell death. In contrast, D-KLA-R showed negligible cytotoxicity toward A549 cells (low MMP2 levels) and siMMP2-transfected H1299 cells. The D-KLA-R peptide induced necrotic cell death in H1299 cells by disrupting plasma and mitochondrial membranes. Furthermore, the selective therapeutic efficacy of the peptide induced by MMP2 was also corroborated in vivo. Therefore, the D-KLA-R peptide can be utilized to develop highly selective therapeutic agents.