Article Cite This: Bioconjugate Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/bc
Conjugate of Enkephalin and Temporin Peptides as a Novel Therapeutic Agent for Sepsis A. Golda,† P. Kosikowska-Adamus,‡ O. Babyak,† M. Lech,†,§ M. Wysocka,‡ A. Lesner,‡ J. Potempa,†,∥ and J. Koziel*,†
Bioconjugate Chem. Downloaded from pubs.acs.org by UNIV OF RHODE ISLAND on 12/11/18. For personal use only.
†
Department of Microbiology, Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian University, 30-387 Krakow, Poland ‡ Faculty of Chemistry, University of Gdansk, 80-309 Gdansk, Poland § Department of Nephrology, Medizinische Klinik und Poliklinik IV, Klinikum der Ludwig-Maximilians-Universität München, 80366 Munich, Germany ∥ Center of Oral Health and Systemic Disease, University of Louisville School of Dentistry, University of Louisville, Louisville, Kentucky 40202, United States S Supporting Information *
ABSTRACT: Antimicrobial peptides (AMPs) exhibit a wide spectrum of actions, ranging from a direct bactericidal effect to multifunctional activities as immune effector molecules. The aim of this study was to examine the anti-inflammatory properties of a DAL-PEG-DK5 conjugate composed of a lysine-rich derivative of amphibian temporin-1CEb (DK5) and dalargin (DAL), the synthetic Leuenkephalin analogue. Detailed study of the endotoxin-neutralizing activity of the peptide revealed that DAL-PEG-DK5 interacts with LPS and the LPS binding protein (LBP). Moreover, DAL-PEG-DK5 prevented dimerization of TLR4 at the macrophage surface upon LPS stimulation. This inhibited activation of the NF-κB signaling pathway and markedly reduced pro-inflammatory cytokine production. Finally, we showed that aggregation of DAL-PEG-DK5 into amyloid-like structures induced by LPS neutralized the endotoxin proinflammatory activity. Consequently, DAL-PEG-DK5 reduced morbidity and mortality in vivo, in a mouse model of endotoxin-induced septic shock. Collectively, the data suggest that DAL-PEG-DK5 is a promising therapeutic compound for sepsis.
■
INTRODUCTION
results in expression of cytokines and other inflammatory mediators.8 Host defense peptides, also called antimicrobial peptides (AMPs), are classified as innate immune effectors in a variety of species, including plants, insects, amphibians, and mammals.9,10 The Antimicrobial Peptide Database-APD (http://aps.unmc. edu/AP/main.php) lists various AMP families, including defensins, dermaseptins, cathelicidins, and temporins.11 AMPs act as potent antimicrobial molecules against a broad spectrum of microorganisms. In addition their broad anti-inflammatory activity is very well documented.10,12,13 The best investigated is the ability of AMPs to inhibit pro-inflammatory reactions in host tissues triggered by bacterial endotoxins. This activity depends on effective, strong binding of LPS, which prevents the LPS-LBP interaction and subsequent initiation of the TLR4 signaling.14,15 Additionally, AMPs are potent molecules that block organ damage in murine models of septic shock.16,17 Taken together, it is clear that short cationic peptides could be applied as a new
Monocytes and macrophages are part of the host innate immune system and play an important role in establishing effective responses against bacterial, viral, and fungal infections.1,2 Recognition of pathogen-associated molecular patterns (PAMPs) by phagocytes triggers an inflammatory response characterized by secretion of various pro-inflammatory mediators, including cytokines (e.g., TNF-α, IL-1β, IL-6), chemokines (e.g., IL-8), interferons, prostaglandins, and antimicrobial proteins.3 Lipopolysaccharide (LPS), also known as endotoxin, is the most common PAMP. LPS covers more than 70% of the Gram-negative bacterial cell surface.4 It constitutes a physical barrier that protects microorganisms from unfavorable environment. Endotoxin released during bacterial growth or disintegration plays an important role in the development of Gramnegative infections and sepsis.5,6 LPS in serum is recognized by the LPS binding protein (LBP) and then transferred to CD14, a differentiation antigen expressed on the surface of phagocytes. CD14 initiates formation of the TLR4/MD2 (toll-like receptor 4/myeloid differentiation protein 2) complex.7 This leads to activation of multiple signaling pathways, including NF-κB and © XXXX American Chemical Society
Received: October 22, 2018 Revised: November 25, 2018
A
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry generation of antibiotics, as well as potent immunomodulators.18 DAL-PEG-DK5 is a peptide conjugate synthesized from a lysine-rich derivative of amphibian temporin-1CEb (DK5) and dalargin (DAL), the synthetic Leu-enkephalin analogue. Both compounds are linked by a short biologically inert and nonimmunogenic PEG linker (8-amino-3,6-dioxaoctanoic acid).19 Temporin-1CEb is a naturally occurring 12-residue peptide derived from frog skin secretions that display antimicrobial activity against Gram-positive and (to a lesser extent) Gram-negative bacteria.20,21 Moreover, a recent study reported its anti-inflammatory properties.22 DAL exerts cytoprotective and pro-proliferative effects that facilitate wound healing.23,24 DAL-PEG-DK5 has a cationic charge of +7 and forms an α-helical structure in the presence of detergents. Its antibacterial efficacy was demonstrated against both Gramnegative and Gram-positive pathogens;19 however, its effect on innate immune responses had not yet been studied. In this study we examined the immunomodulatory properties of two conjugates, DAL-PEG-DK5 and DK5-PEG-DAL, and compared they activity with that exerted by the nonconjugated peptides (DAL and DK5). Our data indicate that DAL-PEGDK5 prevented development of sepsis by acting, both in vitro and in vivo, as a potent anti-endotoxin agent. The endotoxinneutralizing mechanism relied on LPS-induced DAL-PEG-DK5 aggregation into amyloid-like structures, which blocked recognition of endotoxin by host cells. This led to inhibition of intracellular signaling and, ultimately, reduced expression of pro-inflammatory cytokines.
■
RESULTS Anti-Inflammatory Activity of Mimetic Peptides and Their Conjugates. The innate immune response is rapidly activated upon recognition of pathogen-derived, conserved molecules, including among others lipid A (part of LPS), peptidoglycans (PGN), and lipoteichoic acid (LTA). To evaluate the protective effects of four peptides of interest against TLR agonists, we used a murine macrophage (RAW264.7) model. Cells were treated with selected PAMPs in the presence or absence of AMPs and stimulation of expression of TNF-α and NO was measured. All investigated peptides, except DAL, abrogated LPS-induced secretion of TNF-α and NO (Figure 1A and B). By contrast, only DAL-PEG-DK5 interfered with inflammatory responses induced by LTA and PGN (Figure 1A,B). Further studies using primary human monocyte-derived macrophages (hMDMs) confirmed the observed phenomenon (SI Figure 1). To exclude the possibility that observed results were due to peptide-mediated toxicity we determined whether the peptides can affect cell viability. As shown in SI Figure 2, within applied doses, the peptides exerted no cytotoxic effects. These results argue that each peptide that contained the DK5 motif inhibited activation of phagocytes in the response to bacterial components, although with differing efficacy. Moreover, our data indicate that DAL-PEG-DK5 was the most efficient compound against a broad range of TLR agonists. Therefore, we focused on the role of this peptide in cell-signaling pathways induced by LPS. Effect of Mimetic Peptides on LPS-Induced Activation of Macrophages. Detailed examination of the anti-endotoxin effects of the investigated peptides was performed using both human and murine macrophages. Primary hMDMs and RAW264.7 cells were stimulated with E. coli-derived LPS
Figure 1. Ability of the investigated peptides to inhibit inflammatory reactions induced by TLR agonists in RAW264.7 cells. Mouse macrophages were stimulated with 10 ng/mL LPS, 10 μg/mL LTA, or 2 μg/mL PGN in the presence of 25 μg/mL peptide. The level of (A) TNF-α and (B) NO in the culture supernatant was determined in an ELISA or Griess assay, respectively, at 20 h post-stimulation. Mean ± SD n = 3. *P < 0.05, ***P < 0.001, ****P < 0.0001; one-way ANOVA.
together with each peptide and amounts of cytokines secreted into cell supernatants were measured. As shown in Figure 2A, the strongest anti-endotoxin activity in the hMDMs model was exerted by DAL-PEG-DK5, as manifested by a significant reduction of secreted TNF-α, IL-6, IL-8, and IL-10. In the case of RAW264.7 cells, we observed equally efficient inhibition of LPS regardless of the test compound that contained the DK5 sequence (data not shown). To verify the specific anti-LPS activity of DAL-PEG-DK5, we used a scrambled peptide (SCR), which showed no effect on the cell response (Figure 2A). We also examined the anti-endotoxin activity of peptides in whole human blood by evaluating TNF secretion. A protective effect was documented for two conjugates: DK5-PEG-DAL and DALPEG-DK5 (Figure 2B). The latter showed the greatest antiinflammatory efficacy. These results confirmed the findings using hMDMs as the model (Figure 2A) and revealed that both conjugates neutralized LPS proinflammatory activity in an environment that resembles physiological conditions. Finally, we documented that DAL-PEG-DK5 inhibited the response of macrophages to E. coli infection (Figure 2C). Taken together, the data clearly demonstrate that DAL-PEG-DK5 is the potent anti-endotoxin agent. Comparison of the Anti-Endotoxin Activity of the Peptides. To quantify the anti-inflammatory efficiency of the B
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Figure 2. Anti-endotoxin activity of peptides. (A) Human MDMs were stimulated with 10 ng/mL LPS in the presence of each peptide (10 μg/mL). The level of TNF-α, IL-6, IL-8, and IL-10 in the culture supernatants was measured by ELISA at 24 h post-stimulation. Mean ± SD n = 3. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; one-way ANOVA. (B) Human blood was treated with 10 ng/mL LPS in the presence of each peptide (10 μg/mL) and TNF-α concentrations in plasma were measured 20 h later. Mean ± SD n = 2. *P < 0.05, **P < 0.009; one-way ANOVA. (C) HMDMs were infected with E. coli (MOI 1:50) and the TNF-α concentration in the culture supernatant was measured in an ELISA. Mean ± SD n = 2. *P < 0.05, ***P < 0.001; one-way ANOVA. SCR, scramble peptide.
test compounds, we measured the inhibitory concentration 50% (IC50) of LPS-induced macrophage stimulation. Human MDMs and RAW264.7 cells were incubated either with LPS alone or with LPS plus different concentrations of each peptide. We observed dose-dependent inhibition of the endotoxin-mediated inflammatory reaction with the strongest effect exerted by DALPEG-DK5 (Figure 3A,B; Table 1). The IC50 values for TNF-α measured in both human (2.015 μg/mL) and murine macrophages (1.182 μg/mL) were comparable. The IC50 of inhibition of NO secretion in murine cells was on the same level. (3.098 μg/mL).
Conversely, in comparison to DAL-PEG-DK5, DK5 and DK5-PEG-DAL were far less effective in inhibition of TNF-α and NO release, whereas DAL and SCR did not show any inhibitory effect (Figure 3A,B; Table 1). Collectively, these results clearly demonstrate that DAL-PEG-DK5 is nine times more potent than DK5 in terms of inhibiting endotoxinmediated hMDMs stimulation and six times more potent than DK5 in terms of inhibiting stimulation of RAW264.7 cells. Moreover, DAL-PEG-DK5 is at least eight times more potent than DK5-PEG-DAL in terms of inhibiting LPS-mediated TNFα by hMDMs (Table 1). Together, we showed that the C
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
using fluorescently labeled LPS (LPS-AF488). Flow cytometry analysis revealed that binding of LPS to the macrophage membrane was blocked efficiently by DAL-PEG-DK5, since only 14% of cells were LPS-positive. By contrast, 74% of control cells not exposed to the peptide bound fluorescently labeled endotoxin (Figure 4C). SCR had no effect on LPS binding. Moreover, the results of a LBP block assay revealed limited interaction between LPS and LBP in the presence of DAL-PEGDK5 (in contrast to SCR) (Figure 4D). Collectively, these results indicate that DAL-PEG-DK5 inhibited both binding of endotoxin to cell surface receptors and its interaction with the main LPS-cargo molecule in serum. DAL-PEG-DK5 Prevents LPS-Induced Dimerization of TLR4/MD2 and Inhibits LPS-Induced NF-κB Signaling. LPS binding to the cell membrane induces dimerization of the TLR4/MD2 complex because LPS creates an additional binding interface between TLR4 and MD2.25,26 Therefore, we used specific anti-TLR4/MD2 antibodies and FACS analysis to verify whether DAL-PEG-DK5 is able to inhibit LPS-mediated TLR4/ MD2 dimerization.27 The antibodies are specific only for monomeric TLR-4/MD2 and therefore do not recognize dimers. RAW264.7 cells expressed TLR4 receptors (Figure 5A; green line). Dimerization occurred upon binding of LPS to the TLR4/MD2 complex, as demonstrated by a decrease in fluorescence intensity (Figure 5A; pink line). We observed the reverse effect in the presence of DAL-PEG-DK5, as illustrated by restoration of fluorescence intensity (Figure 5A; blue line). This indicates that DAL-PEG-DK5 inhibits dimerization of TLR-4/ MD2 receptors. By contrast, SCR (Figure 5A; orange line) did not affect TLR4 dimerization induced by LPS binding. LPSinduced dimerization of the TLR4/MD2 complex leads to activation of multiple signaling components, including NFκB.3,25 Therefore, we employed a specific reporter cell line to study the effect of DAL-PEG-DK5 on LPS-induced NF-κB activation. The RAW264.7 NF-κB sensor cell line was stimulated with LPS in the presence or absence of DAL-PEGDK5 or SCR. As shown in Figure 5B, NF-κB activation in LPSstimulated RAW264.7 cells was inhibited by DAL-PEG-DK5. By contrast, SCR did not show any inhibitory effect. Since the activity of NF-κB regulates expression of defined inflammatory genes, we measured the levels of TNF-α transcript. Our data demonstrated that expression of mRNA encoding TNF-α was abolished when LPS-stimulated hMDMs were exposed to DALPEG-DK5 (Figure 5C). Cumulatively, these data show that DAL-PEG-DK5 exerted a strong anti-inflammatory effect by blocking recognition of endotoxin by host cells and consequent activation of intracellular signaling. Direct Interaction between DAL-PEG-DK5 and LPS. To further explore the mechanism underlying DAL-PEG-DK5mediated inhibition of endotoxin activity, we examined the direct interaction between both molecules. Because DAL-PEGDK5 has a cationic charge (+7), we anticipated that it would interact with negatively charged molecules such as the anionic lipid A domain of endotoxin. When a FITC-labeled peptide conjugate was incubated with E. coli LPS in sodium phosphate buffer, we observed dose-dependent increases in fluorescence intensity, reflecting aggregation of FITC-labeled DAL-PEGDK5 and LPS (Figure 6A). Moreover, we did not detect significant aggregation of the peptide alone for up to 2 h, which contrasted with the observation of LPS-induced aggregation occurring already after 1 min (Figure 6B). Likewise, a significant increase in ThT (thioflavin T) binding was observed after
Figure 3. Peptides show a marked difference in their ability to inhibit endotoxin activity. HMDMs (A) or RAW264.7 cells (B) were stimulated with 10 ng/mL LPS plus different concentrations of peptide (2.5−50 μg/mL). TNF-α (A) or NO (B) concentrations in the culture supernatants at 20 h post-stimulation were measured in an ELISA or Griess assay, respectively. Because none of the peptides alone induced release of TNF-α or NO, these controls are not shown in the figure. Data are expressed as the mean ± SD of two independent experiments.
Table 1. Inhibition (IC50) of LPS-Induced Macrophage Stimulationa IC50 [μg/mL]
hMDMs
RAW264.7
DK5 DAL DK5-PEG-DAL DAL-PEG-DK5 SCR
18.68 nd 15.58 2.015 nd
17.70 nd 19.83 3.098 nd
a
nd, not determined.
examined peptides possess the broad anti-inflammatory activity regardless of the used in vitro model. Since the LPS-neutralizing activity by DAL-PEG-DK5 was the most significant, we examined this compound in more detail. DAL-PEG-DK5 Prevents LPS Binding to the Cell Surface. The ability of DAL-PEG-DK5 to suppress inflammatory responses elicited by LPS suggests that the peptide affects LPS recognition by phagocytes. Therefore, we stimulated hMDMs and RAW264.7 cells with LPS in the presence of DAL-PEG-DK5. We pretreated cells with LPS, followed by stimulation with DAL-PEG-DK5. Alternatively, we pretreated cells with DAL-PEG-DK5 and then stimulated them with LPS. Interestingly, TNF-α and NO secretion were inhibited markedly by DAL-PEG-DK5 administered together with LPS, or when cells were preincubated with the peptide (Figure 4A,B). However, when macrophages were exposed to LPS followed by DAL-PEG-DK5, the anti-inflammatory effect of the peptide was abolished (Figure 4A,B). The results clearly suggest that DAL-PEG-DK5 may inhibit the interaction between endotoxin and the cell membrane. Therefore, we examined the LPS−cell membrane interaction D
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Figure 4. DAL-PEG-DK5 prevents LPS binding to the cell surface. (A) HMDMs and (B) mouse RAW264.7 macrophages were stimulated with 10 ng/ mL LPS in the presence of DAL-PEG-DK5 (LPS + DAL-PEG-DK5) at the indicated concentrations (10 μg/mL for hMDMs or 25 μg/mL for RAW264.7 cells). Alternatively, macrophages were pretreated with LPS for 2 h, followed by stimulation with DAL-PEG-DK5 (LPS/DAL-PEG-DK5), or macrophages were stimulated with DAL-PEG-DK5 for 2 h, and then stimulated with LPS (DAL-PEG-DK5/LPS). The TNF-α and NO concentrations in the culture supernatants were determined in an ELISA or Griess assay, respectively, at 20 h post-stimulation. Mean ± SD n = 2. *P < 0.05, ***P < 0.001, ****P < 0.0001; one-way ANOVA. (C) The peptide blocked binding of AF488-conjugated LPS to RAW264.7 cells. RAW264.7 cells (5 × 105 cells/mL) were incubated for 15 min at 37 °C with 100 ng/mL AF488-conjugated LPS (AF488-LPS) in the absence or presence of peptide (25 μg/mL) in DMEM containing 5% FBS. After washing, binding of AF488-LPS was analyzed by flow cytometry. Background fluorescence was determined using RAW264.7 cells incubated in the absence of AF488-LPS. A representative histogram from one of three independent experiments is shown, along with the mean percentage of macrophages that bound LPS. Mean ± SD n = 3. ****p < 0.0001; one-way ANOVA. (D) LBP block assay. Peptides (50 μg/mL) were incubated with LPS (5 μg/mL) and the interaction with LBP (60 ng/mL) was investigated. Polymixin B (100 μg/mL) was used as a positive control. Mean ± SD n = 3. ****p < 0.0001; one-way ANOVA. SCR, scramble peptide. E
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Figure 5. Effect of DAL-PEG-DK5 and SCR on TLR4 dimerization and the NF-κB pathway. (A) Effect of DAL-PEG-DK5 on TLR4 dimerization. Flow cytometry analysis of TLR4 dimerization using an antibody specific for TLR4/MD2-AF488 (MTS510); the antibody recognizes only monomeric TLR4/MD2 complexes. RAW264.7 cells (1 × 106 cells) were incubated for 30 min at 37 °C with 100 ng/mL LPS in the absence or presence of DAL-PEG-DK5 or SCR (25 μg/mL) in DMEM containing 5% FBS. Histograms show the isotype control (filled area), along with the AF488-labeled anti-TLR4/MD2 (green), LPS (pink), DAL-PEG-DK5 (blue), and SCR (orange). One representative experiment of three is shown (n = 3). (B) Inhibition of NF-κB activation by DAL-PEG-DK5. A stable RAW264.7 NF-κB sensor cell line was used to study the signal transduction pathway. Mouse macrophages were stimulated with 10 ng/mL LPS in the presence of DAL-PEG-DK5 (25 μg/mL) or SCR (25 μg/mL), and NF-κB activity was measured 6 h later. Mean ± SD n = 3. ****p < 0.0001; one-way ANOVA. (C) HMDMs were incubated with medium only or with LPS (10 ng/mL) in the absence or presence of DAL-PEG-DK5 (10 μg/mL). After the indicated time, total RNA was isolated and RT-qPCR was performed to estimate the level of TNF-α transcripts. SCR, scramble peptide. Mean ± SD. One representative experiment out of three is shown (n = 3). ****p < 0.0001; two-way ANOVA.
Examination of serum and peritoneal exudate revealed a significant increase in endotoxin levels in animals treated with LPS plus DAL-PEG-DK5 compared with animals injected with LPS alone (Figure 7B), suggesting that LPS is not bound to the cell membrane of leukocytes. In addition, WBC counts in LPStreated mice were significantly lower (1.8 × 105 cells/μL) than those in untreated control mice (6.1 × 105 cells/μL). Remarkably, WBC counts were restored in animals injected with endotoxin plus DAL-PEG-DK5 (Figure 7C). Further serum analysis revealed that animals injected with LPS showed significantly higher cytokine levels than control mice (the mean cytokine levels in control animals are marked in the figure by dotted gray lines). Analysis of serum cytokine levels in mice challenged with LPS plus DAL-PEG-DK5 showed a significant reduction in pro-inflammatory mediators such as IFN-γ and MCP-1. By contrast, we noted increased expression of IL-6 and IL-10; TNF-α levels remained unchanged (Figure 7D). These data provide evidence that the DAL-PEG-DK5 conjugate was a potent endotoxin-neutralizing agent that may prevent fatal complications associated with septic shock mediated by bacterial endotoxin. Stability of the Peptides. There are several obstacles to developing AMP-based therapies for infections. One of them is a
addition of LPS to DAL-PEG-DK5. Moreover, we did not detect significant aggregation of SCR after adding LPS (Figure 6C). Thus, we postulate that LPS induces aggregation of the peptide, which in turn leads to LPS neutralization. Taken together, the results show that DAL-PEG-DK5 was able to aggregate and form amyloid structures in the presence of LPS, thereby preventing endotoxin recognition and ensuing immune activity. DAL-PEG-DK5 Activity in Vivo. Finally, we used the wellestablished D-galactosamine-sensitized mouse model to investigate the ability of DAL-PEG-DK5 to prevent the lethal effects of endotoxin in vivo. When D-galactosamine-sensitized C57BL/6 mice were injected intraperitonealy with E. coli-derived LPS, 90% of the animals died within 26 h post-injection (Figure 7A). As our in vitro data revealed that TNF-α and NO secretion was markedly inhibited by the peptide administrated together with LPS or when cells were pretreated with DAL-PEG-DK5, we applied endotoxin and DAL-PEG-DK5 simultaneously. We observed, a marked improvement in survival after treatment with DAL-PEG-DK5. All D-galactosamine-sensitized C57BL/6 mice injected with E. coli-derived LPS and DAL-PEG-DK5 survived (Figure 7A). F
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Figure 6. Direct interaction between DAL-PEG-DK5 and LPS. (A) Dose-dependent FITC-labeled DAL-PEG-DK5 aggregation triggered by 25 μg/ mL LPS. (B) Aggregation assay with FITC-labeled DAL-PEG-DK5 (40 μg/mL) with or without LPS (25 μg/mL). (C) ThT assay identified a significant increase in amyloid formation by DAL-PEG-DK5 (40 μg/mL) after addition of 100 μg/mL LPS (n = 3). Mean ± SD n = 3. ***P < 0.001, ****P < 0.0001; two-way ANOVA. SCR, scramble peptide; au, arbitrary units.
anti-inflammatory activity of the tested conjugate and its derivatives, we used murine and human monocyte-derived macrophages. We noted a significant reduction in expression of major inflammatory mediators such as TNF-α, IL-6, IL-8, and NO when the peptide was added to LPS-treated phagocytes. The efficiency of LPS inhibition was 9- (in hMDMs) and 5.7times (in RAW264.7 cells) higher than that of the DK5 peptide alone, which has documented anti-endotoxin effects.21 DALPEG-DK5 has a higher net charge than DK5 (+7 vs +6) and more α-helicity (35% vs 29%), along with comparable hydrophobicity.19 Notably, the anti-endotoxin effect of the tested compound was stronger than that of the reverse conjugate (DK5-PEG-DAL), indicating that the sequence of peptides plays a role in the observed phenomenon and suggesting that an anti-inflammatory peptide is more likely to be effective when localized to the C-terminus of the conjugate. A previous study showed that another temporin-1CEb analogue, LK6, was able to inhibit IL-8 and TNF-α production by THP-1 human monocytic cells. The anti-inflammatory activity of LK6 was associated with the binding of the peptide to LPS and subsequent inhibition of LPS-induced pro-inflammatory responses by monocytes.22 Considering the fact that oxidants affect all stages of the inflammatory response,29 it is worth mentioning that inclusion of DAL as a therapy for patients with coronary heart disease led to a significant reduction in oxidative stress parameters.30 Further studies of the mechanism underlying the ability of DAL-PEG-DK5 to neutralize LPS revealed that the peptide prevents interaction between LPS and soluble and membrane-anchored molecules that are crucial for activation of macrophages; examples include LBP, CD14, and TLR-4. This prevents activation of intracellular signaling pathways such as the NF-κB pathway, thereby blocking production of pro-inflammatory cytokines. Taken together,
reduction in peptide activity after systemic administration/ distribution. Peptides may bind to serum proteins and aggregate, can be sensitive to salt, serum, or pH changes, or may be easily degraded by endogenous proteases.18 The data presented above clearly indicate that the tested peptides remained active in the presence of serum as well as in vivo; nevertheless, we examined their stability in human serum and cell culture medium (5% FBS in DMEM). The data are presented in Table 2. All peptides had half-lives >24 h in culture medium. In human serum, two conjugates (DK5-PEG-DAL and DAL-PEG-DK5) had half-lives of 4 h, with DK5-PEG-DAL being slightly more stable than DAL-PEG-DK5. The native peptides were degraded faster: DAL had a half-life of 3 h and DK5 had a half-life of 1 h. Collectively, these data show that the half-life of DAL-PEG-DK5 was 3.6 times longer than that of DK5.
■
DISCUSSION AMPs, besides playing a role as potent bactericidal molecules, effectively modulate host responses during infection. Among immunomodulatory functions of AMPs, neutralization of bacterial endotoxin is an important feature that determines development of infection. LPS initiates a strong immune response and serves as an early signal of bacterial infection;25 however, it can also trigger uncontrolled host reactions, followed by tissue damage and organ dysfunction (as observed in sepsis).28 In this study, we demonstrate for the first time the potent endotoxin-neutralizing ability of a conjugate of enkephalin and temporin analogues (DAL-PEG-DK5). The conjugate was selected based on the results of previous studies in which we attempted to design bifunctional peptide-based compounds that share/maintain the antimicrobial functions of DK5 and the cytoprotective activity of DAL.19 To examine the G
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Figure 7. Ability of DAL-PEG-DK5 to prevent mortality and morbidity associated with septic shock. D-Galactosamine (18 mg/mice)-sensitized C57BL/6 mice were injected intraperitoneally with LPS (0.1 μg/g) without or with DAL-PEG-DK5 (25 μg/g). (A) Mortality is expressed as a percentage. Survival of mice injected with LPS plus DAL-PEG-DK5 or with LPS alone. Survival statistics were calculated using the Mantel-Cox test. ***p < 0.0002 (ten mice per test group). (B) Levels of endotoxin in mouse serum and peritoneal cavity. Data are expressed as the mean ± SD **p < 0.01, ***p < 0.001; one-way ANOVA. (C) White blood cells (WBCs) in mouse blood. Data are expressed as the mean ± SD. **p < 0.003; one-way ANOVA. (D) Serum levels of TNF-α, IFN-γ, MCP-1, IL-10, and IL-6 were measured using a cytokine bead array system. Values were compared between animals injected with LPS or LPS plus DAL-PEG-DK5. The levels of each individual cytokine in control mice are illustrated by dotted lines. Data are expressed as the mean ± SEM (five mice per group). **p < 0.0012 (IFN-γ), ***p < 0.0009 (MCP-1), **p < 0.0020 (IL-10), ***p < 0.0002 (IL-6); unpaired t test.
these results indicate that DAL-PEG-DK5 interferes with LPS binding to the macrophage cell surface. The net cationicity and amphipathicity of AMPs mediate their interaction with highly anionic LPS. Therefore, we examined the
interaction between DAL-PEG-DK5 and LPS. We found marked dose-dependent aggregation of DAL-PEG-DK5 after addition of LPS (Figure 6A,B). At the same time, we detected no aggregation of the pure peptide. Lee and co-workers observed H
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Interestingly, we found higher endotoxin levels in serum and the peritoneum of mice injected with LPS plus DAL-PEG-DK5 than in animals treated with LPS alone (Figure 7B). A likely explanation is that the peptide blocks binding of endotoxin to cellular membranes. We also observed lower levels of proinflammatory mediators in serum (IFN-γ and MCP-1) at 3 h post-challenge with LPS plus the peptide than after challenge with LPS alone. By contrast, levels of circulating IL-6 and IL-10 were higher in the presence of peptide (Figure 7D), confirming that the peptide protects against LPS (both of these cytokines play an anti-inflammatory role in local and systemic acute inflammatory responses by controlling the levels of proinflammatory cytokines).39 Indeed, IL-6 plays an important anti-inflammatory role in pneumococcal meningitis.40 Taken together, the data presented in this study show that DAL-PEGDK5 controls inflammatory reactions and protects mice against the effects of endotoxin. To summarize, we show here that DAL-PEG-DK5 exhibits strong endotoxin-neutralizing activity both in vitro and in vivo. Moreover, the data suggest that LPS induces amyloid-DALPEG-DK5 aggregation, thereby neutralizing its own activity. Taken together, the data show that the ability of DAL-PEG-DK5 to modulate inflammatory responses elicited by LPS makes it a potential treatment for sepsis. Moreover, DAL-PEG-DK5 is a promising therapeutic compound because it inhibits not only LPS, but also PGN and LTA; therefore, its inhibitory activity is not restricted to a specific TLR.
Table 2. Peptide Stability in Human Serum Obtained from Healthy Donors and in Cell Culture Medium (5% FBS DMEM) compound
half-life in human serum [min]
stability in culture medium [h]
DK5 DAL DK5-PEG-DAL DAL-PEG-DK5
67 ± 12 182 ± 21 271 ± 32 243 ± 18
>24 >24 >24 >24
similar effects when studying the interaction between a CMA3 peptide and endotoxin.17 Moreover, several AMPs form amyloid structures upon interaction with LPS; one of these is Halictines2, which in the presence of LPS adopts a β-strand structure and forms amyloids.31 Amyloid refers to abnormal fibrous, extracellular, proteinaceous deposits found in organs and tissues. Amyloid fibrils composed of cross-β-sheet structures are associated with Alzheimer’s disease, Parkinson’s disease, and prion diseases. In marked contrast to disease-associated amyloids, other amyloids have native biological activity. For example, E. coli form extracellular amyloid fibrils called curli, which are involved in surface and cell−cell contact and promote community behavior and host colonization.32,33 For over 50 years, the fluorescent dye ThT has become one of the most widely used dyes for selectively monitoring amyloid formation both in vivo and in vitro.34 Therefore, we monitored formation of amyloid fibrils using a ThT binding assay and observed a significant increase in ThT staining when endotoxin was added to DAL-PEG-DK5; this suggests that the peptide forms aggregates that are, at least in part, organized into amyloid-like structures (Figure 6C). Notably, because aggregation occurred after LPS was added, and activity depended on the peptide concentration, we propose a nucleation-dependent aggregation process. However, we cannot exclude the possibility that LPS aggregates in the presence of the peptide. This conclusion is supported by biological observations showing a lack of endotoxin binding to the CD14 receptor, presumably due to the fact that the binding pocket of CD14 is too small to accommodate large LPS aggregates; therefore, it is likely to bind the monomeric form of LPS.25 Further detailed studies examining the interaction between DAL-PEG-DK5 and LPS and other components of the bacterial cell membrane (such as LTA or PGN) are needed. Physicochemical methods, including circular dichroism, nuclear magnetic resonance, surface plasmon resonance, or thermophoresis are required to better understand the interaction between DAL-PEG-DK5 and LPS. Although the anti-endotoxin mechanism mediated by aggregation of DALPEG-DK5 is highly plausible, we cannot exclude direct interaction between the peptide and the host cell membrane. Indeed, treatment of both human and mouse dendritic cells with LL-37 selectively inhibits TLR4 activation, which is associated with alteration of cell membrane function and structure.35 Moreover, LL-37 binds to murine CD14 and inhibits binding of LPS to CD14+ cells.36 The results obtained using the mouse model of endotoxin shock showed that the tested peptide exhibits efficient endotoxin-neutralizing activity when administered together with LPS comparable to well-known host defense peptide LL37.37 A marked improvement in the survival rate of the animals was observed after treatment with DAL-PEG-DK5 (Figure 7A). In addition, WBC depletion from blood at 3 h post-LPS treatment38 was rescued by DAL-PEG-DK5 (Figure 7C).
■
EXPERIMENTAL PROCEDURES Reagents. Gentamicin, endotoxin (LPS from Escherichia coli 01111:B4; 026:B6), lipoteichoic acid (LTA), peptidoglycan (PGN), TRI Reagent, and Griess reagents were from SigmaAldrich (St. Louis, MO). FBS, DMEM, RPMI 1640, calcium, and magnesium-free phosphate-buffered saline (PBS without Ca2+ and Mg2+), penicillin-streptomycin (PEST), and lymphocyte separation medium were obtained from Gibco (Life Technologies). The CytoTox96 nonradioactive cytotoxicity assay kit was obtained from Promega. Peptides Synthesis. All peptides were synthesized manually by means of a solid phase method based on Fmoc (fluorenyl-9methoxycaronyl) chemistry under standard conditions. 2-Chlorotrityl chloride (substitution 0.48 Mequiv/g; GL Biochem, Shanghai Ltd.) and S RAM (substitution 0.23 Meq/g; RAPP Polymere, Germany) resins were used as a solid support. Fmoc-protected amino acids were attached to the resin using N,N′-diisopropylcarbodiimide (DIPCI) and N-hydroxybenzotriazole (HOBt). After deprotection with 20% piperidine in a mixture of DMF/NMP (1:1, v/v), the peptide chain was elongated using the same DIC/HOBt method and a 3-fold molar excess of Fmoc-protected amino acids or PEG-fragment (Fmoc-O2Oc-OH, Iris Biotech, Germany) over the resin active sites. An N-terminal fluorescein moiety was introduced via its succinimide derivative. Cleavage of peptides from the resin was achieved using a TFA/phenol/triisopropylsilane/H2O mixture (88:5:2:5, v/v). The purity of the peptides was analyzed by reverse phase high performance chromatography (RP-HPLC) on a Varian Prostar HPLC System equipped with an Aeris Widepore 3.6 μm XB-C18 (100 × 4.60 mm) column (Phenomenex, USA) and a UV−vis detector. A linear gradient from 10−90% B over 40 min (A: 0.1% TFA; B: 80% acetonitrile in A; flow rate, 1 mL/min) was employed. The peptides were monitored at 226 nm. Mass spectra were recorded using a Biflex III MALDI TOF mass spectrometer (Bruker, Germany). αI
DOI: 10.1021/acs.bioconjchem.8b00763 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
540 nm was measured using a plate reader. Phenol red-free DMEM supplemented with FBS and antibiotics was used as a blank. A standard curve was prepared using 0−80 μM sodium nitrite solution in H2O. Supernatants from stimulated RAW264.7 cells were analyzed simultaneously by ELISA to measure TNF-α content (BD Biosciences). Measurement of Cytokine Levels in Whole Blood. To measure cytokine release by human blood cells, fresh venous blood taken from healthy donors (collected in EDTA) was treated with 10 ng/mL E. coli (0111:B4)-derived LPS in the presence or absence of DK5, DAL, DK5-PEG-DAL, or DALPEG-DK5 (10 μg/mL). After 10 h of incubation at 37 °C/5% CO2, tubes were centrifuged for 10 min at 200 × g without breaking and plasma was collected for cytokine analysis. The amount of TNF-α was measured in an ELISA (BD Biosciences). Measurement of Cytokine Levels after Bacterial Infection. HMDMs (3 × 105) were incubated in 24-well tissue culture plates in 0.5 mL RPMI 1640 supplemented with 10% human serum. Cells were stimulated with E. coli strain ATCC 25922 (MOI 1:50) in the presence or absence of DK5, DAL, DK5-PEG-DAL, or DAL-PEG-DK5 (10 μg/mL). HMDMs and bacteria were cocultured for 2 h at 37 °C/5% CO2. Then, bacteria were removed by washing the hMDM cultures with icecold PBS. Cells were then cultured in DMEM containing PEST. The level of TNF-α was measured in an ELISA (BD Biosciences) at 22 h post-infection. Binding of LPS to RAW264.7 Cells. The effect of peptides on LPS binding to cells was determined as described previously.37 To this end, RAW264.7 cells (5 × 105 cells) were incubated at 37 °C for 15 min with Alexa Fluor 488conjugated LPS (100 ng/mL, 055:B5, Molecular Probes) in DMEM supplemented with 5% FBS in the absence or presence of each peptide. Cells were then washed twice with ice-cold PBS and LPS binding was analyzed by flow cytometry (FACSCalibur, BD Biosciences). The mean fluorescence intensity and percentage of cells labeled with Alexa Fluor 488-conjugated LPS was measured in each group. LPB Block Assay. The effects of DAL-PEG-DK5 and SCR on the interaction between LPS and LBP were determined as previously described.14 The capture antibody (Human LBP DuoSet ELISA (R&D Systems)) was adsorbed onto a MaxiSorp Loose Nunc-Immuno module (Thermo Fisher Scientific) overnight at 4 °C. Next, plates were blocked for 1 h at 37 °C with 1% BSA in PBS. Recombinant LBP (60 ng/mL, SigmaAldrich) was added to the wells for 1.5 h at 37 °C. After washing the plates, biotinylated LPS (InvivoGen) was added in the absence or presence of peptide (preincubated at 37 °C for 30 min). Polymixin B (InvivoGen) was used as a positive control. The binding of LPS to LBP was measured using HRPconjugated streptavidin (diluted 1:2000) and a TMB Substrate Reagent Set (BD Biosciences). Absorbance at 450 nm was measured in a SpectraMax microplate reader (Molecular Devices). Dimerization Assay. RAW264.7 cells (1 × 106) were stimulated with 100 ng/mL E. coli (0111:B4)-derived LPS in the presence of 25 μg/mL DAL-PEG-DK5 and SCR for 30 min at 37 °C. Next, cells were washed twice with cold PBS/0.1% (w/v) BSA (Sigma-Aldrich) and stained for 30 min at room temperature using a monoclonal rat anti-mouse TLR4/MD2AF488 (clone MTS510, Bio-Rad) or the corresponding isotype control IgG2a-AF488 (Bio-Rad). Cells were analyzed in a FACSCalibur flow cytometer (BD Biosciences).
Cyano-4-hydroxycinnamic acid (CCA) and/or 2,5-dihydroxybenzoic acid (DHB) was used as the matrix. Cell Culture. PBMCs were isolated from human blood obtained from the Red Cross (Krakow, Poland). The Red Cross deidentified blood materials as appropriate to ensure confidentiality. Thus, the study did not require consent from the study subjects. Briefly, PBMCs were isolated from EDTAtreated blood using a lymphocyte separation medium (PAA) density gradient as described previously.37 This yielded a fraction highly enriched in monocytes (90% CD14-positive). Cells were plated in 24-well plates (Sarstedt) at a density of 3 × 106/well and incubated in RPMI 1640 (Gibco) supplemented with 2 mM L-glutamine, 50 μg/mL gentamicin (Sigma-Aldrich), and 10% autologous human serum. After 24 h, nonadherent PBMCs were removed by washing with complete medium and adherent cells were differentiated to hMDMs in this medium for 7 days. Fresh medium was added every 2 days. After nonenzymatic detachment, the phenotype of each batch of hMDMs was routinely monitored by immunofluorescence staining for CD14 (clone DJ130c, DakoCytomation), CD11b (clone ICRF44; BD Biosciences), and CD209 (clone DCN46; BD Biosciences), followed by flow cytometry analysis. The cultures selected for further experiments were at least 90% positive for the first two markers and