Identification and Screening of Potent Antimicrobial Peptides in Arthropod Genomes

Abstract

Using tBLASTn and BLASTp searches, we queried recently sequenced arthropod genomes and expressed sequence tags (ESTs) using a database of known arthropod cecropins, defensins, and attacins. We identified and synthesized 6 potential AMPs and screened them for antimicrobial activity. Using radial diffusion assays and microtiter antimicrobial assays, we assessed the in vitro antimicrobial effects of these peptides against several human pathogens including Gram-positive and Gram-negative bacteria and fungi. We also conducted hemolysis assays to examine the cytotoxicity of these peptides to mammalian cells. Four of the six peptides identified showed antimicrobial effects in these assays. We also created truncated versions of these four peptides to assay their antimicrobial activity. Two cecropins derived from the monarch butterfly genome (Danaus plexippus), DAN1 and DAN2, showed minimum inhibitory concentrations (MICs) in the range of 2–16 μg/ml when screened against Gram-negative bacteria. HOLO1 and LOUDEF1, two defensin-like peptides derived from red flour beetle (Tribolium castaneum) and human body louse (Pediculus humanus humanus), respectively, exhibited MICs in the range of 13–25 μg/mL against Gram-positive bacteria. Furthermore, HOLO1 showed an MIC less than 5μg/mL against the fungal species Candida albicans. These peptides exhibited no hemolytic activity at concentrations up to 200 μg/ml. The truncated peptides derived from DAN2 and HOLO1 showed very little antimicrobial activity. Our experiments show that the peptides DAN1, DAN2, HOLO1, and LOUDEF1 showed potent antimicrobial activity in vitro against common human pathogens, did not lyse mammalian red blood cells, and indicates their potential as templates for novel therapeutic agents against microbial infection.

1. Introduction

Global mortality and morbidity rates have greatly increased over the past few decades due to the emergence of antibiotic-resistant bacteria. According to the Center for Disease Control and Prevention (CDC), there are 2 million new infections in the U.S. every year that are a result of drug-resistant bacteria, and approximately 23,000 deaths annually [1]. Globally, the Antimicrobial Resistance review board estimates that around 10 million people will die annually as a result of drug resistant microbes by the year 2050 [2]. This calls for not only the prudent use of antimicrobial drugs, but also, and more urgently, for the development of novel antimicrobial agents.

One promising strategy for the development of effective antimicrobial agents is an investigation into naturally occurring host defense peptides. Many organisms including insects, plants, amphibians and mammals secrete antimicrobial peptides (AMPs) [3,4]. Antimicrobial peptides are almost always 12–100 amino acids long, cationic, exhibit a net charge of +2 to +9, and amphipathic [5–7]. They are a key component of the innate immune system and are used by organisms as the first line of defense. In particular, the host defense peptides of arthropods provide a useful model for the investigation and synthesis of potent antimicrobial peptides [8,9].

Studies on the antimicrobial effects of invertebrate derived host-defense peptides indicate that they have advantages over conventional antibiotics in fighting against infectious agents. AMPs as a group have broad-spectrum activity against a wide range of microorganisms including Gram-positive and Gram-negative bacteria, protozoa, yeast, fungi and viruses [10,11]. These host defense peptides display a rapid action against various microbes and often have low toxicity to animal cells [12]. Since these peptides have several modes of action and most of them do not target metabolic pathways, it is less likely that the pathogens will develop resistance to them [13].

Many host defense peptides found in nature have been classified on several distinct commonalities such as their amino acid composition and folding patterns. Two groups of AMPs that we are currently investigating in this study are cecropins and defensins. Cecropins are a family of cationic antimicrobial peptides that consist of 31–39 amino acids. They lack cysteine residues and are effective against Gram-negative bacteria [14–16]. In contrast, defensins are 36–46 amino acids long, cationic, contain cysteine residues and can form disulfide bridges [14,16,17]. They are reported to be effective against gram-positive bacteria.

We hypothesized that newly sequenced arthropod genomes contain putative antimicrobial peptide sequences that are effective against human pathogens. In this report, we describe how we identified and screened a group of putative antimicrobial peptides that were derived from the genomic sequence data of recently sequenced arthropods.

2. Materials and Methods

2.1 Bioinformatics and design

We created a library of known cecropins, defensins, attacins, and lysozymes of arthropod origin and downloaded them from the Antimicrobial Peptide Database [18]. Using this database, we queried recently sequenced arthropod genomes and expressed sequence tags (EST) using tBLASTn and BLASTp searches. In 2013, we used the i5K project website to identify arthropod genomes to query [19]. We identified six putative peptides that had significant homologies to queries with known antimicrobial effects (Table 1). LOUDEF1 was a defensin-like molecule identified in the genome of human body louse (Pediculus humanus humanus). DAN1 and DAN2 were cecropins identified in the genomic sequence data from the monarch butterfly (Danaus plexippus), HOLO1 was identified in the genomic sequence from the red flour beetle (Tribolium castaneum). INVICT-1 was identified in the genomic sequence from the red imported fire ant (Solenopsis invicta). We synthesized the truncated versions of DAN2 and HOLO1 by breaking each peptide into segments that were 14–17 amino acid long peptides with 5 amino acid overlaps (Table 2). When these truncated versions were created, we used PEPFOLD [20–22] to predict that each truncated peptide still included an alpha-helix.

Table 1.0.

Amino acid sequences of peptides that were synthesized and screened for antimicrobial activity, and their respective genome of origin.

Peptide Name	Genome of Origin	Amino Acid sequence
DAN1	Danaus plexippus (monarch butterfly)	KWKIFKKIEKVGRNVRDGIIKAGPAVQVVGQATSIAK
DAN2	Danaus plexippus (monarch butterfly)	RWKFLKKIEKVGRKVRDGVIKAGPAVGVVGQATSIYK
HOLO1	Tribolium castaneum (red flour beetle)	VTCDLLSAEAKGVKVNHAACAAHCLLKRKRGGYCNKRRICVCRN
LOUDEF1	Pediculus humanus humanus (human body louse)	ATCDLLSASTPWGSLNHSACAAHCLTKRYKGGRCRNGICRCRR
INVICT1	Solenopsis invicta (red imported fire ant)	ATCDLMSGLGVDHSACAAHCILKGKTGGHCSSTGVCNCRK
IX1	Ixodes scapularis (deer tick)	GFGCPLNQGACHNHCRSIKRRGGYCSGIIKQTCTCYRK

Table 2.0.

Amino acid sequences of the truncated peptides of DAN2 and HOLO1

Truncated peptides	Amino Acid Sequences
Dan 2*	RWKFLKKIEKVGRKVRDGVIKAGPAVGVVGQATSIYK
Dan 2_1	RWKFLKKIEKVGRKVRD
Dan2_2	GRKVRDGVIKAGPAVGV
Dan2_3	GPAVGVVGQATSIYK
Holo1*	VTCDLLSAEAKGVKVNHAACAAHCLLKRKRGGYCNRRICVCRN
Holo1_4	VTCDLLSAEAKGVKV
Holo1_5	KGVKVNHAACAAHCL
Holo1_6	AAHCLLKRKRGGYCN
Holo1_7	GGYCNKRRICVCRN
Melittin*	GIGAVLKVLTTGLPALISWIKRKRQQ

^*Amino Acid Sequences of the parent peptides

2.2 Synthesis of peptides

The peptides used in the research were commercially synthesized by GenScript (Piscataway, NJ), which utilized solid phase peptide synthesis, and purified the peptides using HPLC to >75% purity. All peptides were synthesized with C-terminal amidation to reduce enzymatic degradation and promote stability. Upon obtaining the peptides, we dissolved each in 0.01% acetic acid solution and stored them at −70 °C in working stock solutions for further antimicrobial assays.

2.3. Bacterial strains

We used the following strains of bacteria for screening antimicrobial peptides in this study: Escherichia coli (DH5α), Staphylococcus aureus (ATCC# 29213), Bacillus subtilis (Carolina Biological), Pseudomonas aeruginosa (ATCC# 27850), Klebsiella pneumoniae (ATCC# 700603), Candida albicans (ATCC# 90028).

2.4. Radial Diffusion Assay

The antimicrobial activities of the peptides were evaluated by a modification of the sensitive radial diffusion assay described by Steinberg and Lehrer [23]. Prior to the assay, bacteria were grown overnight to the stationary phase at 37° C in 50 mL Tryptic Soy Broth (TSB) (Carolina Biological) with shaking at 200 rpm. We transferred 100 μl of the overnight culture to 50 mL fresh TSB broth and incubated the culture for several hours at 37°C to obtain a mid-logarithmic phase culture. We transferred the subculture to a 50-mL conical tube and centrifuged for 10 min at 4°C at approximately 880 × g. After washing the bacterial pellet with 10 mL of cold sterile 10 mM sodium phosphate buffer (pH 7.4), we suspended bacteria in 5 mL of sodium phosphate buffer. We transferred 10-mL aliquots of the sterile, molten underlay agar to a 15-mL conical plastic centrifuge tube and inoculated with 4 × 10⁶ CFU of washed bacteria. We vortexed the mixture vigorously for 15 s, and then poured into a 100 mm × 15 mm petri dish on a level table. After the underlay gel solidified, the plates were placed over a paper template and 8 wells, 3.5 mm in diameter, were punched by suction, using pipet tips that were cut to the specified diameter. We prepared two fold serial dilutions of the peptides and added 5 μl of peptide of desired concentration to each well in turn. We incubated plates for 3 h at 37°C and then, added 10 mL of nutrient-rich-overlay gel. After incubating the plates overnight, we measured the diameter of the clear zone surrounding each well using digital calipers with the help of 7× magnifying glass. We repeated each experiment three times.

2.5. Broth microtiter assay (MIC determinations)

We measured the MIC of candidate peptides, both parent and truncated, using a microtiter broth dilution method as described by Conlon and Sonnevend [24]. Briefly, the microbes were grown overnight in Mueller-Hinton Broth (MH broth) (Sigma-Aldrich) at 37°C to stationary phase; plates were shaken at 200 rpm. On the day of the assay, we transferred 200 μL of the overnight culture to a flask with 20 mL MH broth and incubated the culture at 37°C to obtain a mid-exponential growth. We then diluted microbial cultures to a concentration of 10⁶ CFU/mL. We added 50 μL of the serially diluted peptide solution into each well containing 50 μl of microbial agent (10⁴ CFU). Subsequently, we incubated the 96 well plates for 20 hours at 37°C and measured OD₆₀₀. The MIC was defined as the lowest concentration of the peptide required to inhibit the growth of the microbial agent (OD₆₀₀ < 0.04). We determined the MIC of each peptide from at least three independent experiments. Melittin, a potent microbicidal peptide found in honeybee venom, was used as our positive control. We included broth only and microbe only wells to test for contamination and adequate growth in each assay respectively.

2.6 Hemolysis assay

We determined the hemolytic activity of peptides against mammalian cells by using sheep erythrocytes as described previously [25]. Briefly, we centrifuged the whole sheep blood (Fisher Scientific) at 400 × g for 10 minutes at 4°C to isolate erythrocytes. We washed 3 mL of packed erythrocytes with Phosphate buffered saline (PBS) until the supernatant was clear and subsequently re-suspended erythrocytes in 20 mL of PBS. We serially diluted the peptides in PBS and added to 180 μL of erythrocyte suspension in a 96-well plate. Melittin served as our positive control and 0.2% Triton-X was used to determine 100% lysis. We incubated the plate at 37°C for 18–24 hours and then centrifuged the cells treated with peptides for 5 minutes. We pipetted an equal volume of supernatant from each well (20 μl), diluted ten-fold with PBS (180 μl) and measured the absorbance of the solution at OD 567 nm.

Results

3.1. In-vitro Antimicrobial Activity of the peptides

3.1.1. Peptide sequences and their notable properties

The peptides investigated in this research vary in origin, sequence length, composition; as well as net positive charge and number of hydrophobic residues at a physiological pH. Table 1 and 2 describe the amino acid sequences of the peptides screened in this study and their genome of origin.

3.1.2. Parent peptides inhibit microbial growth

The peptides, DAN1, DAN2, HOLO1, and LOUDEF1, exhibited antimicrobial activity in both radial diffusion assays and broth microtiter assays to varying degrees as indicated by zone of inhibition (supplemental data) and MIC measurements (Table 3). IX1 and INVICT1 did not produce any measureable antimicrobial activity against S. aureus and E. coli and were not further screened. DAN1 and DAN2 had MIC values in the range of 2.1–16.2 μg/ml, when screened against Gram-negative bacteria: E. coli, P. aeruginosa, and K. pneumoniae. Markedly, DAN2 was quite effective against Gram-negative bacteria as indicated by the low MIC value of 2.1 μg/ml against E. coli. DAN1 and DAN2 showed less potential to suppress the growth of Gram-positive bacteria we tested, B. subtilis, S. aureus, and the fungus C. albicans as suggested by higher MIC values. The MIC values of defensin-like peptides, HOLO1 and LOUDEF1, against Gram-positive bacteria and a fungus species were in the range of 5–35 μg/ml. HOLO1 showed potent antimicrobial activity against Gram-positive bacteria with MIC value as low as 6 μg/ml and against C. albicans with an MIC value as low as 5 μg/ml. HOLO1 and LOUDEF1 were not effective in inhibiting the growth of Gram-negative bacteria as suggested by higher MIC values. INVICT1 and IX1 did not exhibit any antimicrobial properties and had MIC values > 100 μg/ml for all microbes they were screened against (data not shown). The positive control, Melittin displayed MIC values in the range of 7–27 μg/ml when screened against various microbes.

Table 3.0.

Minimum Inhibitory Concentrations (MIC) for the AMPs against a range of microbes was determined by broth micro-dilution assay. Melittin was used as a positive control. Average MIC values were calculated from at least three independent experiments.

	MIC (μg/ml)
	Escherichia coli	Staphylococcus aureus	Bacillus subtilis	Candida albicans	Pseudomonas aeruginosa	Klebsiella pneumonia
AMPs	(Gram − bacteria)	(Gram + bacteria)	(Gram + bacteria)	(fungal pathogen)	(Gram − bacteria)	(Gram − bacteria)
DAN1	4.9 ± 2.3	>200	131.3 ± 46.1	>200	7 ± 0.8	16.2 ± 1.9
DAN2	2.1 ± 0.7	>100	50 ± 19	72.5 ± 35.2	5 ± 2.5	4.3 ± 1.4
HOLO1	>75	13.2 ± 1.8	6 ± 1.8	4.6 ± 1.1	>200	>200
LOUDEF 1	>100	35 ± 1.3	7 ± 1.6	19.1 ± 9.6	>100	>200
Melittin
(Positive Control)	12.5 ± 1.8	6.8 ± 2	5.8 ± 0.7	7.8 ± 1.8	43.7 ± 4.9	26.6 ± 5.5

In an attempt to determine a shorter sequence that would retain the antimicrobial property of the parent peptides, we proceeded to truncate two of the most effective peptides: DAN2 and HOLO1. The peptides were 14–17 amino acid long with 5 amino acid overlaps in between the neighboring fragments (Table 2). The truncated versions of DAN2 and HOLO1 failed to exhibit any MIC values less than 50 μg/ml against E. coli, S. aureus, or C. albicans.

3.1.3. The peptides are not hemolytic at concentrations up to 200 μg/ml

We tested the peptides’ ability to lyse mammalian cells using sheep erythrocytes. All peptides tested were nontoxic to mammalian cells at concentrations ten or twenty times higher than the MICs required for the antimicrobial activity. Figure 1.0 depicts the hemolytic activity of the AMPs compared with the positive control melittin. Melittin lysed RBCs at concentrations of 12.5 μg/ml and higher. However, there was no observable cytotoxicity of the designed peptides at concentrations less than 200 μg/ml.

Figure 1.0.

The antimicrobial peptides DAN1, DAN2, HOLO1, and LOUDEF1 had no hemolytic activity at concentrations ranging from 12.5 μg/ml to 200 μg/ml. Melittin showed hemolytic activity beginning at 12.5 μg/ml. The erythrocytes suspended in PBS were added in 96 well plates with serially diluted peptides concentration ranging from 12.5 μg/ml to 200 μg/ml. Absorbance was read at 567 nm. The error bars are the standard error of the mean.

4. Discussion

Antimicrobial resistance to antibiotics is a serious and urgent global health threat. The rate at which infectious agents evolve to acquire resistance to conventional antibiotics has outpaced the rate of development of new therapeutic agents [26,27]. AMPs that are recruited by innate immune system of invertebrates have received attention as potential therapeutic agents because they exhibit broad antimicrobial activity and low hemotoxicity [3,7,28,29]. With the development of next generation genomic sequencing techniques, it is increasingly feasible to bioinformatically identify AMPs in various organisms, generate, and then screen candidate peptides for their antimicrobial properties.

We investigated the antimicrobial activity of six host defense peptides derived from arthropods. The results demonstrate that four out of six peptides - DAN1, DAN2, HOLO1 and LOUDEF1 - that were tested exhibited potent antimicrobial activity against clinically relevant microbes. On the contrary, INVICT1 and IX1 displayed no antimicrobial activity. Both cecropins, DAN1 and DAN2, had lower MICs than melittin when screened against gram-negative bacteria. The defensins, HOLO1 and LOUDEF1, also showed considerably low MIC, values comparable to that of melittin.

The MIC values of DAN1 and DAN2 are consistent with the literature values of other cecropins. For instance, the MIC values of cecropin A, cecropin B, and cecropin P1 were determined to be in the range of 2–20 μg/ml against different strains of various gram-negative bacteria, mostly E.coli and P. aeruginosa (Table 5.0). DAN1 and DAN2 exhibited MIC values in the range of 1–20 μg/ml against gram-negative bacteria in our studies.

Table 5.0. (A & B).

Comparisons of MIC values of insect cecropins (A) and defensins (B) with LOUDEF1 and HOLO1.

A.
Arthropod Species	Peptide’s Name	MIC values (μg/ml)			Reference
		E.coli	S.aureus	P.aeruginosa
Hyalophora cecropiae	Cecropin B	2	n.d	>100	[30]
	Cecropin B	16	>128	n.d	[31]
	Cecropin A	2.5	n.d	2.5	[28]
	Cecropin B	2.5	n.d	2.5	[28]
	Cecropin P1	0.12–4	8–>128	4–64	[32]
Danaus plexippus	DAN1	4.9	>200	75	Current study
	DAN2	2.1	>100

B.
Arthropod Species	Peptide name	MIC values (μg/ml)				References
		E.coli	S.aureus	Bacilius subtilis	Fungus
Hemipteran insect (Podisus Maculiventris)	C-Thanatin	1	256	ND	ND	[33]
The black soldier fly (Hermetia illucens (L.)	Trx-stomoxynZH1	15–30	27–54	ND	> 98 against R. solani Khun	[34]
Taiga Tick, lxodes persulcatus	Defensin	41.9	3.3	2.1	ND	[35]
Flesh Fly (Sarcophaga peregrine)	Sapecin	>78.0	5.0	20–39	ND	[36]
Fruit fly (Bactrocera dorsalis)	Bactrocerin-1	>232.5	6.5	13.0	26.0–104.7	[37]
Great wax moth (Galleria mellonell)	Galleria defensin	ND	ND	ND	14.7–29.5	[38]
Spider (Ornithoctonus haina)	Oh-defensin	1.3	1.25	ND	1.3	[39]
Red flour beetle	HOLO1	>75	13.2	6	4.6 against C. albicans	Current study
Human body louse	LOUDEF1	>100	35	7	19.1 against C. albicans	Current Study

Similarly, the defensin derivatives, HOLO1 and LOUDEF1, had antimicrobial activities against Gram-positive bacteria and a fungus (C. albicans). The MIC values of these two peptides were comparable with the MIC values reported in literature. As seen in the Table 5.0 (B), the MIC concentrations of insect defensins fall in the range of 1.3–39 μg/ml when tested against different strains of Gram-positive bacteria, mostly S.aureus and B. subtilis. HOLO1 and LOUDEF1 showed activity in a similar range: 6–36 μg/ml, against S.aureus and B. subtilis. The insect defensins reported in the literature showed activity against different strains of fungi at concentrations as low as 1.3 μg/ml and as high as 98μg/ml. HOLO1 and LOUDEF1 exhibited antifungal activity at concentrations higher than 4.5 μg/ml and 19 μg/ml respectively.

Several studies have attributed the activity of AMPs to their net positive charge and hydrophobicity (30, 31). Our peptides have an average positive charge of 7.5 and an average hydrophobicity of 62.5%, which we believe contributed to the observed antimicrobial activity. As reported in the literature, a charge driven interaction between the negatively charged bacterial cell wall and the positively charged peptide induces the activity of the peptides (18, 32). As reported in the literature, the potency of a peptide is in part dependent on its net positive charge, an explanation supported by the observations in this research; DAN2 and HOLO1 (+8) had a more potent antimicrobial activity than DAN1 and LOUDEF1 (+7) [41]. The truncated peptides of DAN2 and HOLO1 did not retain any antimicrobial activity as indicated by higher MIC values. Their lack of antimicrobial activity is likely due to decreased net positive charge and hydrophobic residues.

The intrinsic ability of defensins to form disulfide bonds in the presence of cysteine residues distinguishes defensins from cecropins in terms of their conformations. In the absence of any cysteine groups, the cecropins derivatives form random coil structures in an aqueous environment but convert to alpha-helical structure in an hydrophobic environment, which enables them to penetrate the LPS of the Gram-negative bacteria [14,42]. On the other hands, defensins form channels in the phospholipid membrane and promote lipid clustering in Gram-positive bacteria. The LPS present in the outer membrane of Gram-negative bacteria might permit the interaction of defensins with the inner phospholipid membrane [14].

Hemolytic activity of these peptides against sheep erythrocytes was not detected. This suggests that the peptides are non-toxic to mammalian cells even at concentration up to twenty times greater than the MICs of the peptides. The lack of cytotoxicity of the peptides to mammalian cells is likely due to the absence of charge driven attraction between the peptides and the neutral cell surface of the mammalian cells. Lack of cytolytic activity against mammalian cells supports expected safety for therapeutic use.

In conclusion, our results demonstrated that the peptides derived from arthropods genomes exhibited potent antimicrobial activities and low hemolytic activity. Sequenced arthropod genomes provide a vast array of new antimicrobial peptides to screen and develop. Our method of searching arthropod sequence databases for potential antimicrobial peptides, synthesizing these peptides, and screening them for antimicrobial properties has already proved fruitful and should yield additional candidate peptides in the future. In future studies, we plan to test these antimicrobial peptides against multi-drug resistant strains of pathogens. We will also use rational design and directed evolution to modify some of these peptides to further improve their antimicrobial properties.

Highlights.

• Several cecropin and defensin derivatives were identified via BLAST searches in recently sequenced arthropod genomes such as the monarch butterfly (Danaus plexippus), human body louse (Pediculus humanus humanus), and red flour beetle (Tribolium castaneum). Predicted peptide sequences were synthesized and screened for antimicrobial activity.

• Four out of six peptides synthesized and tested, LOUDEF1, HOLO1, DAN1, and DAN2, showed potent antimicrobial activity against bacterial and fungal human pathogens. Also, these peptides exhibited no hemolytic activity at a concentration as high as 200 μg/ml.

• Our study shows that sequenced arthropod genomes provide amino acid sequence information for a vast array of new antimicrobial peptides to screen and develop as potential therapeutic agents.