Exploring Galleria mellonella larval model to evaluate antibacterial efficacy of Cecropin A (1-7)-Melittin against multi-drug resistant enteroaggregative Escherichia coli

ABSTRACT

High throughput in vivo laboratory models is need for screening and identification of effective therapeutic agents to overcome microbial drug-resistance. This study was undertaken to evaluate in vivo antimicrobial efficacy of short-chain antimicrobial peptide- Cecropin A (1–7)-Melittin (CAMA) against three multi-drug resistant enteroaggregative Escherichia coli (MDR-EAEC) field isolates in a Galleria mellonella larval model. The minimum inhibitory concentration (MIC; 2.0 mg/L) and minimum bactericidal concentration (MBC; 4.0 mg/L) of CAMA were determined by microdilution assay. CAMA was found to be stable at high temperatures, physiological concentration of cationic salts and proteases; safe with sheep erythrocytes, secondary cell lines and commensal lactobacilli at lower MICs; and exhibited membrane permeabilization. In vitro time-kill assay revealed concentration- and time-dependent clearance of MDR-EAEC in CAMA-treated groups at 30 min. CAMA- treated G. mellonella larvae exhibited an increased survival rate, reduced MDR-EAEC counts, immunomodulatory effect and proved non-toxic which concurred with histopathological findings. CAMA exhibited either an equal or better efficacy than the tested antibiotic control, meropenem. This study highlights the possibility of G. mellonella larvae as an excellent in vivo model for investigating the host-pathogen interaction, including the efficacy of antimicrobials against MDR-EAEC strains.

antimicrobial peptide, Cecropin A (1-7)-Melittin, enteroaggregative Escherichia coli, Galleria mellonella, multi-drug resistance

INTRODUCTION

The search for reliable and ergonomic in vivo laboratory models that could simulate the clinical manifestations of humans are essential to investigate the host–pathogen interaction and to identify novel antimicrobial agents (Andrea, Krogfelt and Jenssen 2019; Cutuli et al. 2019). Although mammalian models like mice, rats and guinea pigs are widely exploited due to its similarities in human physiology, the budgetary, logistical and ethical complications create a hurdle in large scale settings. Considering the concerns in lab animal welfare and complications associated with their use, several invertebrate models were explored as preliminary in vivo screening models to reduce the candidate molecules for its further evaluation in mammalian models (Andrea, Krogfelt and Jenssen 2019). The nematode Caenorhabditidis elegans, Drosophila melanogaster flies and of late, the larvae of Galleria mellonella are widely employed in vivo invertebrate models (Cutuli et al. 2019). Owing to the short life span, ease in maintenance and ability to mimic the human host while investigating the pathogens at 37°C, G. mellonella larval model has recently gained more attention. Hence, this larval model, which is considered as a new ‘laboratory rat’, has extensively been employed in exploring the pathogenicity of infectious agents, host-pathogen interaction studies and identification of novel antimicrobial agents against pathogens of public health significance (Piatek, Sheehan and Kavanagh 2020; Wojda et al. 2020).

Of late, an escalating trend in the antimicrobial resistance (AMR) has been reported throughout the world. Apart from various national as well as international strategic policies, research employing antimicrobial alternatives (phages, phytochemicals, antibodies, prebiotics and probiotics, vaccines and peptides) has been documented to combat this public health issue of AMR (Ghosh et al. 2019). In recent times, cationic antimicrobial peptides (AMPs) are evaluated as potential therapeutic candidates on a large-scale basis to treat drug-resistant pathogens of public health significance (Haney, Straus and Hancock 2019). AMPs are typical amphipathic peptides with a net positive charge at physiological pH and evolutionarily conserved molecules found in a wide range of organisms (from lower-order prokaryotes to higher-order human beings). Generally, these cationic peptides are indicated as impending antibiotic substitutes with reportedly less resistance towards pathogens (Haney, Straus and Hancock 2019; Yazici et al. 2019). Nevertheless, the cytotoxicity as well as protease susceptibility of AMPs hinders it's in vivo efficacy. For instance, although Cecropin-A (37 amino acids) and melittin (26 amino acids) have a broad-spectrum of action against pathogens, the susceptibility of Cecropin-A towards protease degradation ability as well as the haemolytic property of melittin hinders its utility as an antimicrobial agent for treatment (Mahdavi Abhari, Pirestani and Dalimi 2019). Hence, Cecropin–melittin hybrid peptide viz. Cecropin A (1–7)-Melittin (CAMA), a pentadecapeptide comprised of the cationic region of ‘Cecropin-A’ and the hydrophobic as well as a non-haemolytic region of the bee venom peptide ‘melittin’ is employed in antimicrobial efficacy studies. Besides, this peptide has been reported to exhibit it's in vitro efficacy against Gram-negative and Gram-positive bacteria, fungi (Ali and Reddy 2000), Leishmania spp. (Chicharro et al. 2001), Entamoeba histolytica (Mahdavi Abhari, Pirestani and Dalimi 2019) and produce membrane penetration effect (Silva et al. 2018). Nonetheless, CAMA has not extensively been employed against drug-resistant pathogens, particularly multi-drug resistant (MDR) pathogens, barring a few studies against Pseudomonas aeruginosa, methicillin-resistant Staphylococcus aureus, Acinetobacter baumannii (Geitani et al. 2019; Neshani et al. 2020). The mammalian models require a higher amount of peptides for its in vivo translation in suitable laboratory animals which would cost exorbitant. Hence, synthetic short-chain AMPs are often employed as therapeutic alternatives for drug-resistant pathogens owing to the high cost of synthesis and research involved while using long-chain peptides (de la Fuente-Núñez et al. 2016). On the other hand, the in vivo translation of CAMA has yet to be explored. Such a situation demands to screen peptides in high throughput in vivo models, like G. mellonella larvae, which could mimic humans in terms of host-pathogen interaction to arrive at an immediate conclusion.

Enteroaggregative Escherichia coli (EAEC) is a food-borne pathogen quite often neglected, however, emerging and is associated with both endemic as well as epidemic diarrhoeal episodes. EAEC is responsible for causing damage to the intestinal epithelium, poor absorption, malnutrition and eventually intellectual deficits in human infants, whereas diarrhoeal episodes are prominent features in animals (Cabrera-Sosa and Ochoa 2020). The Shiga-toxin-producing EAEC strain—O104: H4 recovered from the European outbreak of 2011 has reported considerable economic losses (Borgersen et al. 2018). EAEC infection progresses with an initial adhesion to the intestinal epithelium leading to biofilm formation which would ultimately result in the toxin release consequent to the inflammatory response. The biofilm-forming ability of EAEC, illustrated by ‘stacked-brick’ pattern on HEp-2 cells, is closely associated with its persistence of infection and thereby, drug resistance (Kong, Hong and Li 2015; Cabrera-Sosa and Ochoa 2020). Even though quinolones and beta-lactam antibiotics have resorted as first-hand therapeutic agents for EAEC associated diarrhoea, drug resistance has commonly been reported towards these antimicrobials (Lima, Medeiros and Havt 2018). In recent times, EAEC infection has been well established in the G. mellonella model (Jønsson et al. 2017). Nonetheless, the antimicrobial efficacy of AMPs against MDR-EAEC strains remains unattempted to date, barring our own study (Vergis et al. 2019); hence, the present study was conducted to characterize an in vivo G. mellonella larval model for evaluating the antimicrobial efficacy of CAMA against MDR-EAEC strains.

MATERIALS AND METHODS

Bacterial strains

The typical EAEC field isolates maintained in the repository of Comparative Pathology and Biomedicine laboratory at Division of Veterinary Public Health of ICAR, Indian Veterinary Research Institute, Izatnagar were revalidated using PCR as well as HEp-2 cell adherence assays (Vijay et al. 2015) and tested for antibiotic susceptibility (CLSI 2018). The EAEC strains allocated with NCBI GenBank accession numbers KY941936.1 (MDR 1); KY941937.1 (MDR 2) and KY941938.1 (MDR 3) were used in this study. Lactobacillus rhamnosus MTCC 1408 and L. acidophilus MTCC 10307 strains were procured from the Institute of Microbial Technology (IMTECH, India).

Materials and reagents

Antimicrobial peptide and antibiotic

The amino acid sequence of CAMA (Table S1, Supporting Information) was retrieved from BaAMPs (di Luca et al. 2015) database, synthesized commercially (purity ≥ 95%) from Shanghai Science Peptide Biological Technology, China, resuspended in phosphate-buffered saline (PBS; pH 7.40) with a final stock concentration of mg/mL and stored at −20°C until further use. The antibiotic control, meropenem (Macleods Pharmaceuticals Pvt. Ltd., Mumbai, India) was used in this study to compare the results of CAMA instead of a control peptide.

Media

The media used in this study viz., eosin methylene blue (EMB) agar, cation-adjusted Mueller Hinton (CA-MH) broth and de Man, Rogosa and Sharpe (MRS) agar were procured from HiMedia Laboratories Pvt. Ltd., Mumbai, India.

Analytical reagents and kit

The analytical reagents used in this study were procured from HiMedia Lab Pvt. Ltd. and Sisco Research Laboratories (SRL) Pvt. Ltd., India, while the QuantiChrom LDH cytotoxicity assay kit was procured from BioAssay Systems, USA.

Characterization of CAMA

CAMA was characterized for minimum inhibitory concentration (MIC), minimum bactericidal concentration (MBC; Table 1), in vitro stability (temperature, proteases and physiological concentration of salts) and safety (haemolysis and cytotoxicity) assays (Figures S1 and S2, Supporting Information). The membrane permeabilization effect of CAMA on MDR-EAEC strains was assessed by flow cytometry, while, the outer and inner membrane permeability of MDR-EAEC isolates treated with MIC and MBC values of CAMA was carried out based on the nitrocefin activity as well as the release of cytoplasmic β-galactosidase activity, respectively (Figures S3–S5, Supporting Information). Further, CAMA was evaluated for its antibacterial effect against commensal gut lactobacilli (L. rhamnosus MTCC 1408 and L. acidophilus MTCC 10307; Figure S6, Supporting Information).

Table 1.

In vitro antibacterial activity of CAMA against MDR-EAEC isolates.

NCBI GenBank accession no.	CAMA		Meropenem
	MIC (mg/L)	MBC (mg/L)	MIC (mg/L)	MBC (mg/L)
KY941936.1 (MDR 1)	4.0	4.0	<0.50	<0.50
KY941937.1 (MDR 2)	2.0	4.0	<0.50	<0.50
KY941938.1 (MDR 3)	2.0	2.0	<0.50	<0.50

MIC- Minimum Inhibitory Concentration; MBC- Minimum Bactericidal Concentration

In vitro growth kinetics of MDR-EAEC with CAMA

The in vitro growth kinetics of each of MDR-EAEC isolates were evaluated by incubating the log-phase bacterial cultures (1 × 10⁷ CFU/mL) in CA-MH broth with MIC and MBC levels of CAMA, in triplicates. The desired bacterial numbers for each MDR-EAEC isolate and CAMA were suspended in CA-MH broth as follows: Group I, 10⁷ CFU of MDR- EAEC (50 µL; final concentration of 5 × 10⁵ CFU/mL) with 50 µL of CAMA (final concentration of 1X MIC); Group II, 10⁷ CFU of MDR- EAEC (50 µL; final concentration of 5 × 10⁵ CFU/mL) with 50 µL of CAMA (final concentration of 1X MBC); Group III, 10⁷ CFU of MDR- EAEC (50 µL; final concentration of 5 × 10⁵ CFU/mL) with meropenem (final concentration of 1X MIC; 50 µL); Group IV, 10⁷ CFU of MDR- EAEC (50 µL; final concentration of 5 × 10⁵ CFU/mL) with CA-MH broth (50 µL). Similar groups were also made for the other two MDR-EAEC isolates. The respective groups along with the appropriate controls were incubated at 37°C up to 72 h. To study the antibacterial effect of CAMA on MDR-EAEC isolates, an aliquot (10 µL) from all the four groups were taken at 0, 30, 60, 90, 120, 150, 180 min, 24 h, 48 h and 72 h and were serially diluted 10-fold in 0.85% sodium chloride; the last three dilutions placed on EMB agar plates containing 100 µg of ampicillin were counted after 24 h of incubation at 37°C, and the bacterial counts were expressed as log₁₀CFU/mL (Miles, Misra and Irwin 1938).

In vivo assays using G. mellonella model

In vivo assays were performed using the fifth instar of homegrown G. mellonella larvae (200–250 mg weight; 2.0–2.5 cm length) stored in wood shavings at 15°C in the dark before the experiment. The larvae were kept in a germ-free environment and were provided with ad libitum food during the course of the experiment. Initially, the LD₅₀ dose of each MDR-EAEC strains was determined in G. mellonella larvae, and the validated LD₅₀ dose was used further in the in vivo studies to evaluate the antibacterial effect of CAMA (Figure S7, Supporting Information).

Galleria mellonella larvae (n = 40) per group per experiment were used. Thus, for a single experiment, we used 200 larvae. The experiment was performed three independent times, so overall 600 larvae were used. The larvae were grouped as follows: Group I (infected group), Group II and III (infection + treatment groups), Group IV (PBS control) and Group V (AMP control). Larvae from groups I to III were infected with a LD₅₀ dose (10⁶ CFU/larvae) of cocktail mixture prepared by mixing equal concentration (0.08 OD600 nm) of the three MDR-EAEC strains; groups II and III were administered 3 h post-infection (pi) with MIC dose of CAMA and meropenem, respectively; Group IV were injected with sterile PBS whereas, group V was administered with MIC dose of CAMA.

Galleria mellonella larvae were then subjected to the following assays, at an interval of 6 h up to 24 h, followed by 24 h interval until 96 h pi.

Melanization assay

For determination of melanin production (immune marker), the haemolymph of G. mellonella larvae (n  = 3 per group) was collected at each time point in a sterile pre-chilled eppendorf tube. N-phenylthiourea (10–20 crystals) was added to the tubes to prevent further melanization of the larval haemolymph. An aliquot of the pooled haemolymph (100 µL) was transferred into a 96-well microtitre plate for measuring the absorbance value at a wavelength of 450 nm using an ELISA plate reader (Multiskan; Thermo Scientific, GO, USA) with an absorbance of haemolymph collected from apparently healthy un-inoculated larvae as the background control (Gibreel and Upton 2013).

Enumeration of MDR-EAEC counts

The aseptically collected haemolymph of G. mellonella (n = 3 larvae per group) at respective intervals was serially diluted 10-fold in sterile normal saline solution (NSS), vortexed thoroughly and the bacterial burden was assessed on EMB agar plates supplemented with 100 µg of ampicillin per plate (Miles, Misra and Irwin 1938). The MDR-EAEC colonies on the plates were counted and expressed as log₁₀CFU/mL of larval haemolymph.

Enumeration of haemocytes

The haemocyte density of G. mellonella (n = 3 per group) larvae was quantified at corresponding intervals (Wand et al. 2013). No attempt was made to discriminate various haemocyte subtypes.

Lactate dehydrogenase (LDH) cytotoxicity assay

Galleria mellonella larvae (n = 3 per group) were analysed for the production of cytotoxicity marker-LDH, at specific intervals using QuantiChrom LDH cytotoxicity assay kit, according to the manufacturer's instructions (Wand et al. 2013). Distilled water and 20% Triton X-100 served as the positive and negative controls, respectively. The absorbance was read using ELISA plate reader at 500 nm; the haemolymph from untreated larvae served as background control and the cytotoxicity was calculated as, Cytotoxicity (%) = (OD_Sample—OD_Control)/(OD_{Total Lysis}—OD_Control) x 100, wherein, a sample is the control absorbance of treated cell; control is the experimental absorbance of the untreated cell control and total lysis is the absorbance of Triton X-100 treated cells.

Histopathological Examination

The larvae at each time point were subjected to histopathological examination to study the tissue-level changes (Perdoni et al. 2014). The neutral buffered formalin (10%)-fixed whole larvae were dissected transversally into two halves employing anatomic pincers and by using a new lancet blade for each larva. The procedure was carefully performed to avoid the squeeze of the larval tissues; each paraffin-embedded section (3 µ) was then sectioned and stained using Haematoxylin and Eosin (H&E) to evaluate tissue morphology observing standard laboratory protocols. The microscopic visualization was performed (Leica Microscope DMLB) and the image acquisition was carried out (NanoZoomer-XR C12000, Hamamatsu Photonics, Japan).

Statistical analysis

All the experiments were repeated individually and independently thrice and the data obtained is reflected as mean ± standard deviation for each assay. The data was statistically analysed using GraphPad Prism 5.01 (GraphPad Software Inc., San Diego, CA). A one-way analysis of variance (ANOVA) with Bonferroni multiple comparison post-test was used to compare the differences between cytotoxicity of control and AMP-treated cell lines (Fig. 1A). The association of CAMA on commensal gut lactobacilli was measured by a paired two-tailed ‘t’ test (Fig. 1B). A two-way (repeated measures) ANOVA with Bonferroni multiple comparison post-test was used to compare the differences between control and AMP-treated tests for the in vitro (Fig. 2) and in vivo time-dependent antimicrobial assays (Fig. 4). In vivo G. mellonella larval survival curves were analysed by logrank (Mantel–Cox) test and logrank test for trends (Fig. 3) while, the LD₅₀ of the MDR-EAEC isolates were determined by the probit-regression model.

Figure 1.

Characterization of CAMA. In vitro cytotoxicity (A) observed for CAMA at different concentrations on HEp-2 and RAW 264.7 cells. In vitro effect of CAMA on commensal gut lactobacilli (B). Untreated L. acidophilus and L. rhamnosus serve as positive growth control while, media (MRS broth) serve as negative control. (C) denotes in vitro inner and outer membrane permeability assay of MIC (1X and 2X) concentrations of CAMA against MDR-EAEC strains employing ONPG and Nitrocefin expressed as a function of time at 37°C. Hydrolysis of ONPG by β-galactosidase was used to monitor inner membrane permeabilization by absorbance at 420 nm. Hydrolysis of nitrocefin by β-lactamase was used to monitor outer membrane permeabilization by absorbance at 486 nm.

Figure 2.

Dose- and time-dependent killing kinetics of MDR-EAEC isolates co-cultured with CAMA at different time intervals. Three MDR-EAEC isolates were co-cultured with: MBC of CAMA (A); MIC of CAMA (B) in CA-MH broth at 37°C under static conditions. Simultaneously, respective controls of MDR-EAEC isolates (untreated and meropenem-treated) were incubated in CA-MH broth. Data expressed as the mean ± standard deviation (log₁₀CFU/ml) of three independent experiments (***P < 0.001).

Figure 3.

Survival plot of Galleria mellonella larvae infected with LD₅₀ dose of MDR-EAEC isolates (10⁶ CFU) and treated with MIC dose of CAMA 3 h post-infection. Survival plots of MDR-EAEC infection treated with MIC (1X) of CAMA, keeping respective controls (infection, meropenem-treatment, CAMA control, PBS control) were expressed as the mean of three independent experiments (***P < 0.001).

RESULTS

The three typical MDR-EAEC isolates included in the study were resistant to four or more classes of antibiotics and were ESBL-producers (Table S2, Supporting Information).

Characterization of CAMA

The MBC values for CAMA were either equal or 2-fold greater than the MIC values (Table 1) against MDR-EAEC strains. The MIC values for the antibiotics (ampicillin, ceftazidime and ciprofloxacin) used against E. coli ATCC 25922 were within the prescribed CLSI range (data not shown).

CAMA was found to be thermostable as evidenced by its retained antimicrobial activity after incubation at 70°C and 90°C for 5 min, 15 min and 30 min, respectively (Table S3, Supporting Information). Moreover, a 2-fold increase in the antimicrobial activity (MIC and MBC) of CAMA was noticed upon trypsin exposure (Table S3, Supporting Information) as well as on exposure to proteinase-K (Table S3, Supporting Information); hence, CAMA was found to be protease stable. Further, CAMA retained the antimicrobial activity during incubation at higher concentrations of salts (150mM NaCl and 2mM MgCl₂), though a slight decrease was observed in the MBC values for the two EAEC strains tested (Table S3, Supporting Information).

CAMA was found to be non-haemolytic at 1X and 2X MIC concentrations; however, minimal haemolysis (<6%) was noticed at 4X, 5X and 10X MIC levels (Table 2). CAMA marginally decreased the cell viability of the human epithelioma cell line (HEp-2) and murine macrophage cell line (RAW 264.7), in a concentration-dependent manner (Fig. 1A). The most typical morphological changes caused by CAMA were shrinkage and vacuolization of cytoplasm and loss of monolayer at higher concentrations (4X, 5X and 10X MIC). Cells treated with lower peptide concentrations (1X and 2X) did not exhibit remarkable morphological changes.

Table 2.

In vitro cytotoxicity of CAMA (at different MIC concentrations) against sheep RBCs.

Concentration	Haemolysis (%)
MIC (1X)	0.00
MIC (2X)	0.00
MIC (4X)	1.27
MIC (5X)	3.47
MIC (10X)	5.89

The commensal lactobacilli (L. acidophilus and L. rhamnosus) revealed a similar growth pattern, regardless of CAMA treatment (Fig. 1B). Overall, a non-significant (P > 0.05) antimicrobial effect was observed against L. acidophilus and L. rhamnosus, suggesting the safety of CAMA against commensal lactobacilli.

CAMA, at 1X MIC, exhibited membrane damage in MDR-EAEC strains by flow cytometry; furthermore, CAMA exhibited a moderate percentage of propidium iodide-positive cells (<40%) indicating moderate membrane damage (Figure S1, Supporting Information). In this study, CAMA permeated the inner membrane of all the MDR-EAEC strains in a concentration- and time-dependent increase in the cleavage of ONPG (Fig. 1C). The onset and progress of inner membrane permeabilization of CAMA were quicker than the positive meropenem control. CAMA permeabilized the outer membrane of the MDR-EAEC strains in a concentration and time-dependent manner (Fig. 1C). Also, the onset and progress of the permeabilization by CAMA was much rapid than the positive meropenem control. Moreover, the concentration of nitrocefin permeation for the AMP-treated strains was much higher than the meropenem-treated strains.

In vitro killing kinetic assay of MDR-EAEC with CAMA

In vitro growth kinetics of MDR-EAEC isolates with CAMA was evaluated by incubating the log-phase bacterial cultures (1 × 10⁷ CFU/mL) in CA-MH broth with the peptide at MIC and MBC levels, in triplicates. MDR-EAEC isolates in CA-MH broth were used as untreated control whereas, MDR-EAEC isolates each treated with meropenem (MIC) were used as the positive treatment control.

In groups, I and II, the antimicrobial effect of CAMA was highly significant (P < 0.001) at 30 min of post-inoculation; none of the isolates exhibited visible growth after 30 min post-inoculation. Group III exhibited a highly significant (P < 0.001) reduction at 30 min of co-incubation; none of the isolates exhibited visible growth after 60 min post-inoculation. However, in group IV, all the MDR-EAEC isolates exhibited an increasing growth pattern at 30, 60, 90, 120, 150 and 180 min of incubation (Fig. 2). Since the MIC value of CAMA was observed to be equally effective when compared with the MBC to inhibit the growth of all the three MDR-EAEC strains (Fig. 2A and B), further studies employed the use of MIC levels to investigate the in vivo antimicrobial activity.

Determination of LD₅₀ dose of MDR-EAEC strains in G. mellonella larvae

Concentration-dependent larval mortality was observed upon inoculation of MDR-EAEC strains in G. mellonella (Figure S2, Supporting Information). Based on the pilot survival data, 10⁶ CFU/larvae was determined as the LD₅₀ dose of MDR-EAEC strains in G. mellonella larvae.

Antimicrobial efficacy of CAMA against MDR-EAEC in G. mellonella

The MDR-EAEC infection control larval group exhibited a survival rate of 52.50%, whereas an enhanced survival rate (85%) was exhibited in meropenem- treated larval group (Fig. 3) up to 120 h pi. Nevertheless, a highly significant log-rank Mantel–Cox test (P < 0.001) and a log-rank test for trend (P < 0.05) was appreciable in the infected larval groups treated with CAMA as evidenced by an increased survival rate (92.50%). Further, in the uninfected control groups (CAMA and PBS control) the larvae were healthy with a 100% survival rate up to 120 h pi (Fig. 3).

Melanization assay

MDR-EAEC infection control larval group exhibited a lower melanization rate at 6 h pi, subsequently increased and peaked at 24 h pi. and declined at 48 h pi (Fig. 4A). Nevertheless, the meropenem-treated infected larval group exhibited a lower melanization rate at 6 h pi, attained its peak at 18 h pi and then it reduced gradually in a highly significant (P < 0.001) manner (Fig. 4A). In CAMA-treated larval group, a highly significant (P < 0.001) increase in melanization was observed at 12 to 24 h pi, thereafter, the intensity of melanization was gradually declined (Fig. 4A). Further, at 6 h pi, a slight surge in the melanization intensity was found in the CAMA control larval groups which were retained from 12 to 96 h pi (Fig. 4A).

Figure 4.

In vivo assays using G. mellonella model. Melanization rate (A), MDR-EAEC counts (B), Haemocyte density (C) and LDH cytotoxicity assay (D) of G. mellonella larvae infected with 10⁶ CFU (LD₅₀ dose) of MDR-EAEC isolates and treated with MIC (1X) dose of CAMA 3 h post-infection, keeping respective controls (infection, meropenem-treatment, CAMA control, PBS control). Data expressed as mean ± standard deviation of three independent experiments (*P < 0.05; **P < 0.01 and ***P < 0.001). Melanization rate was assessed by absorbance monitored at 450 nm, MDR-EAEC counts as log₁₀CFU/mL of haemolymph on EMB agar plates supplemented with ampicillin (100 µg/plate), haemocyte density as cells/mL of haemolymph and LDH cytotoxicity assay as the cytotoxicity (%) of larval haemolymph.

Enumeration of MDR-EAEC counts

Though the larval survival could be a suitable parameter to assess the efficacy of antimicrobial agents, haemolymph bacterial burden in larvae serves to be a useful marker to assess the infection dynamics. CAMA-treated infected larval group exhibited a decline in the MDR-EAEC counts in a highly significant manner (P < 0.001) at 24, 48 and 72 h pi. Moreover, MDR-EAEC could not be detected in all the uninoculated larval groups until 96 h pi (Fig. 4B).

Enumeration of haemocytes

Notwithstanding infection and treatment groups, a highly significant (P < 0.001) increase in haemocyte density was observed at 6 h pi, peaked at 12 h and subsequently significant (P < 0.001) reduction was observed (Fig. 4C). Nonetheless, no significant difference in the haemocyte density was observed (P > 0.05) between any of the groups from 72 to 96 h pi (Fig. 4C).

LDH cytotoxicity assay

MDR-EAEC infection control larval group exhibited an increased LDH cytotoxicity in a highly significant (P < 0.001) manner at 6 h pi, peaked at 12–18 h pi that retained up to 96 h pi (Fig. 4D). However, the significant (P < 0.001) increase observed for LDH assay at 6 h pi remained prominent until 48 h pi and declined progressively thereafter in the CAMA- as well as meropenem-treated larval group (Fig. 4D). Whereas, the CAMA control larval groups exhibited a significant increase in the LDH cytotoxicity from 6 to 18 h pi which reduced subsequently (Fig. 4D).

Histopathology

Histopathological examination of G. mellonella would be an ideal approach to divulge the exact host–pathogen interaction subsequent to CAMA treatment in the infected larval groups. No alteration in the histological architecture could be observed in all the larval groups, except for infected control at 6 h and 12 h pi; nevertheless, haemocytes were distributed scanty with minimal melanization at 12–18 h pi.

However, in the infected larval group, the presence of bacterial load around tubular organs, phagocytosis and melanization corresponded with the haemocyte distribution at 12–18 h pi in the sub-cuticular region. Besides, haematoxylin and eosin (H&E) stained larval sections from the infected group at 18 h pi revealed haemocyte clusters as finely stippled blue dots in the sub-cuticular area exhibiting bacterial phagocytosis with clear evidence of melanization. Contrarily, larval cross-sections from CAMA treatment as well as PBS and CAMA control groups looked healthy with individually distributed haemocytes exhibiting no appreciable aggregates or melanization (Fig. 5).

Figure 5.

Histopathological analysis of G. mellonella larvae exhibiting the effect of CAMA on MDR-EAEC using H&E staining at 24 h p.i. The images denote larvae inoculated with PBS (A), uninfected larvae treated with CAMA (B), LD₅₀ dose of MDR-EAEC (C), MDR-EAEC-induced larvae treated with meropenem 3 h pi (D), MDR-EAEC-induced larvae treated with CAMA 3 h p.i (E). Representative images are shown from histological analysis of three larvae for each condition from three independent experiments. Scale bars, 100 µ.

As the infection progressed at 24 h p.i., a comparatively higher cluster of haemocytes was observed in the sub-cuticular area in MDR-EAEC infection control; fat bodies with focal points of melanization and increased bacterial load were also noticed around the tubular organelle (Fig. 5). In CAMA- treated larval group, mild bacterial accumulation was observed around the organelle with a scanty distribution of haemocytes exhibiting no evident aggregates or melanization. In contrast, groups IV and V were found healthy with individually distributed scanty haemocytes exhibiting no noticeable aggregates or melanization. Further, at 48 h p.i., the clustered haemocytes in the sub-cuticular area and fat bodies with evidence of melanization and an increased bacterial load around tubular organelle were observed in the infected control group. In CAMA-treated group, scanty haemocyte distribution with no noticeable aggregates or melanization was observed with a mild accumulation of bacterial load around the organelle. Remarkably, the control larval groups exhibited individually distributed scanty haemocytes with no appreciable aggregates or melanization. Later, at 72 h p.i., mild bacterial accumulation was appreciated in the infected control group with a comparative reduction in melanization and haemocyte aggregates within the fat bodies, while scanty haemocyte distribution with no noticeable aggregates or melanization as well as reduced haemocyte accumulation was noticed in the meropenem-treated group. Interestingly, all the other groups (treatment and control) exhibited scanty haemocytes with no visible melanization.

In short, pronounced histopathological changes observed at 24 h and 48 h pi in MDR-EAEC infected larval groups declined subsequently by 72 h pi. While, at 24 h pi, mild to moderate histopathological changes were evident in CAMA-treated larval groups which were reduced at 48 and 72 h pi. Additionally, no pathological changes could be observed in infected groups treated with CAMA, uninfected control groups and the PBS control group.

DISCUSSION

The insect immune system is supposed to provide more insights about mammalian infections and pathogenesis (Piatek, Sheehan and Kavanagh 2020; Wojda et al. 2020). Owing to its similarities in the innate immune system with humans, the G. mellonella larval system is chosen as an appropriate pre-clinical in vivo model for investigating host–pathogen interaction preceding mammalian models. The haemocytes of larval haemolymph forming an integral part of cellular immunity are involved in phagocytosis and encapsulation, whereas melanization and larval AMPs as a result of phenoloxidase (PO) system play a vital role in the humoral response (Cutuli et al. 2019; Wojda et al. 2020).

Drug resistance among bacterial pathogens of public health importance creates a major obstacle in the pre-clinical and clinical development of therapeutic candidates. With the limited discovery of antibiotics, intensive research has now been directed towards the identification of novel and non-conventional therapeutics (Haney, Straus and Hancock 2019; Petronio Petronio et al. 2020). Among various approaches identified, cationic AMPs have gained considerable momentum due to their antimicrobial, antibiofilm and immunomodulatory functions (Kumar, Kizhakkedathu and Straus 2018; Haney, Straus and Hancock 2019). Besides, AMPs have not significantly been associated with the development of resistance that enables them as suitable candidates for drug-resistant and persistent infections (Ageitos et al. 2017). Nonetheless, extracellular proteins (proteolytic degradation), export by resistance-nodulation-cell division (RND) family of efflux pumps, impedance by exopolymers (alginate, polysialic acid) and biofilm matrix molecules, circumvention of attraction by cell surface or membrane alteration could result in resistance to AMPs (Cheung et al. 2018) viz., LL-37 (Gebhard 2012; Taneja et al. 2013), CRAMP (Lauth et al. 2009), hBD-3 (Gebhard 2012). EAEC, one of the neglected food-borne pathogens, is responsible for chronic as well as persistent diarrhoea that eventually damages the intestinal epithelium of humans and animals (Cabrera-Sosa and Ochoa 2020). Recent resurging trends of AMR, particularly among MDR-EAEC strains and its dissemination has been reported from varied sources that may represent alarming public health threat and therapeutic challenges in resource-limited countries (Lima, Medeiros and Havt 2018). However, in vivo screening of therapeutic compounds in the G. mellonella larval model could be advantageous as it could explore more molecules in less time with minimal quantity. Although the efficacy of various antimicrobial agents has been explored against Carbapenem-resistant Enterobacteriaceae, P. aeruginosa, Klebsiella pneumonia, methicillin-resistant S. aureus in G. mellonella larvae (Betts et al. 2014; Hill, Veli and Coote 2014; Benthall et al. 2015), the efficacy of AMPs as alternative therapeutic candidates have not yet been systematically studied, barring a few studies (do Nascimento Dias et al. 2020; Greber et al. 2020). Conversely, in vivo characterization of the G. mellonella larval model for evaluating the antimicrobial efficacy of short-chain hybrid peptide CAMA against MDR-EAEC strains appears to be the first of its kind.

The short-chain cationic hybrid AMP- ‘CAMA’ with high hydrophobicity residues retrieved from the regularly updated biofilm active AMPs (BaAMPs) database (di Luca et al. 2015) was explored for investigating its antimicrobial activity against MDR-EAEC strains. The proposed mechanism of antimicrobial activity of CAMA was mediated through the bacterial membrane disruption by forming ‘toroidal’ pores and/or detergent-like mechanism (Silva et al. 2018). The LD₅₀ dose of MDR-EAEC strains determined in this study was found to be higher than reported earlier (1.11 × 10⁴ CFU/larvae), which could be due to the strain variation (Jønsson et al. 2017). Moreover, varied pharmacokinetic parameters in comparison to the humans, improved larval bioavailability for antibiotics or the synergistic effect of carbapenems with the natural AMPs circulating in the larval haemolymph could result in a significant survival rate observed in meropenem-treated larvae as reported in earlier studies (Betts et al. 2014; Hill, Veli and Coote 2014; Benthall et al. 2015). Complete survival and lack of melanization were noticed in CAMA control larval groups. The results of survival data indicated that CAMA had either an equal or better efficacy as compared to the antibiotic control, meropenem.

The bacterial counts in the haemolymph of infected larvae treated with CAMA reduced significantly over 24 h and 48 h pi, which could be due to the bactericidal effect of CAMA and/or its intermediates produced during melanization. The melanization in invertebrates is primarily brought about by the activation of the prophenoloxidase (PPO) cascade pathway wherein, the enhanced larval superoxide production leads to the release of host AMPs that could ward off pathogenic microbes. The haemocyte-mediated aggregation and nodulation or phagocytosis of the MDR-EAEC strains would result in bacterial clearance, resulting in larval secretion of AMPs, cell death and melanization (Barnoy et al. 2017; Wojda et al. 2020). During the initial stages of infection (6–18 h pi), the MDR-EAEC stimulated haemocytes might have phagocytosed the bacteria; hence, the findings of haemocyte density correlated with the bacterial enumeration assay, wherein, no significant difference in the MDR-EAEC counts were observed between infection control group as well as treatment (CAMA and meropenem) groups. At a later stage of infection (72–96 h pi), the observed reduction in the circulating haemocytes in all the tested groups could probably be due to the cytotoxic effect of MDR-EAEC strains on the larval cells. This reduction in the density of haemocytes might be credited either to the damage of infected haemocytes and/or its sequestration in the nodules (Hill, Veli and Coote 2014).

The findings of melanization assay correlated well with the haemocyte enumeration assay. MDR- EAEC activated haemocytes secrete AMPs in the insect fat body and triggers the PO cascade which corresponds to the increased melanization rate noticed in the treatment groups (12–48 h pi). Besides, it is imperious from the in vivo results that CAMA offered larval immunomodulation, as the melanization intensity was retained up to 96 h pi in those infected as well as uninfected larval groups treated with CAMA. Likewise, mammalian system also exhibited a similar immunomodulatory effect employing AMPs that suggests the role of AMPs as potential therapeutic candidates (Cutuli et al. 2019; Piatek, Sheehan and Kavanagh 2020). Nonetheless, an increased LDH cytotoxicity was appreciable in those larval groups treated with meropenem similar to ampicillin against P. aeruginosa, in an earlier study (Hill, Veli and Coote 2014).

The sequential events describing host-pathogen interaction, haemocyte distribution and migration of EAEC to different sites are better illustrated by the histopathological examination of the G. mellonella larvae (Barnoy et al. 2017; Piatek, Sheehan and Kavanagh 2020). The haemocyte-directed phagocytosis in the MDR-EAEC infection control group occurred at a faster pace with the haemocytes being directed towards the heart region binding to the cardiac muscle and adjoining organs (fat body and pericardial cells) with continuous bacterial phagocytosis during 24 h and 48 h pi. As a result of the evoked immune response mechanism to eliminate the bacterial infection, these deployed haemocytes might correspond to a decrease in a load of MDR-EAEC, larval melanisation rate and haemocyte density at 72 h pi. Intriguingly, the larval histopathology was found interlinked with the estimation of bacterial burden and immune markers. Hence, it would be judicious to interpret that those factors responsible for microbial survival within this insect larval host were most likely to be directly pertinent to similar infections in mammals (Sheehan and Kavanagh 2018; Piatek, Sheehan and Kavanagh 2020).

CONCLUSION

This study exploits the G. mellonella larval model for evaluating the antimicrobial efficacy of short-chain hybrid peptide CAMA against MDR-EAEC strains. In vivo analysis revealed a significant reduction in the bacterial counts both at 24 h and 48 h pi in CAMA-treated larval groups in comparison to the MDR-EAEC infection control. Besides, CAMA-treatment exhibited enhanced larval survivability, immunomodulation and further proved non-toxic to host larval cells. CAMA exhibited either an equal or better efficacy than the tested antibiotic control, meropenem. The findings of the present study highlight the possibility of the pre-mammalian G. mellonella larval model in investigating host–pathogen interaction and the efficacy of antimicrobial agents against MDR-EAEC strains. Further, in vivo studies and clinical trials employing ethical vertebrate models (mice/piglets) and targeted drug-delivery systems are warranted to translate the peptide to practice.

AUTHORS’ CONTRIBUTION

DBR, SVS and SBB contributed conception and design of this study; JV and RP organized and performed the experiments; MK performed the statistical analysis; JV and MK wrote the first draft of the manuscript; JV, RP, MK and NVK wrote sections of the manuscript; DBR, SVS, NVK and SBB edited the manuscript. All authors contributed to manuscript revision, read and approved the submitted version.

ACKNOWLEDGEMENTS

The authors thank the Director, ICAR-Indian Veterinary Research Institute, Izatnagar, India for providing facilities for the research. The authors are thankful to Dr Chobi Debroy and Dr Bhushan Jayarao, Pennsylvania State University, State College, USA for providing EAEC DNA. We thank Dr Ajay Kumar, Scientist and Dr (Mrs) Meeta Saxena, Senior Technical Officer, Division of Biochemistry for their expertise for flow cytometry. We also thank Dr Indira Devi, Director (Research), and Dr Gavas Ragesh, Assistant Professor, Kerala Agricultural University, Thrissur, Kerala for providing G. mellonella larvae. The technical assistance provided by Mr K.K. Bhatt and Dr Deepa Ujjawal are duly acknowledged.

FUNDING

The research work was supported by grants received from CAAST-ACLH (NAHEP/CAAST/2018–19) of ICAR-World Bank-funded National Agricultural Higher Education Project (NAHEP).

Conflicts of Interest

None declared.

Author notes

Present Address: Department of Veterinary Public Health, College of Veterinary and Animal Sciences, Pookode, KVASU, Wayanad- 673 576, India.