Characterization of the immune induced antimicrobial peptide in Drosophila melanogaster and Drosophila ananassae

Insects can recognize invading pathogens and initiate an immune response. Among them, Drosophila has emerged as an invertebrate model for investigating innate immune responses in which antimicrobial peptides play a crucial role. In the present study, immune-induced antimicrobial peptides were characterized in D. melanogaster and D. ananassae using the agar well diffusion method, HPLC, SDS-PAGE and LC-MS/MS after infection with either S. aureus or E. coli. The HPLC revealed two and three differentially induced components, respectively, in D. melanogaster and D. ananassae fl ies infected with S. aureus and E. coli. The tricine SDS-PAGE analysis also revealed two and fi ve differentially induced proteins, respectively, in D. melanogaster and D. ananassae infected with E. coli. In E. coli infected fl ies, the ~6 kDa band was produced at higher level. Based on LCMS/MS and Mascot analysis, the peptide was identifi ed as a putative cecropin A-like peptide, and the data suggested that both species of Drosophila have exhibited a clear immune response. The fl ies were also able to discriminate between bacteria, as this putative cecropin A-like peptide was produced in fl ies infected with E. coli but not S. aureus. * Corresponding author; e-mail: knagarajv@gmail.com INTRODUCTION The fruit fl y, Drosophila melanogaster has innate immunity against invading microbes. This includes both cellular and humoral immune responses (Lye, 2018; Meghashree & Nagaraj, 2020). Antimicrobial peptides (AMPs) are an important component in the fi rst line of defence (Yuchen et al., 2019). AMPs are endogenous peptides with a molecular weight (MW) of ~2–22 kDa and they are released by the fat body (analogue of the liver) into haemolymph to clear off the microbial infections (Troha et al., 2019). The interactions of AMPs with Gram-positive and Gramnegative bacteria differ. The positively charged AMPs selectively interact with prokaryotes having a negatively charged bacterial cell-wall, including lipopolysaccharides (LPS) and phospholipids. Based on the available data (FlyBase), nine distinct classes of AMPs (23 members) are identifi ed in Drosophila (Thurmond et al., 2019). Among them, attacin, diptericin, cecropin and drosocin are produced in response to Gram-negative bacterial infections (Imd pathway), metchnikowin and defensin in response to a Gram-positive bacterial infection (Toll pathway) and drosomycin only in response to fungal infection (Sheehan et al., 2018). As fruit fl ies are genetically similar in the way they combat diseases as humans, they can be used to evaluate miEur. J. Entomol. 118: 355–363, 2021 doi: 10.14411/eje.2021.037


INTRODUCTION
The fruit fl y, Drosophila melanogaster has innate immunity against invading microbes. This includes both cellular and humoral immune responses (Lye, 2018;Meghashree & Nagaraj, 2020). Antimicrobial peptides (AMPs) are an important component in the fi rst line of defence (Yuchen et al., 2019). AMPs are endogenous peptides with a molecular weight (MW) of ~2-22 kDa and they are released by the fat body (analogue of the liver) into haemolymph to clear off the microbial infections (Troha et al., 2019). The interactions of AMPs with Gram-positive and Gramnegative bacteria differ. The positively charged AMPs selectively interact with prokaryotes having a negatively charged bacterial cell-wall, including lipopolysaccharides (LPS) and phospholipids. Based on the available data (Fly-Base), nine distinct classes of AMPs (23 members) are identifi ed in Drosophila (Thurmond et al., 2019). Among them, attacin, diptericin, cecropin and drosocin are produced in response to Gram-negative bacterial infections (Imd pathway), metchnikowin and defensin in response to a Gram-positive bacterial infection (Toll pathway) and drosomycin only in response to fungal infection (Sheehan et al., 2018).
As fruit fl ies are genetically similar in the way they combat diseases as humans, they can be used to evaluate mi-

Protein quantifi cation
The protein concentration in the haemolymph from control and infected (24 h) fl ies was quantifi ed in fi ve independent experiments (n = 50) by direct concentration measurement in which the haemolymph was diluted 1 : 5 with double distilled water. A 2 μL of diluted haemolymph was placed on a μDrop™ plate and absorbance was measured at 280 nm using Multiskan Sky spectrophotometer (Thermo Scientifi c™, USA). Protein concentration was determined based on this instrument's built-in protocol for the extinction coeffi cient of BSA.

In vitro antibacterial activity determined using the agar well diffusion method
The freeze-dried homogenate from Drosophila spp. obtained 24 h after injection with either S. aureus or E. coli or PBS (Control) were examined for antibacterial activity against the respective bacteria using the agar well diffusion method (Sewify et al., 2017). In brief, 5-mm diameter wells were made with a sterile cork borer (6 mm diameter) in nutrient agar plates spread with S. aureus or E. coli. The lyophilized homogenate (5 mg) was suspended in 50 μL PBS. Ciprofl oxacin (100 μg/mL) was used as a standard. After incubation for 24 h at 37°C, the zone of inhibition was measured in terms of its diameter in mm.

HPLC analysis
The haemolymph from control and infected fl ies (24 h) was subjected to HPLC analysis. The lyophilized sample was dissolved in 0.1% TFA solution at a concentration of 25 mg/mL. A 20 μL of the crude sample was injected using a glass syringe into a C 18 reverse-phase analytical column (5 μm particle size; 250 × 4.6 mm column) placed over an HPLC (Shimadzu). The solvent system included 0.1% TFA in Milli Q water (Solvent-A) and 80% aqueous Acetonitrile (ACN) with 0.1% TFA (Solvent-B) and the fl ow rate for the mobile phase was set at 1 mL/min. The elution was carried out with a linear gradient of 5-95% of solvent-B over a 60 min period. The eluted peaks were detected at 214 nm using an UV-DAD detector (SPD-M20A).

SDS-PAGE analysis
The profi le of proteins in haemolymph was assessed by both Tris glycine (Laemmli, 1970) and tricine SDS-PAGE techniques (Schägger, 2006). A 30 μg of haemolymph from control and infected fl ies (24 h) were placed in wells. For Tris-glycine, 4% stacking and 12% resolving gel were used and ran at 150 V for 75 min. For tris-tricine SDS-PAGE, 4% stacking, 10% spacer and 16% resolving gel were used and ran at 150 V for 195 min. To determine the MW, high (11-245 kDa; Himedia) and low range protein markers (3-45 kDa; SRL) were used. After electrophoresis, separated protein bands were detected by silver staining method (Gromova & Celis, 2006). The density and number of bands were determined using Gelanalyzer software, version 19.1. The gel band of interest was cut out and placed in 1% acetic acid solution until required.

In-gel digestion
The piece of gel band was washed with 500 μL of wash solution (50% acetonitrile, 50 mM ammonium bicarbonate) and vortexed for 15 min at RT until it became opaque and stuck together. Then the gel band was spun down and the supernatant removed. A 3 mL of the Dithiothreitol solution was added to completely cover the piece of agar and incubated for 30 min at 56°C in an air thermostat. A 5 mL of Acetonitrile (ACN) was added, incubated for 10 min at RT and then the liquid was removed. A 3 mL of the iodoacetamide solution was added and incubated for 20 min at 2003), which are highly conserved in insects. There are also many research studies focused on identifying novel cecropin-like peptides in insects (Wu et al., 2015;Park & Yoe, 2017;Manniello et al., 2021).
Hence, the objective of this study was to determine whether the induction of immune-induced AMP in D. melanogaster and D. ananassae fl ies infected with either E. coli or S. aureus differed. To authenticate its antimicrobial effi cacy, homogenates of fl ies were tested for antibacterial activity in vitro. In addition, the phylogenetic relationships of the identifi ed peptide are also discussed.

Fly stocks
D. melanogaster (1.002) and D. ananassae (11.001) were reared on an instant Drosophila diet supplemented with yeast and kept at room temperature (RT) under 12L : 12D conditions. The fl ies were obtained from the Drosophila Stock Center, University of Mysore, Mysore, Karnataka, India. For all experiments, 4-5 day old male and female adult fl ies (1 : 1) were used.

Bacterial species
Escherichia coli (MTCC 723) and Staphylococcus aureus (MTCC 7443) were obtained from MTCC, Chandigarh. All bacterial cultures were maintained on a nutrient agar medium. For liquid culture, bacteria were grown in sterile tubes containing 5 mL of nutrient broth (beef extract -3g/L; peptone -0.5g/L; NaCl -0.5g/L), which was incubated for 24 h at 37°C before use. An optical density of 0.5 (OD 600 ) having 1 × 10 7 CFU/mL was used as an infectious dose and was obtained using a spectrophotometer (Multiskan Sky, Thermo scientifi c).

Bacterial infection
Flies were anesthetized and infected by inserting a tungsten needle into the lateral side of the thorax that had been dipped into either S. aureus or E. coli suspended in phosphate buffer saline (PBS). The treated fl ies were kept at RT by placing each of them in a fresh vial, laying the vial on its side until all fl ies recovered from the anesthesia in order to avoid the fl ies from becoming stuck in the food (Khalil et al., 2015). For the control group, fl ies were pricked with PBS dipped needle to create a non-septic injury.

Preparation of crude extract and isolation of haemolymph
Infected fl ies (n = 50) were homogenized in 120 μL 0.1% trifl uoroacetic acid (TFA) at an ice-cold condition (Bhagavathula et al., 2017). The disrupted homogenate was further sonicated (QSonica 125, Thermo Scientifi c) at 20% amplitude for 5 cycles at an interval of 5 s. A 100 μL supernatant of each extract was collected after centrifugation at 10000 × g for 15 min at 4°C, freeze-dried using a lyophilizer (FreeZone, Labconco) for 10 h and stored at -80°C until further use. These lyophilized samples were used for HPLC analysis and antibacterial activity.
The haemolymph was collected by means of centrifugation (Dhar & Mishra, 2020) in which each fl y after 24 h of bacterial infection was pricked with a needle to release the haemolymph. A 0.5 mL vial was punched with 4-5 tiny holes using a 24G syringe needle and all the pricked fl ies (n = 50) were added to it. This vial was put inside a 1.5 mL microcentrifuge tube from which haemolymph was collected after centrifugation for 10 min at 2000 × g and stored at -80°C (Damrau et al., 2015). From each vial, 2 μL of haemolymph was extracted and freshly isolated samples were used for protein quantifi cation and SDS-PAGE analysis.
RT in the dark. The piece of agar was placed in acetonitrile and then centrifuged to remove all the liquid. Trypsin buffer (13 ng/ μL) was added until the gel band was covered and then kept in an ice bucket for about 90 min. A 1 μL of ammonium bicarbonate buffer (100 mM) was added to cover the gel band and keep them wet during enzymatic cleavage. 5% formic acid/ACN (1 : 2) was added and incubated for 15 min. After centrifugation, the supernatant was transferred to a new vial. Then the samples were lyophilized and dissolved in 20 μL of 2% acetonitrile/0.1% formic acid solution.

LC-MS/MS analysis
The sample was analysed using an ultra-high-performance LC with mass selective detection and an Ultimate 3000 series LC (Dionex, USA) coupled with ESI tandem mass spectrometer (micrOTOF-Q II) (Bruker, Germany). A 3 μL of sample was injected into LC precolumn (Pep map TM 100; 75 μm × 2 cm; Nanoviper C18, 3 μm; 100Å) and LC analytical column (EASY SPRAY PEPMAP RSLC C18 3 μm; 50 cm × 75 μm; 100Å) of an EASY-nLC 1200 LC instrument. Mobile phase A of 0.1% Formic acid in HPLC water and mobile phase B of 0.1% formic acid in acetonitrile was used. A linear gradient starting from 5% to 95% in 60 min with an 0.2 mL/min fl ow rate was recorded. The MS scan was carried within the 200-1800 m/z range and the data acquired in MS/MS (auto) scanning mode.

Mascot Search
The data analysis was carried out using MASCOT search engine. In MS/MS ions search, the SwissProt database was used with all entries option for Taxonomy. The other parameters used included trypsin as a proteolytic enzyme, Cysteine carbamidomethylation as fi xed modifi cation, methionine oxidation as variable modifi cation, the error window for peptide mass was 10 ppm and fragment ion mass 0.6 Da. The decoy database was selected to calculate the false discovery rate. Only top rank peptide hits for given precursors were used for further protein identifi cations.

Phylogenetic analysis
The phylogenetic tree was constructed for seven cecropins: cecropin A (64 aa; P01507), A1 (63 aa; C0HKQ7), A2 (63 aa; C0HKQ8), B (63 aa; P14956) and C (63 aa; O16829), cecropin-2 (63 aa; XP_001955554.1) and sarcotoxin-1C (63 aa; XP_001955556.1) using the Maximum Likelihood method and JTT matrix-based model (Jones et al., 1992). The bootstrap consensus tree inferred from 1000 replicates was taken to represent the evolutionary history of the cecropins analysed. The percentage of replicate trees in which the associated cecropins clustered together in the bootstrap test was shown next to the branches (Felsenstein, 1985). Initial tree(s) for the heuristic search were obtained automatically by applying Neighbour-Joining and BioNJ algorithms to a matrix of pairwise distances estimated using the JTT model and then selecting the topology with the highest loglikelihood value. The evolutionary analyses were conducted in MEGA X v10.1.8 (Kumar et al., 2018).

Statistical analysis
The data were analysed using a two-way ANOVA for the protein concentration and antimicrobial activity with Bonferroni posthoc test using Graph pad Prism software 5.0. All the values are means ± SEM. Values were considered signifi cant when *P < 0.05, **P < 0.01 and ***P < 0.001.

The concentration of protein in haemolymph was greater in infected fl ies
The average total concentration of protein in haemolymph based on fi ve independent experiments was found to be signifi cantly higher in both D. melanogaster and D. ananassae 24 h after infection with E. coli or S. aureus (Fig. 1). In D. melanogaster, infection with E. coli resulted in a signifi cantly greater protein production (82.1 mg/mL; P < 0.01) than infection with S. aureus (69.4 mg/ mL). However, in D. ananassae protein concentration was higher after infection with S. aureus (114 mg/mL) than E. coli (104 mg/mL). In addition, there is a very signifi cant difference in total protein concentration in the two species Drosophila following bacterial infections.

The Drosophila homogenate inhibits the growth of bacteria
The freeze-dried homogenate of bacteria-infected fl ies inhibited the growth of bacteria differently as seen in the zone of inhibition against both S. aureus and E. coli compared with standard ciprofl oxacin. An inhibition zone was not observed in PBS control fl ies. However, the degree of inhibition of bacteria is different. The E. coli infected D. melanogaster and D. ananassae homogenates resulted in a larger zone of inhibition than that of the S. aureus infected fl ies (Table 1 and Fig. S1).

HPLC profi le showed differential expression of immune induced molecules
The HPLC profi le of haemolymph showed that D. melanogaster and D. ananassae infected with either S. aureus or E. coli had two and three differentially produced molecules, respectively. This shows that there is a clear quantitative difference in the expression of AMPs. In addition, the elution time indicated (within 30 min) that these molecules are possibly polar (Fig. 2).

Detection of a cecropin A-like peptide in E. coli infected fl ies using SDS-PAGE and LC-MS/MS analysis
The SDS-PAGE analysis of haemolymph protein using the Tris-glycine method revealed several electrophoretic bands with MW ranging from ~11-242 kDa, based on a densitometry analysis. Several proteins are up and downregulated during infection with either E. coli or S. aureus. Among them, three (58, 34, and 13 kDa) and fi ve proteins (45, 33, 27, 14, and 11 kDa) were markedly produced in D. melanogaster and D. ananassae, respectively (Fig. 3).
As the Tris-glycine method doesn't resolve the low MW peptides, haemolymph samples from D. melanogaster and D. ananassae infected with S. aureus or E. coli were separated using the tris-tricine method. The data showed that a single protein band of ~6 kDa (based on the retention factor calculated using gelanalyzer software) was differentially produced in both D. melanogaster and D. ananassae injected with E. coli, but not with S. aureus or PBS (Fig.  4A). In addition, one protein band in D. melanogaster and four protein bands within 14 kDa, were highly produced in D. ananassae infected with E. coli (Fig. 4B) and these protein bands were not recorded in the two species of fl ies infected with S. aureus. The data also showed that the control groups of these two fl ies have distinct protein profi les. The higher molecular weight protein bands were also differentially produced in the E. coli infected fl ies. Here,   D. melanogaster and D. ananassae infected with E. coli showed three and two highly produced proteins, respectively, compared to PBS injected fl ies (Fig. 5). The Mascot search of the LC-MS/MS analysis of a protein band (~6 kDa) from both E. coli infected D. melanogaster and D. ananassae gave a fragment ion with sequence AG-PAVAVVGQATQIAK, which is similar to cecropin A of 6952 Da with a protein sequence coverage of 26% (Table  2 and

Phylogenetic analysis of the putative cecropin A-like peptide sequence
Cecropins occurred in insects before the divergence of Diptera, Lepidoptera and Coleoptera. The phylogenetic tree (Fig. 7A) indicate that Drosophila cecropins are present in one branch and Hyalophora's cecropin in another branch. The result of a multiple sequence alignment and phylogenetic analysis indicate that the putative cecropin A-like peptide has an approximately 11.7% sequence similarity with other Drosophila cecropins (Fig. 7B). The result also support the independent evolution of the cecropin peptide family in these insects (Tassanakajon et al., 2015).

DISCUSSION
S tudies on the immune system of insects help in understanding the complexity of the immune system and the vital role of AMPs in insects' innate immunity. Drosophila has been a suitable model for studying the role of antimicrobial peptides in neutralizing circulating pathogens. The measurement of protein concentration in complex mixtures other than cell lysates can be better assessed at 280 nm. The increased total protein level recorded could be due to bacterial infection resulting in a higher metabolic rate in these fl ies. The type of signal generated by E. coli could be stronger and have resulted in a higher metabolic rate and hence increased total protein level in E. coli infected fl ies than in S. aureus infected fl ies.
The agar well diffusion method is commonly used for screening the antibacterial activity of AMPs. Here, a putative cecropin A-like peptide produced due to bacterial infection could be the reason for the antimicrobial activity recorded against bacterial pathogens. In a recent study (Park & Yoe, 2017), the minimum inhibitory and bactericidal   concentrations evaluated for a cecropin-like peptide has higher antibacterial effects against Gram-negative bacteria. In the S. aureus infection, though a cecropin A-like peptide was not recorded, there were other immune-induced proteins with higher MWs, which could also account for the in vitro antimicrobial activity recorded in these infections. In addition, the HPLC profi le of haemolymph also confi rms the presence of immune-induced components in fl ies infected with S. aureus. Drosophila is considered to be good a model for understanding the variability in conserved genes expressed in closely related species (Hodgins-Davis et al., 2009). The tricine SDS-PAGE method has been less used for identifying immune-induced AMPs in Drosophila, but as shown here, can clearly be used to identify low MW peptides. The production of a specifi c AMP against E. coli by both fl ies indicates that inducible immune genes may have been conserved in these two fl ies (Hanson et al., 2016). AMPs are not produced in uninfected or PBS-pricked fl ies (Li et al., 2019;Feng et al., 2021;Kapila et al., 2021). They are not constitutively expressed in Drosophila haemolymph. Further, the level of induced immunity was stronger against E. coli than S. aureus. This is in accordance with earlier reports in which several AMPs are elicited against Gram-negative bacterial infections, but not non-fl agellated Grampositive bacterial infections (Lemaitre et al., 1997). In another report, cecropin is strongly expressed in Drosophila cell lines by bacterial lipopolysaccharide and fl agellin, but weakly by peptidoglycan (Samakovlis et al., 1992). This confi rms that E. coli can elicit a stronger immune response than S. aureus in Drosophila spp. Thus, as in the mammalian immune system, where different pattern recognition receptors are involved in the identifi cation of lipopolysaccharide (TLR-4) and peptidoglycan (TLR-2), Drosophila might be able to discriminate the different compositions of the membranes of Gram-positive and Gram-negative bacteria (Takeuchi et al., 1999).
Among several antimicrobial peptides of Drosophila origin, cecropin A, a 4.3 kDa peptide (active form), which was fi rst isolated from the haemolymph of the Lepidopteran H. cecropia (Mylonakis et al., 2016) is mainly expressed during Gram-negative bacterial infections (Wen et al., 2019), as it is an α-helical antimicrobial peptide that mainly kills Gram-negative bacteria (Fu et al., 2004). Diptericin (9 kDa) and drosocin (2.19 kDa) are also produced in response to Gram-negative bacterial infections. In this study, the putative cecropin A-like peptide expression was recorded only in haemolymph from Drosophila infected with E. coli.
We have shown marked similarities and some notable differences in the immune responses to bacterial infection by D. melanogaster and D. ananassae. Though the protein profi les were different in these two species, the AMP was recorded in both species in response to infection with E. coli, but not S. aureus. Previous research also indicates that many of the differentially expressed genes in D. melanogaster during the parasitoid-specifi c immune response have similar transcriptional responses in other closely related species of Drosophila. The gene expression profi les in D. melanogaster and D. simulans are very similar (Salazar-Jaramillo et al., 2017). In addition, 83% of proteincoding genes in D. ananassae are homologous to those in D. melanogaster (Uniprot, 2021a), which could explain the similar protein patterns recorded in SDS gels of the two species studied.
In this study, the LC-MS/MS-based mascot analysis detected a peptide with 16 amino acid sequences (AGPA-VAVVGQATQIAK), which is similar to the cecropin-A from H. cecropia, with a good Mascot score and sequence coverage. However, the identifi ed putative cecropin A-like peptide has only 37-41% similarity with other Drosophila cecropins (Uniprot, 2021b). Hence, the present fi nding of a cecropin A-like peptide is an addition to the pool of already known cecropins, produced by Lepidoptera like H. cecropia, Bombyx mori, Dipteran Musca domestica and Coleop-teran Acalolepta luxuriosa, etc. In addition, as a part of the evolutionary link, the Lepidopteran cecropin is similar to those reported for Diptera. For example, the cecropin from Brachycera (Diptera) is closely related to the Lepidopteran cecropin (Brady et al., 2019) and the mosquito cecropin (Diptera) is more similar to the B. mori cecropin-D (Lowenberger et al., 1999). Notwithstanding, that the remaining unidentifi ed region of this putative cecropin A-like peptide may have a similar or different amino acid sequence to the Hyalophora's cecropin. There are many other cecropin like peptides, such as sarcotoxin-I (Buonocore et al., 2021), papiliocin (Kim et al., 2010), stomoxyn (Boulanger et al., 2002bLandon et al., 2006) and hinnavin (Yoe et al., 2006), etc.
Further genomic analysis could validate this peptide as either cecropin A or its ortholog in Drosophila. The detection of a particular cecropin in several species of Diptera, like Drosophila, is not uncommon. Cecropin B is produced by both D. sechellia and D. simulans, and cecropin C by D. takahashii, D. simulans and D. sechellia (Uniprot, 2021b). Similarly, in this study, a putative cecropin A-like peptide was detected in both D. melanogaster and D. ananassae. This data support the tree topology as the obtained sequence is most similar to H. cecropia. In other words, each sequence is more similar to its ortholog in another species than to other members in the same species. That cecropin-A1, A2, B, C are conserved in D. melanogaster indicates that both paralogs are from duplication of a cecropin-like peptide ancestor. The homology of the sequences of cecropin A (H. cecropia) and cecropins (A 1 , A 2 , B, and C) from D. melanogaster, indicate that the detected peptide region of cecropin A has two amino acids similar to known cecropins of D. melanogaster and D. ananassae. This study has shown how different bacterial infections generate distinct immune responses in different genetic backgrounds. This study is novel as there are no reports on the evaluation of immune responses in D. ananassae and this is the only report of the production of putative cecropin A-like peptide in Drosophila spp.

CONCLUSION
This study has shown there is an antibacterial immune response in D. melanogaster and D. ananassae against S. aureus and E. coli. The production of a putative cecropin A-like peptide in both species of Drosophila is reported here for the fi rst time against infection with E. coli, but not against S. aureus. The role of this peptide in innate immunity with comparison to other paralogues needs to be investigated.