The soil collection procedures in Mae Wong Nation Park were approved and permitted by the Department of National Parks, Wildlife and Plant Conservation, Thailand (Permission number 0907.4/7432). A total of 550 soil samples were randomly collected from 110 different sites in Mae Wong National Park, Kamphaeng Phet Province, Thailand. Soil samples were collected between August 2014 and July 2015. The process for soil collection followed Thanwisai et al. (2012). Site locations, soil temperatures, pH, and moisture were recorded. The samples were transported to the Department of Microbiology and Parasitology, Faculty of Medical Science, Naresuan University, Phitsanulok Province, Thailand for isolation of EPNs.

Entomopathogenic nematodes were obtained from soil samples using the baiting technique with Galleria mellonella larvae (Bedding and Akhurst, 1975). After 5 days as bait, the dead G. mellonella were removed from the soil and were subsequently placed on White traps (White, 1927), in which the emerging IJs were then collected and stored at -20°C for further DNA extraction. The IJs (200–300 nematodes) from each isolate were used for DNA extraction as described (Thanwisai et al., 2012). To identify species of EPNs, polymerase chain reaction (PCR) was performed using the primers TW81_F (5′-GTT TCC GTA GGT GAA CCT GC-3′) and AB28_R (5′-ATA TGC TTA AGT TCA GCG GGT-3′) to amplify a region of the internal transcribed spacer locus (ITS) for Heterorhabditis as well as 539_F (5′- GGA TTT CCT TAG TAA CTG CGA GTG-3′) and 535_R (5′-TAG TCT TTC GCC CCT ATA CCC TT-3′) to amplify a region of 28S rDNA for Steinernema (Stock et al., 2001). The PCR components (30 μl) consisted of 7.5 μl DNA extracted solution, 0.6 μl dNTPs (200 μM), 1.2 μl each primer (5 μM), 4.2 μl MgCl2 (25 mM), 3 μl 10X buffer, 0.3 μl 5 U/ml Tag polymerase, and 12 μl distilled water. The cycling conditions for ITS were as follows: 1 cycle of 95°C for 5 min followed by 35 cycles of 94°C for 1 min, 50°C for 30 s, 72°C for 1 min, and a final extension at 72°C for 7 min. The cycling conditions for 28S rDNA were as follows: 1 cycle of 95°C for 5 min followed by 35 cycles of 94°C for 1 min, 55°C for 30 s, 72°C for 45 s, and a final extension at 72°C for 7 min. Both reactions were performed in the Applied Biosystems thermal cycler (Applied BiosystemsTM VeritiTM thermal cycler, Pittsburgh, PA, United States). PCR products were checked by electrophoresis over 30 min, 100 V on a 1.2% TBE-buffered agarose gel stained with ethidium bromide. PCR products were purified using a NucleoSpin Gel and PCR Clean-up column (Macerey-Nagel Ltd., Düren, Germany) as recommended by the manufacturer. The purified PCR products were sequenced at Macrogen Inc. Service (Seoul, Korea).

Photorhabdus and Xenorhabdus were isolated from the hemolymph of the dead G. mellonella infected with EPNs. Hemolymph was streaked on nutrient bromothymol blue triphenyltetrazolium chloride agar (NBTA) and incubated at room temperature in the dark for 4 days. Green and blue colonies were sub-cultured for further genomic DNA extractions. A single colony was then transferred to a 15 ml centrifuge tube containing 3 ml Luria-Bertani (LB) broth and shaken at 180 rpm for 18–24 h. Genomic DNA of each bacterial isolate was extracted from 3 ml culture using a genomic DNA mini kit (Blood/culture cell; Geneaid Biotech Ltd., New Taipei, Taiwan) as recommend by the manufacturer.

To identify Xenorhabdus and Photorhabdus, the Recombinase A gene (recA) was amplified using primers recA_F (5′-GCT ATT GAT GAA AAT AAA CA-3′) and recA_R (5′-RAT TTT RTC WCC RTT RTA GCT-3′) (Tailliez et al., 2010). The PCR reagents was mixed in a final volume of 30 μl, which consisted of 3 μl DNA extract, 0.6 μl dNTPs (200 μM), 1.2 μl each primer (5 μM), 4.2 μl MgCl₂ (25 mM), 3 μl 10X reaction buffer, and 0.3 μl 5 U/ml Taq polymerase plus 12 μl distilled water. The cycling conditions for Xenorhabdus were as follows: 1 cycle of 94°C for 5 min followed by 30 cycles of 94°C for 1 min, 50°C for 1 min, 72°C for 2 min, and a final extension at 72°C for 7 min. The cycling conditions for Photorhabdus were as follows: 1 cycle of 94°C for 5 min followed by 30 cycles of 94°C for 1 min, 50°C for 45 s, 72°C for 1.30 min, and a final extension at 72°C for 7 min. PCR product analysis, purification and sequencing were performed as for EPN characterization.

Sequence assembly and editing were performed using SeqmanII DNASTAR Inc., (Madison, WI, United States). Sequences were aligned using Clustal W (Thompson et al., 1994) and compared with sequences from other known species in GenBank using the BLASTN algorithm. Phylogenetic analyses of sequence data were performed using MEGA 6.0 (Tamura et al., 2013). A bootstrap consensus tree was inferred from 1000 replicates.

We used 12 strains of drug-resistant bacteria including Acinetobacter baumannii (three clinical strains), Escherichia coli (one clinical strain), E. coli ATCC^® 35218 (one strain), Klebsiella pneumoniae ATCC^® 700603 (one strain), Enterococcus faecalis ATCC^® 51299 (one strain), Pseudomonas aeruginosa (one clinical strain), P. aeruginosa ATCC^® 27853 (one strain), Staphylococcus aureus (two clinical isolates), and S. aureus ATCC^® 20475 (one strain) for screening and assessment using the disk diffusion method. The clinical isolates of A. baumannii, E. coli, P. aeruginosa, and S. aureus were obtained from patients with positive samples processed by the diagnostic microbiology laboratory at a hospital in Thailand between February 2011 and December 2014 (Supplementary Table S1). Speciation of bacteria were determined using Standard Operating Procedures supplied by the Department of Medical Science, Ministry of Public Heath, Thailand and Clinical and Laboratory Standards Institute (CLSI) guidelines (M100-S24 and M100-S25), respectively. Each bacterial strain was streaked on Mueller Hinton agar (MHA) and incubated at 37°C for 24 h. A single colony was picked, dissolved in phosphate buffered saline solution, and the concentration adjusted to 0.08–0.13 (0.5 McFarland standard) (Seier-Petersen et al., 2014) according to spectrophotometric measurement at a wavelength of 600 nm.

We utilized 24 isolates of symbiotic bacteria including 20 isolates of P. luminescens subsp. akhurstii, 2 isolates of X. stockiae, 1 isolate of X. Japonica, and 1 isolate of P. temperata subsp. temperata for screening. A bacterial stock of each isolate was streaked on NBTA medium and incubated at room temperature in the dark for 4 days. A single colony was transferred and grown in 25 ml LB broth and shaken at 180 rpm for 24 h. The concentration of the whole cell suspension was adjusted to 10⁷–10⁸ CFU/ml using phosphate buffered saline solution. To prepare cell-free supernatant, 1 ml whole cell suspension was centrifuged at 12,074 g (Centurion, United States) for 5 min. The supernatant was then filtered using a 0.22-μm filter. The flow-through was used as the cell-free supernatant (Bussaman et al., 2012).

To screen Xenorhabdus (3 isolates) and Photorhabdus (21 isolates) against drug resistant bacteria, 20 μl whole cell suspension and of cell-free supernatant were dropped onto a sterile 6-mm paper disk. The disk was then allowed to dry for 3–5 min and transferred onto the surface of MHA plates that had been plated with 12 strains of drug resistant bacteria. Antibiotic disks were used as positive controls. The MHA plates were incubated at 37°C for 24 h. The diameter of the inhibition zone was measured in millimeter units.

Xenorhabdus (two isolates) and Photorhabdus (nine isolates) that could inhibit the growth of at least one strain of drug resistant bacteria during screening were then subcultured on NBTA medium for further extraction of crude compounds. The plates were incubated at room temperature in the dark for 4 days. A single colony of each isolate was transferred and grown by shaking in a 1000-ml flask containing 500 ml LB at 180 rpm for 72 h. To extract the crude bioactive compound, 1000 ml ethyl acetate was added to the culture and mixed well. The flask was then allowed to stand at room temperature for 24 h. The extraction from each isolate was performed three times to maximize the level of crude compound. All bacterial extracts were concentrated using a rotary vacuum evaporator (Buchi, Flawil, Switzerland). The condensed extracts of all bacterial isolates were weighted and stored at -20°C until used.

Bacterial extracts from each isolate were dissolved in 1 ml dimethyl sulfoxide (DMSO) Then, 10 μl solution was dropped onto a sterile 6-mm disk, which was placed on the MHA plated with drug resistant bacteria. The plates were then incubated at 37°C for 24 h. Antibiotic disks and DMSO were used as a positive and negative control, respectively. The diameter in millimeter of the clear zone was measured using a ruler.

The MIC of nine bacterial extracts was determined using the broth twofold serial dilution method in 96-well micro titer plates. Initial concentrations of all P. luminescens and P. temperata extracts were dissolved in DMSO to generate 500 and 220 mg/ml solutions, respectively. Then, 100 μl of each extract was mixed with 100 μl MH broth in a well of the micro titer plates followed by twofold serial dilution. To each well, 100 μl bacterial suspension (1 × 10⁸ cell/ml) was added and mixed well. Plates were incubated at 37°C for 24 h. Turbidity of each well was observed visually. The clear wells in the micro titer plates with the lowest concentration from each extract were considered as representing the MIC.

To determine the minimum bactericidal concentration (MBC), 10 μl of the bacterial suspension from micro titer plates from the MIC experiment was dropped onto MHA in triplicate to observe viability following aerobic incubation at 37°C for 24 h. The MBC was read by determining the lowest concentration of the bacterial extracts that reduced the viability of the initial bacterial inoculum by ≥99.9%.

For the analysis of natural products present in the ethyl acetate extract, the extracts were evaporated to dryness and dissolved in a 1/10 culture volume of methanol. HPLC-MS analysis was performed using a Dionex Ultimate 3000 system coupled to a Bruker AmaZon X mass spectrometer and an Acquity UPLC BEH C18 1.7 μm RP column (Waters) with an acetonitrile (0.1% formic acid) in H₂O (0.1% formic acid) gradient ranging from 5 to 95% over 16 min at a flow rate of 0.4 ml/min at 40°C. Chromatograms were analyzed using Bruker Compass Data Analysis 4.3.

A total of 550 soil samples collected from 110 locations in Mae Wong National Park, Kamphaeng Phet Province, Thailand yielded 24 EPNs (4.36%) belonging to Heterorhabditis and Steinernema consisting of 21 and 3 isolates, respectively (Figure 1). Most EPNs were isolated from loam with a soil temperature range of 14–31°C (mean 22.8°C), soil pH range of 6.0–7.0 (mean 6.7), soil moisture range of 1.0–7.5 (mean 2.0), and an above mean sea level range of 224–1275 m (mean 359.7 m).

In total, 16 EPN isolates were identified by BLASTN analysis of sequence obtained from a region of the 28S rRNA gene for Steinernema and from the ITS locus for Heterorhabditis. Genomic DNA could not be obtained from the remaining eight isolates owing to contamination by protozoa and fungi.We identified 13 Heterorhabditis isolates as H. indica (seven isolates with 99% identity), H. baujardi (Five isolates with 99% identity), and H. zealandica (one isolate with 100% identity). Phylogenetic tree analysis of Heterorhabditis isolates (Accession No. KY471364–KY471376) obtained in the present study (13 isolates) and Heterorhabditis (20 identified species downloaded from the NCBI database) revealed 3 main groups: group 1 consisted of an isolate of H. indica (KF247222.1) and 7 isolates of Heterorhabditis from the present study; group 2 contained H. baujardi (AF548768.1) and 5 isolates of Heterorhabditis; and group 3 consisted of one isolate of H. zealandica (EF530041.1) and one isolate of Heterorhabditis from our study (Figure 2). In addition, the 3 isolates of Steinernema (Accession No. KY454617–KY454619) were identified as S. websteri (2 isolates, 99% identity) and S. kushidai (1 isolate, 99% identity). The former were closely related to S. websteri (AY841762.1) and the latter was most closely related to S. kushidai (AF331897.1) (Figure 3).

Preliminary identification of symbiotic bacteria at the genus level was based on colony morphology on NBTA agar. In total, 24 EPNs presented as Photorhabdus (green colony) on NBTA agar (21 isolates) and Xenorhabdus (blue colony) on NBTA agar (3 isolates). BLASTN analysis of 588 bp recA sequences revealed that 20 isolates of P. luminescens subsp. akhurstii and 1 isolate of P. temperata subsp. temperata (Accession No. KY436903–KY436923) could be identified with 97–99% identity. Phylogenetic tree analysis of the 21 Photorhabdus isolates demonstrated 2 distinct groups. Group 1 was composed of 20 isolates that were closely related to P. luminescens subsp. akhurstii (FJ862005.1) and group 2 contained only a single isolate that was related to P. temperata subsp. temperata (FJ862014.0) (Figure 4). In turn, the three isolates of Xenorhabdus (Accession No. KY404049–KY404051) were identified as Xenorhabdus stockiae (two isolates) with 97% identity and X. japonica (one isolate) with 98% similarity. Figure 5 shows the phylogenetic tree of Xenorhabdus, which was divided into two groups. Group 1 contained only one study isolate and a recA sequence of X. japonica (FJ823400.1). Group 2 included two study isolates that were closely related to X. stockiae (FJ823425.1).

A clear zone indicating growth inhibition of drug resistant bacteria was demonstrated after exposure to the bacterial extraction from 2 isolates of Xenorhabdus and nine isolates of Photorhabdus (Figures 6, 7 and Table 1). The active isolates were P. luminescens subsp. akhurstii (bMW1.2_TH, bMW8.1_TH, bMW49.3_TH, bMW56.5_TH, bMW59.2_TH, bMW59.5_TH, bMW90.1_TH, and bMW103.2_TH along with P. temperata subsp. temperata bMW27.4_TH). The most potent Photorhabdus isolate was bMW27.4_TH P. temperata subsp. temperata, which could inhibit up to 10 strains of drug resistant bacteria AB320 A. baumannii (extensively drug resistant or XDR), AB321 A. baumannii (mutidrug resistant or MDR), AB322 A. baumannii (MDR), S. aureus ATCC^® 20475, PB36 S. aureus (methicillin-resistant S. aureus or MRSA), PB57 S. aureus (MRSA), E. coli ATCC^® 35218, PB1 E. coli (ESBL+MDR), PB30 P. aeruginosa (MDR), and E. faecalis ATCC^® 51299. All Photorhabdus isolates could inhibit S. aureus ATCC^® 20475, PB36 S. aureus (MRSA), and PB57 S. aureus (MRSA). In addition, the bacterial extraction from bMW27.4_TH P. temperata subsp. temperata could inhibit the growth of S. aureus ATCC^® 20475 with a clear zone larger than 25 mm. However, both isolates of X. stockiae (bMW16.3_TH and bMW16.5_TH) inhibited only the growth of P. aeruginosa.

Based on MIC and MBC testing, eight bacterial extracts from all P. luminescens subsp. akhurstii isolates were evaluated against three strains of drug resistant bacteria including S. aureus ATCC^® 20475, PB36 S. aureus (MRSA), and PB57 S. aureus (MRSA). In addition, one bacterial extract from a P. temperata subsp. temperata isolate was evaluated against five strains of drug resistant bacteria including S. aureus ATCC^® 20475, PB36 S. aureus (MRSA), PB57 S. aureus (MRSA), E. coli ATCC^® 35218, and PB30 P. aeruginosa (MDR). The growth of drug resistant bacteria was inhibited by Photorhabdus extracts with MICs ranging from 7.81 to 0.86 mg/ml. PB36 S. aureus (MRSA) was the most susceptible to all extracts with MICs ranging from 1.95 to 0.86 mg/ml (Table 2) and MBCs ranging from 0.86 to 15.625 mg/ml (Supplementary Figures S1–S4).

Ethyl acetate extracts were checked by HPLC-MS for known compound classes (Supplementary Figure S5 and Table S2). All isolates of Xenorhabdus and Photorhabdus produced GameXPeptide derivatives. All of the Photorhabdus spp. produced isopropylstilbene and all Xenorhabdus isolates produced xenoamicin derivatives. Six isolates of P. luminescens subsp. akhurstii produced xenocoumacin derivatives. Additionally, mevalagmapeptide and phurealipids derivatives were found in many of the isolates (Supplementary Table S2).