PA14 and Escherichia coli strain OP50 were cultured as described previously (Dharmalingam et al., 2012). C. elegans strain CF4059 with genotype fer-15(b26)II; rol-6(su1006)II; fem-1(hc17)IV which is sterile at 25°C and of the roller phenotype to avoid confounding progeny production during screening and aid in worm scoring were obtained from Cynthia Kenyon Lab (University of California, USA). C. elegans strain SAL105 with genotype pha-1(e2123) III;denEx2 whose lys-7 gene was tagged with green fluorescent protein (Alper et al., 2007) were obtained from Caenorhabditis Genetics Center (CGC), USA (https://cbs.umn.edu/cgc/home), respectively. Procedures for maintenance and handling of all worms were approved by the Universiti Sains Malaysia Animal Ethics Committee.

A total of 10 sea bed soil samples were collected from waters at depths ranging from 10 to 20 m deep from Songsong Island, Yan, Kedah, Malaysia (5°48′37.2″N 100°17′47.5″E) on December 2013. The sediment samples were spread on petri plates and dried overnight in laminar flow hood (Valli et al., 2012).

After drying, samples were heated at 70 ± 2°C for 15 min and were grinded lightly with alcohol-sterilized mortar and pestle. Ten-fold serial dilution up to 10⁻⁵ was carried out by diluting 1.0 g of sediment sample in 9.0 mL of 50% artificial sea water (ASW). Approximately, 0.1 mL of the mixture was spread on starch casein agar (SCA) supplemented with 80 μg mL⁻¹of cycloheximide. All plates were incubated at 28 ± 2°C and observed for actinomycetes growth for 28 days (Mincer et al., 2002; Valli et al., 2012). Grown colonies were observed for morphological differences and listed as candidates for screening assays.

The isolates were cultured in M1 medium [ingredients: 10.0 g soluble starch, 4.0 g yeast extract, 2.0 g peptone, and 1.0 L distilled water followed by autoclaving at 121°C for 20 min] and incubated with shaking at 28 ± 2°C, 200 rpm for 7–14 days. The culture broth was freeze-dried and extracted with 1:100 (w/v) methanol (MeOH). The mixture was shaken overnight and then filtered using Whatman grade 1 cellulose filter paper with 11 μM pore size. The filtrate was concentrated using rotary evaporator at 60 ± 2°C. The extracts were stored at −20 ± 2°C and adjusted to working concentration with distilled water.

Anti-bacterial screening of extracts was performed using the modified Kirby-Bauer disc diffusion method. A few colonies of P. aeruginosa PA14 from a culture plate incubated for 24 h were directly inoculated in 0.85% saline. The suspension was compared to a 0.5 McFarland turbidity standard and adjusted with sterile saline. A sterile cotton swab was dipped into the suspension and pressed on the wall of tubes to remove excess bacterial suspension. The swab was repeatedly streaked over the entire surface of Mueller-Hinton agar (Merck, Germany) until the entire surface was streaked. A sterile Whatman antibiotic disc with a 6 mm diameter was placed on the bacterial lawn and 10 μL of 200 μg mL⁻¹ crude extract was transferred onto each disc. The plates were inverted and incubated at 37 ± 2°C for 24 h. The zone of inhibition was measured and recorded after the incubation period.

For the survival assay, C. elegans strain CF4059 was used. Young adult worms were age-synchronized and infected with PA14 on Pseudomonas Infection agar (PIA) as described previously (Dharmalingam et al., 2012). Final concentration of extract in each plate was 200 μg mL⁻¹ while a negative control plate contained only distilled water. Worm survival was scored every 24 h, with mortality designated as failure of worm to react by motion when prodded with platinum wire (Powell and Ausubel, 2008). A total time of 96 h was selected as the screening time-point as it gave the best resolution in identifying a potent hit candidate. Worms that crawled onto the plate wall were not included in the final survival analysis.

The isolate which produced extract contributing to highest worm survival in the slow killing survival assay described above was cultured in M1 broth for 5 days at 180 rpm at 28 ± 2°C. Genomic DNA of the isolate was extracted using Real Biotech Corporation Hi-Yield Genomic DNA Kit. Universal primers 1492R and 27F were used for the amplification of DNA polymerase chain reaction (PCR) amplification with Applied Biosystem Veriti® 96-Well Fast Thermal. The PCR process involves initial denaturation at 94°C for 3 min followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, extension at 72°C for 60 and 40 s, and final extension at 72°C for 5 min. The PCR product was purified using QIA quick PCR Purification Kit by Qiagen. Sequences of purified DNA samples were aligned with the corresponding phylogenetic tree constructed using MEGA6 (Tamura et al., 2013). Comparison of 16S ribosomal RNA gene sequences of the isolate was done using EzTaxon (Chun et al., 2007).

The extract of the isolate contributing to highest survival of worms was further partitioned in n-hexane, dichloromethane, ethyl acetate, and butanol. The methanol-aqueous was mixed with each solvent in 1:1 (v/v) ratio in a separator funnel and shaken vigorously. The funnel was let to stand for 15 min and the resulting layers were collected and dried with a rotary evaporator. The subsequent partitions were then employed in a slow killing survival assay. Subsequently, the partition causing highest percentage of worm survival was subjected to the assays described below.

The CF4059 worms were exposed to the infection plate as described in the slow-killing assay with or without partition supplementation at 50, 200, 400, and 1,000 μg ml⁻¹ final concentration. Worm survival was scored every 24 h.

C. elegans CF4059 was exposed to the infection plates as described in the slow-killing assay. The pharyngeal pumping of three randomly picked worms was observed for 20 s at 12 h interval using Leica Stereomicroscope M205 FA. Pumping rate was measured by counting grinder movement and contraction/relaxation cycles of the bulb (Hobson et al., 2006).

Method for kinetic growth study was a modification of an earlier protocol (Hall et al., 2014). A few PA14 colonies from 24 h freshly cultured plates were inoculated in 250 mL conical flask with 50 mL Mueller-Hinton broth (Merck, Germany), followed by incubation at 37°C, 180 rpm for 24 h. The culture was transferred into 50 mL centrifuge tube and centrifuged at 2,775 × g for 30 min. The supernatant was discarded and remaining cell pellet was washed twice. The cell pellet was then dissolved in fresh MH broth. The treatment well contained 10 μL of bacterial cell culture, 180 μL of MH broth and 10 μL of 8 mg/mL extract. Control well only contained 10 μL of bacterial cell culture and 190 μL of MH broth. The microtiter plates were incubated in the microplate reader at 37°C for 24 h with sampling interval every 4 h at 625 nm (BioTek Synergy Mx, USA).

Biofilm assay was carried out as described previously (O' Toole, 2011). PA14 was cultured in Luria Bertani (LB) broth overnight at 37°C with shaking at 180 rpm. PA14 was cultured in LB broth at 37°C with 180 rpm shaking overnight. A 96-well biofilm assay plate with 400 μg mL⁻¹ of actinomycete partition in LB broth was inoculated with the overnight PA14 culture at 1:100 ratio and further incubated overnight at 37°C. The cells were discarded and plate was rinsed with tap water. Biofilm formed on the wall of the plate was stained with 1% crystal violet and solubilized with 30% acetic acid in water. Optical density was measured using microtiter plate reader (SpectraMax M5) at 550 nm wavelength.

Protease and elastase assay were carried out as described elsewhere (Rudrappa and Bais, 2008). PA14 was cultured at 37°C for 24 h, with or without presence of Streptomyces sp. partition in LB broth. The supernatant was collected and filtered using 0.22 μM nylon filter. About 50.0 μL supernatant was added into the reaction mixture consisting of 0.8% azocasein (Sigma) in 500 μL of 50 mM K₂HPO₄ at pH7. The reaction mixture was incubated at 25°C for 3 h. The reaction was stopped by adding 0.5 mL of 1.5 M HCl into the mixture. The tubes were placed on ice for 30 min and centrifuged at 7,826 × g for 10 min. Finally, 0.5 mL of 1 M NaOH was added into the tubes and the reading was measured at 440 nm.

For elastase assay, 50.0 μL supernatant was added to 1.0 mL of 10 mM Na₂HPO₄ at pH7 and 20 mg of elastin-Congo red. The tubes were incubated for 4 h at 37°C with 180 rpm shaking. The tubes were centrifuged at 7,826 × g for 10 min and the optical density reading was taken at 495 nm.

Pyocyanin assay was carried out as described elsewhere (Essar et al., 1990). PA14 was cultured with or without presence of Streptomyces sp. metabolites and supernatant was collected as above. About 4.5 mL of chloroform was added to 7.5 mL of collected supernatant and vortexed for 20 s. The mixture was centrifuged at 2,880 × g for 10 min. About 3.0 mL of the resulting blue layer at the bottom of the tube was transferred into a new tube, followed by addition of 1.5 mL of 0.2 M HCl and vortexing for 20 s. Tubes were then centrifuged for 2 min at 2,880 × g and 1.0 mL of the ensuing pink layer was transferred into cuvettes and reading was taken at 520 nm (Thermo Scientific Genesys20).

Slow killing assay was carried out using the transgenic lys-7::GFP C. elegans strain SAL105. The fluorescence micrograph of worms was captured using a Leica Microsystem M205 FA following 24 h of pathogen exposure. Images were analyzed using Image J (National Institutes of Health, USA) to quantify lys-7 fluorescent intensity.

TLC plate (Merck TLC silica gel 60 F₂₅₄, Germany) coated with silica was used as the stationary phase. Sample of hexane partition was prepared by diluting 10 mg of extract in 1.0 mL of CHCl₃. The sample was spotted on the plates with a capillary tube. The plates were then put in developing chamber with solvent system of methanol:chloroform of 20:80. After drying, a small portion of the plate was cut, followed by staining with vanillin-sulphuric acid reagent (Yadav and Gupta, 2013). The stained plates were then air-dried for 15 min and oven-dried at 96 ± 2°C for 8 min (Maurya and Srivastava, 2013). Spots formed were aligned on the plate and marked. The marked area was scraped using a scalpel. The collected fractions were dissolved in 100% ethanol, filtered by using Whatman no.1 filter paper and rotated to dryness. The fractions were subjected for survival assay and the fraction with positive result was sent for GC-MS analysis (Agilent 6890, USA) with capillary column of 30 m × 0.25 mm × 0.25 μm (Agilent HP-5 ms, USA). The flow rate was set at 1.2 mL/min, with 10 μL sample injection. Helium gas was used and total run was 40 min. The obtained spectrum was compared with NIST Spectral Library for compound identification.

All numerical data were analyzed using GraphPad Prism 5 and StatView5.0.1 (SAS Institute, Inc) software. Values were presented as mean ± standard deviation (SD) of at least two independent experiments. Data from the killing assays were analyzed with StatView 5.0.1 and plotted using the Kaplan-Meier Cumulative Survival Plot for Time (non-parametric survival analysis). The comparison was analyzed using the GraphPad Prism 5 Log-rank (Mantel-Cox) significance test. Data from dose response assay, pharyngeal pumping assay, biofilm assay, kinetic growth of PA14 and total cell fluorescent count were analyzed with GraphPad Prism 5 unpaired t-test.

A total of 18 morphologically different strains were successfully isolated from the marine sediment samples (Table S1). These isolates produced aerial mycelium, with four of the isolates showing pigmentation. All isolates produced mycelial clump when cultured in broth medium after incubation in shaker at 180 rpm and 28 ± 2°C. All 18 isolates were extracted and subjected to anti-microbial assay. None of these isolates caused a visible inhibition zone on the P. aeruginosa PA14 lawn (data not shown) which means these extracts do not possess bactericidal activities toward PA14.

Compared to untreated PA14 infected worms, infected worms treated with 8 of the 18 marine actinomycetes extract, respectively, (A3, A5, A22, A26, A38, A42, A48, and A50) showed improved survival rates (Figure 1). Among these, statistically significant increase in survival rate was achieved with A3, A5, and A22 (36.99 ± 2.80–57.31 ± 3.85%), with isolate A5 contributing to the highest survival rate. In tandem, A5 also delayed mortality of infected worms, as represented in TD50 value (Table S2). In addition, worms treated with the remaining extracts (A20, A24, A30, A31, A39, A40, A41, A43, A45, and A47), showed increased susceptibility to killing by PA14. Further partitioning of the isolate A5 methanol extract followed by slow killing assay revealed the hexane partition of isolate A5 to be most potent in attenuating the killing of infected worms, with survival rate of 69.65 ± 4.50% (Figure 2) and TD50 = 102.2 ± 5.54 h (Table S3).

Using partial 16S analysis, the A5 isolate showed sequence identity of 99.7–99.85% with several Streptomyces sp. This include Streptomyces sundarbansensis, Streptomyces puniceus, Streptomyces badius, Streptomyces sidensis, Streptomyces rubiginosohelvolus, Streptomyces pluricolorescense, Streptomyces parvus, Streptomyces globisporus, and Streptomyces cyaneofuscatus (Figure 3). The sequence was deposited in NCBI GenBank under accession number KX372372. On ISP2 agar plates, A5 isolate produced white beige aerial mycelia and brownish substrate mycelia with no pigmentation (Figure S1). A5 isolate was named as Streptomyces sp. CCB-PSK207.

Results showed that the hexane partition of the Streptomyces sp. CCB-PSK207 extract promoted survival of PA14-infected worm in a dose dependent manner with a gradual increase observed from 45.33 ± 4.32 to 72.71 ± 4.66% at concentration range of 50–400 μg mL⁻¹(Figure 4). There was no further increase in C. elegans survival promotion at 1,000 μg mL⁻¹ concentration.

Comparison of pharyngeal pumping rate in C. elegans exposed to the Streptomyces sp. CCB-PSK207 hexane partition with worms without presence of partition showed no significant difference for all three time points (Figure 5). This indicated that the Streptomyces sp. CCB-PSK207 hexane partition did not interrupt C. elegans feeding rate.

Similarly, the Streptomyces sp. CCB-PSK207 hexane partition did not impair growth kinetics of PA14, which denote that the rescue of PA14 infected worms did not occur through killing of pathogens (Figure 6).

Results showed no significant difference in the level of biofilm, protease, and pyocyanin production between control PA14 and those exposed to Streptomyces sp. CCB-PSK207 hexane partition (Figure 7). However, a significant increase in elastase production was observed in treated PA-14.

As compared to worms fed with E. coli OP50, PA14-infected worms showed diminished fluorescent signal (Figures 8A,B). However, treatment with Streptomyces sp. CCB-PSK207 hexane partition appear to restore GFP expression in both PA 14 infected worms and OP50 fed worms (Figures 8C,D). Imaging-based software quantification showed that worms exposed to Streptomyces sp. CCB-PSK207 metabolites produced significantly highest intensity of GFP expression (Figure 8E).

Three fractions were collected from preparative TLC and further employed in the C. elegans slow killing survival assay. Among them, fraction A5HB showed a significant increase in worm survival rate at 71.43 ± 4.67% (Figure 9) and TD50 of 93.6 ± 1.9 h (Table S4). Fraction A5HA resulted in < 20% worm survival, while A5HC increased the susceptibility of worms toward mortality caused by PA14, respectively.

Further chemical profiling and compound identification was done using GC-MS (Pollak and Berger, 1996). Seven main compounds were identified, Figure S2 namely (1) tetradecanoic acid methyl ester, (2) pentadecanoic acid 14-methyl methyl ester, (3) tetradecanoic acid, 12-methyl-methyl ester, (4) tridecanoic acid methyl ester, (5) hexadecanoic acid methyl ester (6) octadecanamide, and (7) 1,2-benzenedicarboxylic acid mono(2-ethylhexyl) ester (Table 1).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.