Isolation, characterization, anti-MRSA evaluation, and in-silico multi-target anti-microbial validations of actinomycin X2 and actinomycin D produced by novel Streptomyces smyrnaeus UKAQ_23

Streptomyces smyrnaeus UKAQ_23, isolated from the mangrove-sediment, collected from Jubail,Saudi Arabia, exhibited substantial antimicrobial activity against methicillin-resistant Staphylococcus aureus (MRSA), including non-MRSA Gram-positive test bacteria. The novel isolate, under laboratory-scale conditions, produced the highest yield (561.3 ± 0.3 mg/kg fermented agar) of antimicrobial compounds in modified ISP-4 agar at pH 6.5, temperature 35 °C, inoculum 5% v/w, agar 1.5% w/v, and an incubation period of 7 days. The two major compounds, K1 and K2, were isolated from fermented medium and identified as Actinomycin X2 and Actinomycin D, respectively, based on their structural analysis. The antimicrobial screening showed that Actinomycin X2 had the highest antimicrobial activity compared to Actinomycin D, and the actinomycins-mixture (X2:D, 1:1, w/w) against MRSA and non-MRSA Gram-positive test bacteria, at 5 µg/disc concentrations. The MIC of Actinomycin X2 ranged from 1.56–12.5 µg/ml for non-MRSA and 3.125–12.5 µg/ml for MRSA test bacteria. An in-silico molecular docking demonstrated isoleucyl tRNA synthetase as the most-favored antimicrobial protein target for both actinomycins, X2 and D, while the penicillin-binding protein-1a, was the least-favorable target-protein. In conclusion, Streptomyces smyrnaeus UKAQ_23 emerged as a promising source of Actinomycin X2 with the potential to be scaled up for industrial production, which could benefit the pharmaceutical industry.

a 17 mm zone of inhibition against the test organism, while other tested fermentation media showed no zone of inhibition against the tested organism. To improve antibiotic production, the ISP-4 medium was further optimized, and it was found that only two of the twenty-four tested fermentation broths, the FM-5, and FM-9, supported antibiotic production, with inhibition zones of 17 mm and 20 mm, respectively, against the tested organism. As a result, FM-5 and FM-9 were selected for further finer optimizations. In submerged fermentation, the antibiotic production was significantly variable; therefore, further optimization of media composition and fermentation conditions was carried out in solid-state fermentation. The highest yield (Mean ± SD) of crude antimicrobial extract from solid-state fermentation was 561.3 ± 0.3 mg/kg fermented agar in a fermentation medium (modified ISP-4 agar) consisting of starch maize 1% w/v, starch soluble 0.6% w/v, (NH 4 ) 2 SO 4 0.5% w/v, NaCl 0.75% w/v, K 2 HPO 4 0.6% w/v, CaCl 2 0.15% w/v, MgSO 4 0.05% w/v, CaCO 3 0.015% w/v, FeSO 4 0.010% w/v, ZnSO 4 0.005% w/v, agar 1.5% w/v, pH 6.5, inoculum 5% v/w at temperature 35 °C, and an incubation period of 7 days (see, Supplementary Table S3).

Optimization of antibiotic production by response surface methodology (RSM). The results
showed that the linear and quadratic effects of pH, temperature, inoculum volume, and agar concentrations were highly significant (P < 0.05), but the interactive effects of agar concentrations and inoculum volumes were insignificant (P > 0.05). The optimal level of each variable and the effects of their interactions on antibiotic output were explored using three-dimensional (3D) response surface curves (Fig. 4), and the findings were dependent on the strategy of holding one variable constant at its optimum level while the other two variables were varied across the experimental range. The 3D curves of the measured responses indicated interactions between pH, temperature, inoculum volume, and agar concentrations (Figs. [4][5]. The crude antimicrobial extract's experimental yield was 561.3 ± 0.3 mg/kg of the fermented agar, whereas the RSM predicted yield of the crude antimicrobial extract was 842.0 ± 0.0 mg/kg of fermented agar (see , Supplementary Table S4). An ANOVA analysis revealed P < 0.05, indicating that the model was significant. The 3D response surface and contour presentations were plotted to investigate the interaction of the different physicochemical factors used and evaluate the optimal range of each factor for maximal antibiotic production.
Antibiotic extraction and purification. The solid-liquid extraction produced a crude antimicrobial extract that was reddish-orange in color and amorphous. The TLC analysis revealed that the crude antimicrobial extract contained two prominent compounds, K 1 and K 2 , with Rf values of 0.38 and 0.28, respectively (see, Supplementary Figure S2). The bioautography revealed that both the compounds, K 1, and K 2 had antimicrobial activity against the indicator organism (see, Supplementary Figure S2). Thus, both compounds were selected for further analyses. The comprehensive application of various chromatographic techniques resulted in the isolation of two pure compounds, K 1 and K 2 . The compounds K 1 and K 2 were subjected to further structure elucidation and antimicrobial activity assessments.
Physicochemical characterization and structure elucidation of compounds K 1 and K 2 . The purified compounds, K 1, and K 2 based on their physicochemical and spectroscopic analyses, were identified as actinomycin X 2 and actinomycin D, respectively (Fig. 6), ( Table 1).
The compounds, K 1 and K 2 , were isolated as reddish-orange amorphous powders. Their UV-Visible spectra showed absorption maxima (λ max ) at 239 and 440 nm for the compound K 1 (see, Supplementary Figure S3) and 215 and 442 nm for the compound K 2 (see, Supplementary Figure S4), corresponding to the known actinomycins, X 2 and D 17 . The IR spectra of K 1      www.nature.com/scientificreports/   www.nature.com/scientificreports/ contains Oxo-Pro moiety instead of Pro in one chain (see, Supplementary Figure S7). A literature search revealed that the actinomycin analog with OxoPro in one of the peptide rings has been reported, i.e., Actinomycin X 2 . An LC-MS comparison between compound K 1 and standard Actinomycin X 2 revealed similar retention times (see, Supplementary Figure Figure S10). These experiments demonstrated that there was no difference between the compared compounds. As a result, it was concluded that compound K 1 is similar to Actinomycin X 2 .
The HR-ESI-MS spectrum of K 2 showed molecular ion peaks at m/z 1255.638 [M + H] + and 1277.623 [M + Na] + , which corresponded to Actinomycin D (see, Supplementary Figure S11). These molecular ion peaks corresponded to the molecular formula C 62 H 87 N 12 O 16 [M + H] + of K 2 . The mass and molecular formula were consistent with Actinomycin D. The CID MS/MS fragmentation spectrum of compound K 2 was compared with the standard Actinomycin D 17 (see , Supplementary Table S6).
An LC-MS comparison of compound K 2 with the standard Actinomycin D revealed a similar retention time (see, Supplementary Figure S12) and same MS (see, Supplementary Figure S13), whereas an LC-MS-MS comparison revealed a similar fragmentation pattern (see, Supplementary Figure S14). These observations revealed that there was no difference between the compared compounds. As a result, it was concluded that compound K 2 is similar to Actinomycin D.
NMR spectroscopy. The Nuclear Magnetic Resonance (NMR) is a powerful analytical tool that has been used as a gold standard for molecular identification [18][19][20] , structural elucidation, and drug discovery [20][21][22] . In this study, several 1D and 2D NMR techniques were used for the structure elucidation of the isolated compounds. The NMR spectroscopic data for the K 1 peptide displayed typical characteristics of an actinomycin molecule, including the presence of a phenoxazinone unit and two pentapeptide lactone rings. The 1 H NMR spectrum of K 1 showed two ortho-coupled aromatic protons at δ 6.66 (d, J = 7.7 Hz, H-8) and 6.40 (d, J = 7.7 Hz, H-7) together with two methyl protons at δ 2.59 (s, H 3 -13) and δ 2.25 (s, H 3 -12), characteristic for the phenoxazinone chromophore (see , Supplementary Table S7). Furthermore, the 1 H NMR spectrum of K 1 displayed characteristics of a typical peptide, with NH protons at δ 7.  Table S7). The 13 C NMR and DEPT-135 spectra also showed 16 methyl groups from which two were assigned to C-12 (δ 7.20) and C-13 (δ 14.48) of the phenoxazinone, 5 methylenes from which three were assigned to C-3, C-4, and C-5 of the α-ring proline, and two for the C-3 and C-5 of β-ring proline, indicating that one of the methylene carbon of this proline was modified, a 2 sp 2 methines (C-7 and C-8 of phenoxazinone nucleus), 6 sp3 methines including two oxygenated (C-3 of each threonine), and four nonoxygenated (C-3 of each valine and methyl-valine), and 12 sp 2 quaternary carbons including three carbonyls for C-3, C-11, and C-14 of the phenoxazinone nucleus. The ongoing data are in agreement with spectral features observed for the actinomycins. Additionally, the downfield carbonyl signal at δ 208.27 was assigned to C-4 of β-ring proline, indicating that compound K 1 has dissimilar pentapeptide lactone rings, one ring contains proline, and the other contains oxoproline. After an extensive 2D-NMR (HSQC, COSY, and HMBC) analyses, the amino acid residues in each peptide ring were identified as Thr, Val, Pro, Sar, and Me-Val in α-ring, and Thr, Val, OxoPro, Sar, and Me-Val in β-ring (see, Supplementary Table S7). The amino acid α-protons showed correlations with the corresponding amino acid's α-carbons as shown in the HSQC spectrum of K 1 . Furthermore, the five amino-acids residues in both the pentapeptide lactone rings were identified from their unique spin system, as shown in the COSY spectrum 23 . The sequence of amino acids in these two peptide lactone units in K 1 was identical to those of the Actinomycin X 2 . The sequence of amino acids residue in α-ring was deduced from the following HMBC correlations www.nature.com/scientificreports/ assigned to be identical to that of Actinomycin X 2 (see, Supplementary Figure S15-S21). The compound K 1 was identified as Actinomycin X 2 based on ongoing data and comparison with previously published spectroscopic and chromatographic data 17,23-28 .
Antimicrobial activity of compounds K 1 and K 2 . Primary antimicrobial activity. The primary antimicrobial activity of the compounds K 1 and K 2 demonstrated that Actinomycin X 2 had a larger zone of inhibition, i.e.,14.7 ± 2.7 mm and Actinomycin D had a smaller zone of inhibition, i.e., 9.9 ± 2.8 mm, whereas actinomycinsmixture (X 2 + D, 1:1 w/w) had a medium zone of inhibition, i.e., 13.5 ± 3.0 mm (Mean ± SD), against the Grampositive test bacteria, at a concentration of 5 µg/disc. According to the results, Actinomycin X 2 was the most potent antibiotic, while Actinomycin D was the least potent, and actinomycins-mixture showed mild potency. Levofloxacin (control antibiotic) inhibited non-MRSA bacteria with a zone of inhibition of 25.6 ± 6.91 mm and MRSA bacteria with a zone of inhibition of 28.8 ± 7.14 mm (Mean ± SD) (Figs. 7, 8) (see, Supplementary Table S8). Based on the above findings, Actinomycin X 2 and Actinomycin D were selected for further investigation, and actinomycins-complex (X 2 + D) was omitted.
Secondary antimicrobial activity (MIC and MBC). MIC of Actinomycin X 2 ranged from 1.56 to 12.5 μg/ml for non-MRSA test bacteria and from 3.125 to 12.5 μg/ml for MRSA test bacteria, while MIC of the isolated Actinomycin D ranged from 3.125 to 12.5 μg/ml for non-MRSA test bacteria and from 12.5 to 25 μg/ml for MRSA test bacteria. Levofloxacin (control antibiotic) inhibited the growth of all the test organisms at a concentration of 5 µg/ml ( Table 2). MBC of Actinomycin X 2 ranged from 6.25 to 50 μg/ml for non-MRSA test bacteria and from 12.5 to 50 μg/ml for MRSA test bacteria, while MBC of Actinomycin D ranged from 25 to 50 μg/ml for non-MRSA test bacteria and 50 μg/ml for MRSA test bacteria (Table 2).

Molecular docking of isolated actinomycin X 2 (K 1 ) and actinomycin D (K 2 ). The in-silico molecu-
lar docking experiments predicted affinity in terms of binding energies of the compounds K 1 and K 2 against the DNA and eight other potential antibacterial protein targets ( Table 5). The molecular docking interactions are represented in Fig. 9. Detailed in-silico intermolecular interaction analysis is presented in Table 6. Both compounds, K 1 and K 2, exhibited maximum affinity with isoleucyl-tRNA synthetase showing binding energy of − 11.6 kcal/mol and − 11.3 kcal/mol, respectively. However, penicillin-binding protein-1a was observed as the least favorable target for both docked compounds demonstrating an equal value of binding energy as − 7.2 kcal/ mol; interestingly, both compounds preferred to bind with DNA, indicating equipotent in terms of docking predicted affinity of − 10.6 kcal/mol. In general, a very subtle difference was noted between the binding energies of compounds K 1 and K 2 except dihydrofolate reductase target where compound K 2 (− 8.1 kcal/mol) was observed as more potent than K 1 (− 6.3 kcal/mol) with an appreciable difference in the binding energy (Table 5).

Discussion
The rising prevalence of antibiotic-resistant bacteria hampers the efficacy of the currently available antibiotics and the therapeutic efficiency. This highlights an urgent need to search for new antibiotics with diverse antimicrobial activity to treat the pathogens that are resistant to conventional antibiotics. Ample evidence suggests that terrestrial actinomycetes are still the most promising candidates for discovering new bioactive compounds www.nature.com/scientificreports/ active against a wide range of pathogenic microorganisms. The isolation of biologically active compounds from terrestrial actinomycetes is still a field of interest in the quest for bioactive compounds [42][43][44][45][46] . Previously, we isolated Bacillus pumilus strains from soil samples collected from Unaizah, Saudi Arabia, and found that the isolated strains were bioactive against various human pathogens, including Gram-positive and Gram-negative test bacteria 3,4,8 . These recent studies motivated us to isolate antibiotic-producing soil organisms from the mangrove sediment samples of Jubail, Saudi Arabia. In the present study, twenty-five mangrove sediment samples were collected from Jubail, Saudi Arabia, and then screened for the existence of bioactive  www.nature.com/scientificreports/ actinomycetes strains. Only one strain, Streptomyces smyrnaeus UKAQ_23, exhibited extremely potent antimicrobial activity among the five actinomycetes strains isolated from the twenty-five collected mangrove sediment samples. The selected isolate was employed for antibiotic production in a wide range of fermentation media, resulting in antibiotic production in an inorganic salt medium (ISP-4). The media composition and fermentation conditions were optimized to achieve maximum antibiotic production, resulting in high yield antibiotic production in modified inorganic salt medium (modified ISP-4 agar) by solid-state fermentation. The antibiotics were extracted using a solid-liquid extraction method, and the purification was performed using various chromatographic techniques, resulting in the isolation of two purified antimicrobial compounds, K 1 and K 2 . The structures of the compounds, K 1 and K 2 , were elucidated employing various spectroscopic techniques, including UV, IR, mass and 1D and 2D NMR spectroscopic techniques, resulting in the identification of the compounds, K 1 and K 2 , as Actinomycin X 2 and Actinomycin D, respectively. The antimicrobial screening of isolated actinomycins, X 2 and D, revealed that both actinomycins exhibited potent antimicrobial properties against the tested bacterial strains, and further research revealed that Actinomycin X 2 exhibited higher antimicrobial property than Actinomycin D. As a result, we concluded that the isolated actinomycete strain, Streptomyces smyrnaeus UKAQ_23, is an opulent source for the production of Actinomycin X 2 and Actinomycin D and could be used as an alternative source of said actinomycins, which could be a significant shift in the pharmaceutical industry 17,47-52 .  Our recent findings showed that actinomycins, X 2, and D produced by Streptomyces smyrnaeus UKAQ_23 exhibited potent antimicrobial activities against MRSA and non-MRSA Gram-positive bacterial strains, but no bioactivity was demonstrated against tested Gram-negative bacteria and fungal strains, which is consistent with previous studies; Sharma and Manhas, who demonstrated the antibacterial potential of actinomycins V, X 2 , and D produced by Streptomyces strain M7, and the findings revealed that actinomycins X 2 and D had significant antibacterial potential against VRE, MRSA, and Bacillus subtilis with MIC values ranging from 1.95 to 8.0 μg/ ml 54 ; Wang et al. demonstrated antimicrobial activity of actinomycins X 0β , X 2 , and D produced by Streptomyces heliomycini, and the results showed that actinomycins X 0β , X 2, and D had substantial antibacterial potential against S.aureus, MRSA, Bacillus subtilis, and Bacillus cereus with MIC values ranging from 0.04 to 2.48 μg/ml 56 .
Wang et al. demonstrated the antibacterial potential of four actinomycins, neo-Actinomycin A, neo-Actinomycin B, Actinomycin D, and Actinomycin X 2 produced by Streptomyces sp. IMB094 and the results revealed that actinomycins D and X 2 had substantial antibacterial activity against S.aureus, S.epidermidis, and E.faecalis with MIC values ranging from 0.06 to 0.5 μg/ml, while neo-Actinomycins A and B exhibited less-substantial antibacterial potential with MIC values ranging, 16-32 μg/ml by neo-Actinomycin A and > 128 μg/ml by neo-Actinomycin B, respectively. However, with MIC values ranging from 16 to 128 μg/ml, all four actinomycins had insignificant antibacterial activity against E.coli, K.pneumoniae, P.aeruginosa, S.marcescens, A.calcoaceticus, P.mirabilis, P.rettgeri, P.vulgaris, C.freundii, M.morganii, S.maltophilia, and Enterobacter spp. 28 .
Khieu et al. demonstrated that Actinomycin D produced by Streptomyces sp. HUST012 had substantial antibacterial activity against MRSA ATCC 25923, MRSE ATCC 35984, Escherichia coli ATCC 25922, and Klebsiella pneumoniae ATCC 13883 with MIC values ranging from 0.04 to 2.24 μg/ml 57 , which was partially deprived of our findings, as our isolated actinomycins had shown non-substantial antibacterial activity against Gramnegative test bacteria. Kulkarni et al. demonstrated the antimicrobial potential of Actinomycin D produced by Streptomyces hydrogenans IB310. They showed that Actinomycin D had substantial antibacterial and antifungal potential against plant pathogens and suggested the potential application of Actinomycin D in agriculture to manage plants' bacterial and fungal infections 58 , which was partially deprived of our findings, as our isolated actinomycins had shown non-substantial antifungal activity.
Chen et al. reported that several species of Streptomyces produce Actinomycin D analogs. However, only a few species produce in significant amount, including Streptomyces parvulus (0.152 mg/ml Actinomycin D), Streptomyces griseoruber (0.21 mg/ml Actinomycin D), Streptomyces sindenensis (0.85 mg/ml Actinomycin D), Streptomyces MITKK-103 (0.11 mg/ml Actinomycin X 2 ), Streptomyces sp. JAU4234 (0.62 mg/ml actinomycin X 2 ), Streptomyces nasri strain YG62 (0.15 mg/ml Actinomycin X 2 ), and Streptomyces strain MS449 (1.92 mg/ml Actinomycin X 2 , 1.77 mg/ml Actinomycin D) 24 . This study confirms our findings and suggests that the isolated www.nature.com/scientificreports/ Targets Actinomycin X 2 (K 1 ) Actinomycin D (K 2 ) www.nature.com/scientificreports/ strain Streptomyces smyrnaeus UKAQ 23 is suitable for industrial use as it can produce a significant amount of Actinomycin D. However, prior to its commercial employment, strain improvement is essential. Our findings on antibiotic production and optimization of fermentation conditions indicated that the modified ISP-4 agar was the best fermentation medium for maximum antibiotic production at pH 6.5, temperature 35 °C, inoculum 5% v/w, agar 1.5% w/v, and incubation period of 7 days. Furthermore, the solid-state fermentation method was superior to the submerged-state fermentation method for the production of crude antibiotics, possibly due to the lack of water in solid-state fermentation. The findings obtained by optimizing the fermentation medium using RSM significantly affected antibiotic production by Streptomyces smyrnaeus UKAQ_23. Furthermore, to avoid the formation of fermentation products with poor antibacterial activity, pH, temperature, inoculum concentration, agar concentration, and incubation time, among other variables, must be tightly regulated throughout fermentation. This method may be used to conduct further investigation into antibiotic production. The higher yield of crude antimicrobial extract obtained using the solid-state fermentation method was consistent with previous reports [59][60][61][62][63] . Surprisingly, one study showed that solid-state fermentation conditions are ideal for producing antimicrobial compounds by Streptomyces youssoufiensis SF10 strain 64 .

Residues involved in hydrophobic interactions
Molecular docking is a robust computational technique for predicting bound conformations of drug candidates and their binding affinities. In this study, the molecular interaction of compounds K 1 and K 2 against several drug-targets of bacterial biochemical pathways was studied for the first time by molecular docking using AutoDock Vina 1.1.2 65,66 . Analysis of the docking results (Table 6) revealed that both compounds, K 1 and K 2, possessed sufficient molecular framework for their interaction with numerous drug targets, which might be considered accountable for producing the antibacterial activity (Figs. 10-11). The preferential binding of compound K 1 with IleRS, MurD ligase, and GFAT targets further signifies the importance of an additional keto group in the pyrrolidine ring of the proline moiety. Although both compounds K 1 and K 2 were observed to be most potent against IleRS, it was compound K 1 which offered an additional hydrogen bond accepting platform in the form of oxo-Pro moiety for interaction with Ser593 residue in the binding pocket of IleRS, the most promising antibacterial drug target as per our docking results (Fig. 11). However, PBP-1a was the least desirable macromolecular site for both compounds K 1 and K 2 .
Among potential contacts of compound K 2 within inhibitor binding cavity of IleRS, Asn50, Leu52, Arg391, Gly551, His581, and Ser593 were recognized to afford polar interactions in the form of hydrogen bonds. Residue His581 is positioned at the C-terminus of the last β-sheet of the Rossmann fold of the IleRS protein, is conserved in the bacterial IleRS but is substituted with an asparagine/serine in the eukaryotic IleRS 36 . Both compounds Targets Actinomycin X 2 (K 1 ) Actinomycin D (K 2 )

Residues involved in hydrophobic interactions
Isoleucyl-tRNA synthetase  www.nature.com/scientificreports/ K 1 and K 2 interacted with His581 by polar and hydrophobic contacts, respectively. IleRS plays an important role in the tRNA aminoacylation step of the bacterial protein synthesis, and hence, its inhibition has proven to be an effective antimicrobial strategy, hampering a vital step of protein synthesis 67 . These results may provide a foundation for experimental design for further exploration of the precise antimicrobial mechanism of action of compounds K 1 and K 2 .

Sample collection and isolation of actinomycetes.
A total of twenty-five mangrove sediment samples were collected from Jubail, Saudi Arabia (latitude: 27°00′40″ N; longitude: 49°39′29″ E; altitude: 22 ft; annual rainfall: 97 mm; average temperature: 26.6 °C) 8 . All the mangrove sediment samples were air-dried at room temperature for 1 week and then pre-treated for 1 h in a hot air oven at 60 °C to enrich and selectively isolate actinomycetes 68 . A serial dilution of up to 10 −5 was then prepared by dissolving the 1 g of sample in 0.89% NaCl. From each dilution, a 0.1 ml sample was spread out on ISP-4 agar supplemented with nalidixic acid (50 μg/ml) and cycloheximide (50 μg/ml) to isolate actinomycetes strains. The nalidixic acid and cycloheximide were supplemented to inhibit the growth of the Gram-negative bacteria and fungi, respectively 8,68 . The inoculated plates were then incubated at 28 °C for up to 14 days. The isolated actinomycetes were sub-cultured and purified on ISP-4 slants. The purified cultures were preserved in 20% (v/v) glycerol at − 78 °C for further use.

Preliminary antimicrobial screening of isolated actinomycetes. Preliminary antimicrobial screen-
ing of the isolated actinomycetes was conducted by the spot inoculation method using modified tryptic soy agar (MTSA) plates 3,4,8 . MTSA plates consisted of starch soluble 1 g, the pancreatic digest of casein 1.5 g, papain digest of soya bean 0.5 g, sodium chloride 0.5 g, agar 1.7 g, peptone 0.4 g, tryptone 0.4 g, beef extract 0.3 g, lactose 1 g, L-cystine 0.013 g, bromothymol blue 0.002 g, agar 1.5-1.7 g, ultrapure water 100 ml, and pH 6.8 ± 0.2.  Aspergillus niger (A.niger) ATCC 6275 were used as test organisms. The antibacterial activity was determined by incubating the inoculated plates at 35 ± 2 °C for 24 h., and antifungal activity was determined by incubating the inoculated plates at 28 ± 2 °C for up to 48 h. The antimicrobial activities of all the isolates were recorded by observing the zone of inhibition around the actinomycetes spot 8,68 . The selection of the highly potent actinomycete strain, UKAQ_23, was based on the zone of inhibition around the inoculated spot (growth).

Identification and characterization of selected isolate UKAQ_23. The selected isolate UKAQ_23
was identified by 16S rRNA gene sequencing 8,24 . The phylogenetic analysis was performed using the Maximum Likelihood approach using the Tamura-Nei model, and the phylogenetic tree was constructed using the neighbor-joining (NJ) method MEGA X 24,69 .
The cultural characterization of strain UKAQ_23 was determined by recording the cultural characteristics, i.e., the color of substrate and aerial mycelium, spore formation, diffusible pigment on various growth media, i.e., GM-1… GM-7) at 28 °C for up to 7 days. SEM recorded the morphological characteristics 54 .
Optimization of media composition and fermentation conditions for high yield antibiotic production. Initially, the production of antibiotic was carried out in submerged-state fermentation, but owing to large fluctuation in the production of antibiotic, the optimization was shifted to solid-state fermentation 61 The effects of media constituents and fermentation conditions on antibiotic production were determined by varying the one factor at a time and base fermentation conditions were kept constant, i.e., pH 6.5, temperature 30 °C, carbon source 1% w/v, nitrogen source 0.5% w/v, NaCl 0.50% w/v, inoculum 5% v/w, agar 1.5% w/v, and incubation period of 7 days. After completing each set of fermentation, the yield of antimicrobial extract was measured in mg/kg of fermented agar 53,61,62,73,74 . Optimization of antibiotic production by RSM. After optimizing the media composition and fermentation conditions, the four most influential variables (pH, temperature, inoculum volumes, and agar concentrations) were chosen for further antibiotic-production-optimization by RSM using the Box-Behnken design (BBD) 79 . The predicted model was validated and found to be statistically suitable for use. A total of 27 experiments were carried out to optimize the main factors for a 3-levels-4-factors BBD with three replicates at the center. After evaluating the responses for each trial, each response was fitted to an individual second-order polynomial model. For media optimization, a total of twenty-seven BBD experiments were run in one block. The minimum and maximum ranges of variables investigated in terms of their actual and coded values are mentioned (see, Supplementary Table S10). To assess the impact of process variables on antibiotic production, the BBD was used (see , Supplementary Table S11).
Extraction and purification of antimicrobial compounds. The extraction of antimicrobial compounds from the fermented agar was carried out by solid-liquid extraction method as described earlier 60,61 (see, Supplementary Figure S22). The resulted crude antimicrobial extract was subjected to further purification.
A thin layer chromatography (TLC) was employed to assess the purity and number of compounds present in the crude antimicrobial extract using a solvent system composed of dichloromethane (CH 2 Cl 2 ) and acetone ((CH 3 ) 2 CO) in a ratio of 60:40. The TLC was run on pre-coated silica gel (60 F254, Merck, USA) sheets 2 . The plates were visualized under UV light. The bioautography was performed on a developed TLC plate 80 . S.aureus ATCC 29213 was employed as a test organism.
A solvent system composed of dichloromethane and acetone (60:40) was employed in the preparative TLC and silica gel column chromatography, while pure methanol was employed in the size-exclusion chromatography. The purification steps culminated into two pure antimicrobial compounds named K 1 and K 2 . The homogeneity of purified compounds was assessed by HPLC (Ultimate 3000 HPLC, Thermo Scientific, Waltham, Massachusetts, USA).
The solubility of compounds was determined by dissolving the compounds in various solvents, e.g., water, ethanol, methanol, acetone, acetonitrile, dimethylformamide, dimethyl sulphoxide, chloroform, ethyl acetate, dichloromethane, and n-hexane at a concentration of 10 mg/ml. The melting points of the compounds were determined by the capillary method with the Dynalon DMP100 Digital Melting Point Device (Spectrum Chemical www.nature.com/scientificreports/ Manufacturing Corporation, USA). The compounds were dissolved in methanol at a concentration of 1 mg/ ml, and UV-Visible spectra were recorded with the Ultraspec 8000 spectrophotometer (GE, Pittsburgh, PA, USA). FT-IR spectra were recorded with the Nicolet iS20 FT-IR spectrometer (Thermo Scientific, Waltham, Massachusetts, USA). The MS n analysis was performed on LCMS-IT-TOF (Shimadzu Corporation, Japan Antimicrobial activity. Primary antimicrobial activity. The standard disc diffusion method determined the primary antimicrobial activity of the isolated antimicrobial compounds (K 1 , K 2 , and K 1 + K 2 ) [76][77][78] . The isolated antimicrobial compounds were dissolved in methanol, and then diluted samples were dispensed on sterile paper discs (6 mm size). Each disk consisted of 5 µg of the antimicrobial compound. This experiment investigated the antimicrobial efficacy of isolated antimicrobial compounds on the tested microorganisms. A disk containing 20 μl of methanol was used as a negative control, while Levofloxacin (5 µg/disc) was used as positive control antibiotic. Each test was performed in triplicate. The diameters of the inhibitory zones were measured on an mm scale. The results were recorded in Mean ± Standard Deviation (SD).

Secondary antimicrobial activity (MIC and MBC).
The secondary antimicrobial activity of isolated antimicrobial compounds (K 1 and K 2 ) was determined by performing the MIC and MBC. MIC was determined by the resazurin-based micro broth dilution method, while MBC was determined by standard spot inoculation method 3,4,78,86 . The antimicrobial compounds were dissolved in methanol at a concentration of 200 µg/ml, and then, various concentrations (0.098-50 µg/ml) were prepared in Mueller-Hinton broth (MHB) by following the two-fold serial dilution method in microtiter plates (columns 2-11). Levofloxacin (5 µg/ml) was used as a control antibiotic (column 1). A 100 µl suspension of each test organism (0.5 McFarland) was dispensed in its respective well in columns 1-11. Each well in columns 1-11 contained an equal volume of the test antibiotics and suspensions of the test bacteria. Following the addition of the bacterial suspensions, the plates were incubated at 35 °C for 18-24 h. After incubation, 30 μl of sterile resazurin dye (0.015% w/v) was dispensed into each well of columns 1-11. The plates were kept at room temperature for 5 h. After 5 h, the results of MIC were recorded. The prepared concentrations of the antimicrobial compounds were evaluated for their antimicrobial efficacy against selected test organisms. The lowest concentration of the tested antimicrobial compound showed no color change from blue to pink was considered MIC. The lowest concentration of the tested antimicrobial compound showed no isolated colony on the inoculated plate considered MBC. The results were expressed in µg/ml.

Statistical analyses.
The statistical software package Minitab 19.2020.1 was used to analyze the experimental design applied in RSM during optimization of fermentation conditions for maximum yield of antibiotic. A One-Way ANOVA statistical test statistically analyzed the results of primary antimicrobial activity of the isolated antimicrobial compounds K 1 and K 2 to determine the statistical differences among the means of groups (tested bacteria and isolated actinomycins). The post hoc test (Tukey method) was performed to determine the significance of interactions among the means of groups, where p = 0.05 was considered as statistically significant. The statistical analyses were performed with SPSS software, version 20.0 (IBM, USA) 52,87 .
Molecular docking studies. X-ray crystal structures of several protein targets known to be associated with antimicrobial biological activity elicitation and employed in antimicrobial drug discovery and development, e.g., DNA gyrase, dihydropteroate synthase, dihydrofolate reductase, glucosamine-fructose-6-phosphate aminotransferase, isoleucyl-tRNA synthetase, tyrosyl-tRNA synthetase, penicillin-binding protein-1a, and UDP-Nacetylmuramoyl-L-alanine: D-glutamate ligase were retrieved from the Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB, http:// www. rcsb. org/ pdb/ home/ home. do) with PDB IDs 3TTZ, 2VEG, 3SRW, 2VF5, 1JZQ, 1JIJ, 3UDI and 2X5O. Also, Actinomycin D co-crystallized with DNA fragment (PDB ID: 1MNV) was also included in the study. Biovia Discovery Studio Visualizer 2020 and MGLTools 1.5.6 were used for the preparation of receptors 88 . All the co-crystalized ligands, water molecules, and cofactors were deleted, and Gasteiger charges were added to each receptor individually. Two-dimensional chemical structures of the compounds K 1 and K 2 were drawn in ChemDraw Ultra and converted to their three-dimensional coordinate by using the Chem3D Ultra program, energy minimized by MM2 method saved in PDB format. All non-polar hydrogens of the ligands were merged, and rotatable bonds were defined in MGL Tools 1.5.6. AutoDock Vina 1.1.2 was used for molecular docking simulation using default protocol with exhaustiveness adjusted to 12 65 . In each receptor, a grid box having dimensions of 30 points in all directions was built with a grid spacing of 1 Å at the center of respective co-crystallized ligands. At the end of the docking computation, the best poses were selected from the top ten models from each target by examining their binding energy (ΔG binding, kcal/mol) and non-bond interactions profile 89 www.nature.com/scientificreports/