Hypericum perforatum (Hypericaceae) and Ziziphora capitata (Lamiaceae) plants were collected during the summer (June 2013, the plant’s flowering stage) from Chatkal Biosphere Reserve of Uzbekistan, western part of Tien Shan mountain (41°08′ N; 69°59′ E). This biosphere reserve is situated in the Tashkent Region within the Chatkal mountain range (1.110–4.000 m above sea level) of the West Tien-Shan Mountains and is unique for its significant role in biodiversity conservation and ethnobotany. The climate is characterized by average annual temperatures ranging from 20 to 25°C with increased annual precipitation from plains to mountains, reaching 700–800 mm.

The aerial parts of H. perforatum and Z. capitata were dried in the laboratory excluding direct sun light at room temperature for 6–7 days and ground into a fine powder by mortar and pestle. Approximately, 10 g of plant powder was extracted with 50 ml of methanol for 24 h in a dark room temperature. Subsequently, the solvent was evaporated in a rotary vacuum evaporator at 40°C and re-suspended in dimethyl sulfoxide (DMSO). The homogenate was filtered through Whatman No. 1 filter paper, centrifuged at 5000 g for 15 min and sterilized by filtration through 0.22-μm sterile filters (Millipore, Bedford, MA, USA). The filtrates were stored at -4°C and used for in vitro screening of antimicrobial activity.

The extracts were individually tested against the following pathogenic microorganisms: Klebsiella oxytoca 6653, K. pneumoniae 40602, K. aerogenes NCTC 8172, Citrobacter freundii 82073, Staphylococcus aureus MRSA 16, Enterococcus faecalis NCTC 775, Providencia rettgeri NCIMB 9570, Pseudomonas aeruginosa NCTC 6749, Escherichia coli NCTC 9001 and Fusarium solani, Fusarium oxysporum, and Alternaria alternata. Reference strains and clinical isolates were obtained from the Department of Microbiology, Manchester Metropolitan University, UK, and the National Culture Type Collection (NCTC), UK. The fungal strains were obtained from the Department of Microbiology and Biotechnology, National University of Uzbekistan. Each plant extract was dissolved in dimethyl sulfoxide (DMSO), sterilized by filtration using a sintered glass filter, and stored at 4°C. The antimicrobial activity of the extracts was tested using the agar well-diffusion method. Microorganisms were grown overnight at 30°C in Mueller-Hinton Broth (Oxoid, Basingstoke, UK) supplemented with 5% horse blood, and 100 μl of suspension containing 10⁶ CFU ml^-1 of bacteria was spread on the surface of Mueller-Hinton agar plates. Wells with 6-mm diameters were cut off and filled with 50 μL of each extract (10 mg ml^-1). Ampicillin (Sigma-Aldrich, Steinheim, Germany) (0.5 mg ml^-1), nystatin (Sigma-Aldrich, Steinheim, Germany) (1 mg ml^-1) and DMSO were used as controls. Fungal strains were grown on potato dextrose agar plates (PDA; Difco Laboratories, Detroit, MI, USA) at 28°C for 5 days. Small piece of fungal culture were placed in the middle of Petri plates. Each antimicrobial assay was performed in triplicate. The plates were incubated at an appropriate growth temperature for 2 days for bacterial strains (37°C) and 4 days for fungal strains (30°C). The assessment of antimicrobial activity was based on the measurement of inhibition zones on the agar surface around the well.

Three plants from each species of H. perforatum and Z. capitata including roots (20–30 cm depth) were randomly collected about 1 m apart from each other from an area of 100 m² in the Chatkal Biosphere Reserve. The whole plants, along with root systems, were wrapped in plastic bags, and brought to the laboratory on same day and immediately stored at 4°C. The isolation of bacterial strains was carried out on the next day to minimize storage effects.

The root systems of the collected plants were separated from the shoots, soil adhering to the roots was removed and roots were carefully washed under running water, taking care to minimize root injury. Three plants of each species were used to determine the number of bacterial colonies cultured from the root tissue. For the bacterial isolation, root tissues were pooled from each of three replicate plants. The roots were surface sterilized by immersion in 70% (v/v) ethanol, following by shaking in 5% (w/v) sodium hypochlorite solution for 5 min. Subsequently, the roots were rinsed in sterile distilled water six times. To test the efficiency of sterilization, the sterile roots were incubated in TSA medium for 2 days at 28°C, and no infestation was observed.

Sterilized roots were weighed aseptically (1 g) and macerated in a mortar employing phosphate buffered saline (PBS) (20 mM sodium phosphate, 150 mM NaCl, pH 7.4) in a laminar air flow cabinet. The extracts were placed in a tube containing 9 ml sterile PBS and shaken with a vortex for 1 min. The supernatant was collected and serially diluted (10¹–10⁵) in PBS, and 100 μl from appropriate dilutions were spread on Tryptic Soy Agar (TSA, Difco Laboratories, Detroit, MI, USA) plates in triplicate. The plates were incubated at 28°C, and colony forming units (cfu) g^-1 root tissue were determined on the third day. A representative number of colonies that exhibited differentiable colony morphologies were picked from plates and were re-streaked for the purification of the isolates. The pure bacterial cultures were preserved on plates at 4°C for the further analyses. In addition all bacterial isolates were stored in Tryptic Soy broth (TSB) (Difco) with 30% glycerol at -80°C.

The identification of bacterial isolates was performed by matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) as described previously (Egamberdieva et al., 2016). The sample preparation was performed according to the ethanol/formic acid extraction protocol recommended by Bruker Daltonics (Bremen, Germany) and was described in Egamberdieva et al. (2016). Briefly, the isolates were cultured on TSA medium (Difco Laboratories, Detroit, MI, USA) for 24 h, and approximately 10 mg of cell mass was suspended in 300 μL water (LC–MS CHOMASOLV®; Honeywell) and vortexed to generate a homogenous suspension. The suspension was mixed with 900 μL ethanol (≥99.8% GC; Sigma-Aldrich) and centrifuged. The pellet was resuspended in 50 μL 70% formic acid (v/v) and subsequently carefully mixed with 50 μL acetonitrile. After centrifugation, aliquots of 1 μL supernatant were placed immediately on spots of a MALDI target. Each spot was allowed to dry and subsequently overlaid with 1 μL of matrix (α-ciano-4-hydroxycinnamic acid in 50% aqueous acetonitrile containing 2.5% trifluoroacetic acid). Mass spectra were acquired using a MALDI-TOF MS spectrometer in a linear positive mode (Microflex^TMLT, Bruker Daltonics, Bremen, Germany) in a mass range of 2–20 kDa. A bacterial test standard (BTS, Bruker Daltonics, Bremen, Germany) was used for instrument calibration. The raw spectra were imported into the MALDI Biotyper^TM software and then processed and analyzed using standard pattern matching against the reference spectra in the MALDI Biotyper^TM reference database (version 3.0, Bruker Daltonics, Bremen, Germany). A calculated matching score (score value) provided a measure of the probability of a correct classification.

The production of IAA (indole 3-acetic acid) was determined as described by Bano and Musarrat (2003). The IAA concentration in culture was calculated using a calibration curve of pure IAA as a standard. The cellulose-degrading ability of bacterial isolates was analyzed by streaking inocula on cellulose (Sigma-Aldrich, St. Louis, MO, USA) Congo-Red agar media as described by Pratima et al. (2012). Furthermore, β-1,3 glucanase activity was tested using the substrate lichenan (Sigma-Aldrich, St. Louis, MO, USA) in top agar plates (Walsh et al., 1995), and protease activity was determined by using 5% skimmed milk agar plates (Brown and Foster, 1970). The production of HCN by bacterial isolates was measured using the protocol described by Castric (1975).

The bacterial isolates were tested in vitro for their antagonistic activities against the following pathogenic fungi: Fusarium oxysporum f.sp. radicis-lycopersici (Forl), F. solani, F. culmorum, Gaeumannomyces graminis pv. tritici (Ggt), Alternaria alternata, and Botrytis cinerea and the oomycete Pythium ultimum. The bacterial isolates were grown in TSB broth for 3 days, and 50-μl bacterial cultures were dropped into a hole of PDA plates (4 mm in diameter). Fungal strains for inoculation were grown in peptone dextrose agar (PDA) plates at 28°C for 5 days. Disks of fresh cultures of the fungus (5 mm diameter) were cut out and placed 2 cm away from the hole filled with bacterial filtrate. The plates were sealed with Parafilm^®M and incubated at 28°C in darkness until the fungi had grown over the control plates without bacteria. Antifungal activity was recorded as the width of the zone of growth inhibition between the fungi and the bacteria tested.

Bacterial isolates with antagonistic activity against the majority of tested fungal pathogens, were tested for their ability to control tomato root rot caused by F. oxysporum f.sp. radicis-lycopersici. For the inoculation of soil, F. oxysporum was grown in PDA plates for 5 days. Small pieces of agar from the growing edge of the colony were homogenized and used to inoculate 300 ml of Chapek-Dox medium, which was kept under aeration (110 rpm) at 28°C. After 3 days, the spore suspension was filtrated with sterile glass wool to remove the mycelium. The concentration of spores in the inoculum was adjusted to 10⁷ spores ml^-1 by microscopic enumeration with a cell-counting haemocytometer and mixed thoroughly with potting soil to obtain a concentration of approximately 10⁷ spores kg^-1 soil. The tomato seeds of the cultivar Fuji Pink (Sakata, Japan) were sterilized by stirring with 70% ethanol for 5 min and in household bleach (adjusted to approximately 5% sodium hypochlorite) for 3 min. Subsequently, the seeds were washed several times with sterile distilled water. After germination in sterile Petri plates, the seeds were placed in a bacterial suspension of 1 × 10⁸ CFU ml^-1 prepared as described above and shaken gently for 10 min. The inoculated seeds were sown in plastic pots, and each treatment contained four groups of 24 plants. The plants were grown in a growth chamber under controlled conditions (16 h light, 8 h dark), at temperature light 28°C, dark 20°C and relative humidity 60%. After 3 weeks, the plants were removed from the soil, washed and examined for foot and root rot symptoms as indicated by browning and lesions. Roots without any disease symptoms were classified as healthy.

To test whether bacterial isolates were capable of stimulating plant growth, a pot experiment was conducted in the controlled plant growth chamber. Tomato seeds (Solanum lycopersicum. cv. Fuji Pink, Sakata, Japan) were surface-sterilized as described above. Surface-sterilized seeds were transferred to plastic Petri dishes and germinated for 4 days in a dark room at 25°C. The bacterial isolates were grown overnight in TSB, and 1 ml of each culture was pelleted by centrifugation (10.000 × g for 10 min). Cell pellets were washed with 1 ml PBS, re-suspended in PBS and cell suspensions were adjusted to OD_{620 nm} = 0.1 (0.2 for Bacillus and Arthrobacter) that correspond to a cell density of about 10⁷–10⁸ cells ml^-1. Germinated tomato seeds were placed in the bacterial suspension with a sterile forceps and shaken gently. After 10 min, the inoculated seeds were aseptically planted into a plastic pot filled with potting soil (N 250 mg l^-1, P 120 mg l^-1, K 700 mg l^-1, pH 6.0, Floragard GmbH, Germany) to a depth of approximately 1.5 cm. Non-inoculated plants were used as negative controls. Each experiment included six plants per treatment with three replications (total 18 plants) and pots were set-up in a randomized design. Plants were grown in a growth chamber under the conditions described above.

The data were subjected to one-way analysis of variance (ANOVA) in the software package SPSS-22 statistical software (SPSS, Inc., Chicago, IL, USA). Mean comparisons were conducted by the least significant difference (LSD) (P = 0.05) test.

The inhibitory effect of extracts from Z. capitata and H. perforatum, which were tested against diverse enteric pathogens (A. baumanii 60649, K. oxytoca 6653, K. pneumoniae 40602, K. aerogenes NCTC 8172, C. freundii 82073, S. aureus MRSA 16, E. faecalis NCTC 775, Proteus rettgeri NCIMB 9570, P. aeruginosa NCTC 6749, and E. coli NCTC 9001) at a concentration of 10 mg ml^-1, resulted in different extents of inhibition (Table 1). The strains A. baumanii 60649, E. coli NCTC 9001, E. faecalis NCTC 775, K. oxytoca 6653, K. pneumoniae 40602, P. aeruginosa NCTC 6749, and S. aureus MRSA 16 were inhibited by the extract of H. perforatum. However, extract of Z. capitata did not exhibit any potential antibacterial activity against the 10 tested pathogens. Extracts of H. perforatum exhibited potential antifungal activity against F. oxysporum and A. alternata, whereas the extract of Z. capitata did not exhibit any inhibitory activity against the tested fungal strains.

The total number of endophytic bacterial isolates in the root tissue of Z. capitata was significantly higher (4.5 ± 0.8 × 10³ CFU g^-1 of fresh root tissue) than in H. perforatum roots (2.6 ± 0.71 × 10³ CFU g^-1 of fresh root tissue). Isolates were chosen randomly from the dilution plates exhibiting different colonial morphology, forms, texture, and color from each plate. A total of 18 bacterial isolates were derived from H. perforatum and 15 isolates from Z. capitata. Taxonomic investigation by MALDI-TOF MS revealed that the majority of strains were identified with secure genus identification and probable species identification (Table 2). The endophytes from the root of H. perforatum were affiliated with nine genera, whereas 14 isolates were identified at the species level. Achromobacter was the predominant genus, which was followed by the genus Pseudomonas. Furthermore, isolates affiliated with the genera Arthrobacter, Bacillus, Erwinia, Pantoea, Serratia, and Stenotrophomonas were found. The most abundant species was Achromobacter piechaudii (S22a, S7a, S7) (Table 2). A total of five bacterial genera were isolated from the root of Z. capitata (Table 2). The most abundant isolates of Z. capitata were also identified as A. piechaudii (M11, M6, M31, M24, M41). Members of the genera Serratia, Stenotrophomonas, and Erwinia were not identified among the endophytes from Z. capitata.

All endophytes isolated from H. perforatum and Z. capitata were screened for multiple plant growth-promoting traits. Most of the bacterial isolates exhibited one or more plant growth-promoting activities. The production of phytohormone IAA by bacterial isolates is presented in Table 2. The highest level of IAA production was observed for Arthrobacter crystallopoietes S1 (19.8 μg ml^-1), A. piechaudii S7 (17.5 μg ml^-1), Achromobacter sp. S14 (15.2 μg ml^-1), Pantoea agglomerans S22 (12.3 μg ml^-1), and Bacillus cereus S40 (12.2 μg ml^-1), which were isolated from H. perforatum. Two isolates, Enterobacter cloacae M20 and P. agglomerans M13, isolated from Z. capitata also exhibited high IAA production in culture media (12.2 and 16.1 μg ml^-1, respectively). The presence of tryptophan did not stimulate auxin production in the majority of the isolates, whereas only four isolates from Z. capitata revealed an increase in IAA synthesis: Bacillus sp. S2 (15.5 μg ml^-1), P. agglomerans S22 (μg ml^-1), Serratia liquefaciens S26 (15.0 μg ml^-1), Stenotrophomonas sp. S9 (16.3 μg ml^-1) and two isolates, B. altitudinis M9a (49.9 μg ml^-1) and P. agglomerans M13 (52.5 μg ml^-1) from H. perforatum (Table 2). All isolates isolated from H. perforatum, except Achromobacter spanius S23, Bacillus sp. S2 and isolate S9, were able to produce one or more cell wall-degrading enzymes. In contrast, only four isolates from Z. capitata (A. piechaudii M41, Achromobacter sp. M19, E. cloacae M20 and isolate M8) were able to produce proteases, and only one isolate, A. spanius M18, produced cellulase and β-1,3-glucanase. HCN was not produced by any isolate from Z. capitata, whereas seven isolates isolated from H. perforatum were able to produce HCN (Table 2).

Antagonistic activity was recorded for endophytes against plant pathogenic fungi F. oxysporum f. sp. radicis-lycopersici, F. solani, F. culmorum, G. graminis pv. tritici, A. alternata, and B. cinerea and the oomycete P. ultimum. As presented in Table 3, all isolates from H. perforatum exhibited antagonistic behavior to one or more of the tested plant pathogenic fungi. The isolates A. crystallopoietes S1, A. piechaudii S7, Pseudomonas koreensis S25, Pseudomonas pseudoalcaligenes S24, and Stenotrophomonas sp. S9 were highly effective against six fungal pathogens and exhibited the highest inhibition of mycelial growth (Figure 1). Among the isolates from Z. capitata, only P. agglomerans M13 exhibited antagonistic activity against five fungal pathogens, but the in vitro inhibition of the mycelium was lower than that of the other isolates. In general, H. perforatum, which exhibited a broad spectrum of antimicrobial activity, supported a higher proportion of antagonistic endophytes compared with Z. capitata.

The bacterial isolates that exhibited antagonistic activity against a wide range of fungal pathogens in vitro were selected to evaluate their ability to suppress tomato foot and root rot caused by F. oxysporum f. sp. radicis-lycopersici in a pot experiment. In non-infested soil, the portion of diseased plants was 2%, whereas in the presence of the pathogen, the portion of plants that exhibited disease symptoms increased to 38% (Figure 2). The selected antagonistic bacterial isolates A. crystallopoietes S1, Bacillus sp. S2, B. cereus S40, P. koreensis S25, S. liquefaciens S26, and Stenotrophomonas sp. S9, exhibited a statistically significant (P < 0.05) disease reduction (up to 9%) compared with Fusarium-infected control plants (Figure 2). Several isolates, namely A. piechaudii S7, Pseudomonas putida S19, Pseudomonas thivervalensis S5, P. pseudoalcaligenes S24, and P. agglomerans M13 reduced disease incident, but the effects were not significant.

The antagonistic endophytic bacterial isolates were also effective on the growth of tomato plants under controlled conditions (Figure 3). Statistical analysis showed that growth stimulatory effects of the isolates A. crystallopoietes S1, A. spanius S23, Bacillus sp. S2, P. putida S19, and Stenotrophomonas sp. S9 increased plant biomass significantly (P < 0.05) between 30 and 41%. However, four strains namely A. piechaudii S7, B. cereus S40, P. koreensis S25, and P. pseudoalcaligenes S24 reduced plant growth response, leading to a decrease in plant dry biomass between 5.3 and 11.4% (Figure 3).