Mature leaves of rice plants (Basmati rice variety HUBR 10-9) growing in full sunlight for the whole day were collected in the month of October, 2014 from the Agricultural farm of the Institute of Agricultural Sciences, Banaras Hindu University, Varanasi, Uttar Pradesh (25° 21′N, 83° 1′E and ~76 m above mean sea level) situated in the eastern gangetic plains of India. It is pertinent to mention that rice is mainly grown in rain fed areas during winter (Rabi) throughout India and the crops attain flowering stage during the month of October. For the isolation of bacteria, disease-free fresh leaves (0.5 g) were cut into small pieces and dipped in 10 mL sterile normal saline (0.9% NaCl) in a centrifuge tube. Tubes were vortexed vigorously to detach the associated epiphytic bacteria from the leaf surface. The resulting suspension was centrifuged at 2,000 × g for 5 min and the supernatant was collected in a separate tube. Different dilutions of the supernatant were made and 20 μL from each dilution plated on Luria–Bertani (LB) solid agar–agar medium. Plates were incubated at 37°C in a bacteriological incubator. Plates from each dilution were examined after 48 h of incubation and colonies showing distinct morphology in terms of shape, size, color, and texture were selected. Five colonies with distinct morphotypes were picked up and restreaked on fresh LB agar plates. Pure cultures of all the five putative isolates were obtained by repeated sub-culturing of a single colony on LB agar-agar medium. Growth of all the five isolates was also tested in liquid LB medium at 37°C in a shaker at 120 rpm (Orbitek LT, Scigenics Bioteck. Pvt. Ltd, India).

The source of UV-B radiation was a UV-B lamp (Cat No. 3-4408, Fotodyne Inc., USA), emitting main output at 312.67 nm. UV-B intensity was measured by a Black-Ray Long Wave Ultraviolet Intensity Meter (UVP Inc., USA). The desired UV-B intensity was obtained by adjusting the distance between the UV lamp and test samples. Percentage survival of each of the isolates after UV-B radiation exposure was determined by colony count method. For determining percentage survival, twenty mL of actively growing liquid culture of each isolate was transferred in five Petri dishes (diameter 100 mm, with open lids) and exposed to UV-B radiation (2 Wm^-2) for the desired time periods (1, 2, 3, 4, 6, and 8 h). Exposure of cultures to 2 Wm^-2 of UV-B radiation was based on the fact that the intensity of UV-B in solar radiation averaged about 2.3 Wm^-2 at global level (Madronich et al., 1998). Additionally, the low intensity of UV-B shows delayed effects and total killing of cyanobacteria/bacteria occurs after prolonged exposure, we had been performing experiments by using 2-2.5 Wm^-2 of dose in all the experiments (Singh et al., 2013; Babele et al., 2015). Cultures were gently magnetically stirred to facilitate uniform exposure. Fifty microliter aliquots were withdrawn from each plate at known time intervals after UV-B treatment and spread on solid agar plates. Colonies appearing after 48 h of growth were counted in a colony counter and percentage survival was calculated. Unless otherwise stated, all the experiments were performed with log phase cultures having an initial dry weight of approximately 0.15 mg mL^-1.

Intracellular ROS production was detected by means of the probe 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) (Santos et al., 2012). Ten milliliter culture of the selected isolate was exposed to UV-B radiation (2 Wm^-2) for different time periods up to 4 h. Irradiated and control cultures were harvested by centrifugation at 8,000 × g for 10 min at 4°C and washed with 10 mM potassium phosphate buffer. Pellets were suspended in 10 mL of 10 mM potassium phosphate buffer supplemented with 500 μM of DCFH-DA. Samples were incubated in the dark at 37°C for 1 h followed by brief sonication. One hundred μL of above samples from each set were taken out and protein content was determined by the Bradford method (Bradford, 1976). Fluorescence of the samples was measured with a spectrofluorometer (Varian, Inc., USA) with an excitation wavelength of 485 nm and emission wavelength of 519 nm. The fluorescence intensity at 519 nm was corrected against blank controls (cell extract without DCFH-DA) and thereafter normalized to protein content. ROS production is based on per mg protein.

Whole cell protein was extracted as described previously (Kurian et al., 2006). Cells were harvested from unexposed (control) and UV-B exposed (treated) cultures (1, 2, 3, and 4 h exposure) by centrifugation at 10,000 × g for 10 min at 4°C in a Sigma Refrigerated Superspeed Centrifuge (Sigma, USA). The resulting pellets were suspended in 10 mM HEPES buffer (pH 7.2) supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) and cells were broken by repeated freezing and thawing followed by mild sonication. The resulting suspension was centrifuged at 10,000 × g at 4°C for 10 min to remove cell debris. The supernatant obtained mainly contained cytoplasmic proteins. Chilled 10% (final concentration) trichloroacetic acid (TCA) in acetone was added to the supernatant and kept at -20°C for 2 h for the precipitation of proteins. Precipitate was collected by centrifugation at 14,000 × g at 4°C for 30 min. The pellet obtained was washed thrice with chilled acetone (95% v/v). Finally the sample was air dried and kept at -20°C till further use. If needed, salts and insoluble impurities, if any, were removed by a 2-D Clean-Up kit (GE Healthcare, UK). The protein content present in each sample was measured by Bradford method.

Forty microgram proteins of UV-B irradiated (1, 2, 3, and 4 h) and control cultures were loaded on a 12% SDS-PAGE and electrophoresed using SE 600 Ruby multiple gel-electrophoresis system (GE Healthcare, UK). Electrophoresis was performed at 50 mA for 3 h. Gel was stained with Coomassie Brilliant Blue R-250 (0.05% Coomassie Brilliant Blue R-250, 50% methanol and 10% acetic acid) for 12 h and destained with destaining solution (30% methanol and 10% acetic acid) for 24 h. The image was captured by means of the AlphaImager Gel Documentation unit (Alpha Innotech, USA).

2-Dimensional gel electrophoresis was performed with the proteins of control culture and 3 h UV-B exposed culture. The first dimension isoelectric focusing (IEF) was performed using Immobiline DryStrip (pH 4–7, 13 cm; GE Healthcare Bio-Sciences AB, Sweden). Protein sample (500 μg) dissolved in rehydration buffer (7 M urea, 2 M thiourea, 2% (w/v) CHAPS, 0.3% (w/v) DTT, and 0.5% (v/v) IPG buffer) was loaded onto each IGP strip and rehydrated for 14 h at 4°C in the dark. After rehydration, IEF was performed by the Ettan IPGphor 3 IEF system (GE Healthcare, UK) using the following steps; 2 h at 100 V, 2 h at 200 V, 1 h at 500 V, 1 h at 2000 V, 1 h at 3000 V, 1:30 h at 6000 V and 1:30 h at 8000 V until a total of 60000 Vh was attained. Equilibration of the IPG strips was done as described previously (Nouwens et al., 2002). The second dimension separation was carried out on 12% SDS-PAGE at 25 mA/gel using SE 600 Ruby multiple gel electrophoresis system (GE Healthcare, UK). Three gels were run for each protein sample of UV-B treated and untreated control cultures. Gels were stained with Colloidal Coomassie Blue G-250 (Candiano et al., 2004) and images were captured using the AlphaImager Gel Documentation unit (Alpha Innotech, USA).

The gels were scanned and protein spots analyzed by PDQuest^TM software version 8.0.1 (BioRad Laboratories, USA) for spot detection, quantification, background substraction, and spot matching between various gels. Spot quantification in the control and treated gels was performed by spot volumes (intensity × mm²) as described by Agrawal et al. (2014). The relative spot volume with respect to the UV-B treated and untreated control samples at desired time was compared by means of Student’s t-test. P-values less than 0.05 were considered statistically significant.

Differentially expressed proteins were identified by MALDI-TOF MS/MS (Matrix-assisted laser desorption/ionization-time of flight mass spectrometry). Selected protein spots were excised from the gel by manual picking using sterile one-touch spot picker and placed into separate microcentrifuge tubes. Trypsin digestion of the gel spots was done by the method of Shevchenko et al. (2006). MALDI-TOF MS analysis of protein spots was done at the Interdisciplinary School of Life Sciences, Banaras Hindu University, Varanasi. Data obtained were analyzed by Flex Analysis 3.3 and BioTools softwares (Bruker Corporation, Germany).

The proteins were identified by comparing peptide mass fingerprints at the NCBInr database using the Mascot search engine¹. Search parameters allowed for oxidized methionines, carbamidomethylation of cysteines, and one missed cleavage site of trypsin with mass accuracy of ±30 ppm. Identification was based on the first ranking result and Mascot scores of >74 which indicated that the hits were significant.

UV-B responsive hypothetical proteins identified by MALDI-TOF MS analysis were further characterized by using bioinformatics tools. Protein BLAST (Basic Local Alignment Search Tool) was used to search similar protein sequences with known functions available at NCBI database. Top hits were selected and further analysis was performed using ClustalW to find the alignment of functional residues of proteins of known function with the sequence of hypothetical proteins (Larkin et al., 2007). Physico-chemical properties of proteins were determined using the ProtParam tool² and database of Enterobacter cloacae subsp. cloacae strain 08XA1 as reference (Yan et al., 2012). InterProScan (version 5.0) server³ was used for the elucidation of protein family. Similarity search of both the hypothetical proteins was done by BLASTP. Phylogenetic analysis of the hypothetical proteins was carried out by using MEGA 6.06 software and tree was constructed by UPGMA (Unweighted Pair Group Method with Arithmetic Mean) method. Structure prediction of hypothetical proteins was done using EasyModeller version 4.0 (Kuntal et al., 2010). Three templates (PDBID-3SVL A.PDB- chromate reductase from Escherichia coli, 3S2Y A.PDB-chromate/uranium reductase from Gluconacetobacter hansenii, and 4H6P A.PDB-chromate reductase from G. hansenii) showing query cover above 93%, identity above 64%, and positives above 72% with respect to the hypothetical protein gi 503938301 were used for structure prediction. Validity of the structure was checked by the Ramachandran plot and quality of the predicted structure was checked by using PDBSum server.

Five proteins, out of the 13 proteins identified by MALDI-TOF MS/MS, were selected for gene expression analysis by qRT-PCR for testing their regulation under UV-B stress. Among the selected genes, two were hypothetical protein genes (cp011650 and cp002886) and three known protein genes namely, DNA gyrase inhibitor (sbmC), peroxiredoxin (ahpC), and outer membrane protein (ompC). 16S rRNA was used as a bacterial reference gene (McMillan and Pereg, 2014). All the primer sets were designed by Primer 3 software and specificity was checked by Primer-BLAST (Table 1). Commercially synthesized primers (Integrated DNA Technologies Inc., USA) were used. Total RNA of the UV-B treated and control culture was isolated by means of RNA isolation kit (Ambion Life Technologies, USA) as per the manufacturer’s instructions. DNA was removed by DNase treatment. Complementary DNA was synthesized by employing cDNA synthesis kit (BioRad Laboratories, USA) according to manufacturer’s protocol. cDNA quantification was done by using Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific, USA). qRT-PCR was performed on a MiniOpticon Real-Time PCR System (Bio-Rad Laboratories, USA) using SsoFast EvaGreen Supermix (Bio-Rad). Total reaction volume of 20 μL included: 10 nM of each forward and reverse primers, cDNA template (final conc. 20 ng) and the SsoFast EvaGreen Supermix (according to the manufacturer’s instructions). Each run included a non-template control (NTC). Real-time amplification for each gene including control and treated samples was performed in triplicate under identical PCR conditions. Thermal cycle was set as: 1 cycle at 95°C for 5 min, followed by 40 cycles at 95°C for 15 s, 53°C for 20 s, and 70°C for 30 s. The melting curve analysis was performed for all the assays; and the melting temperature of the product was determined using the melting curve program: 50–95°C, with a heating rate of 0.05°C per second and continuous fluorescence measurement. The Raw C_t values were extracted and analyzed using the CFX Manager Software (Bio-Rad, USA). The C_t values were automatically identified using the “Second Derivative Maximum Method” by the system software set with default parameters. The transcript expression level was normalized by the candidate reference gene (16S rRNA) relative to 0 h using the 2^-ΔΔCt method (Livak and Schmittgen, 2001). All the assays of qRT-PCR were performed under identical conditions for all the target genes.

All experiments were conducted in triplicate under identical conditions. Mean values and standard deviations were determined from three replicates of each treatment. A one-way ANOVA (analysis of variance) helped in confirming the significance of data according to Duncan’s multiple range test (DMRT) at P ≤ 0.05. OriginPro 8.5.0 (Origin Lab Corp., USA) software was used for the DMRT. Unless otherwise stated, values show the mean ± SD (n = 3).

Five bacterial isolates (designated as RL-1 to RL-5) based on morphological characters and certain biochemical tests were isolated from the phyllosphere mature leaves of rice plant. Identification of all the five isolates was made by 16S rRNA gene sequencing and accession numbers were obtained. These isolates were designated as: RL-1 (Microbacterium sp. SBT-6, KU376416), RL-2 (Microbacterium sp. SBT2, KU356922), RL-3 (Enterobacter cloacae RLI, KT161961), RL-4 (Pantoea sp. SBT-4, KU376415), and RL-5 (Agrobacterium sp. SBT-5, KU376420). Before assessing the impact of UV-B stress on protein profile, percentage survival of all the five isolates exposed to UV-B radiation (2 Wm^-2 intensity) for varying time periods was determined to assess the tolerance level. Results showed decrease in percentage survival of all the isolates with increasing duration of UV-B exposure and complete killing of four isolates occurred at 6 h of continuous exposure to UV-B. However, one isolate namely, E. cloacae RLI (RL-3) showed complete killing after 8 h and retained ca. 52 % survival after 2 h of UV-B exposure (Figure 1). The observation that E. cloacae RLI shows higher level of tolerance to UV-B prompted us to select this isolate for all other experiments.

Once it became apparent that UV-B exposure causes loss of survival of E. cloacae, it was desirable to test the generation of ROS as its induction under oxidative stress has been well documented. As expected, ROS generation was enhanced in cultures exposed to UV-B radiation, and the level increased with increasing duration of exposure. The maximum increase was noted at 3 h as is evident from the increase in fluorescence of DCF (Figure 2). However, after 3 h of continuous exposure to UV-B, the fluorescence decreased possibly due to loss of cell viability or death.

That, UV-B exposure causes loss of survival and induction of ROS generation as a result of oxidative stress aroused our interest to test alterations, if any, in the whole cell proteome of UV-B treated cultures. Accordingly, SDS-PAGE of proteins from the cultures exposed to UV-B radiation for varying time periods and control (unexposed) was performed. Analysis of gels based on electrophoretic pattern indicated alterations in the protein profile of UV-B treated cultures as compared to control cultures (Figure 3). Certain new protein bands appeared and a few were lost after exposure of cultures to UV-B radiation for 3 h. Additionally, the intensity of certain protein bands also decreased with increasing duration of exposure.

To gain a better insight of UV-B stress-induced changes in protein profile, 2-D gel electrophoresis was performed to find out changes in the proteome. It is evident from the analysis of gel (Figures 4A,B) that changes in both number and intensity of protein spots indeed occur upon exposure of cultures to UV-B. Analysis of gels showed 378 protein spots in untreated control and 384 spots in UV-B treated cultures of E. cloacae. Besides some changes in total number of spots, analysis of gels showed differential expression of several spots in UV-B treated cultures (Figure 4C). The data of scatter plot indicate that a number of protein spots show higher and/or lower level of expression in response to UV-B stress (Figure 5). Changes in expression level of proteins prompted us to focus our attention mainly on the protein spots which were showing either up- or down regulation on the basis of spot-to-spot comparison. Accordingly, thirteen protein spots showing more than twofold higher and/or lower level of expression from UV-B treated cultures as compared to untreated control (p < 0.05) were selected for MALDI-TOF MS analysis. Among the 13 identified proteins, 11 belonged to up-regulated group including two hypothetical proteins (Figure 4C) and 2 to down-regulated group. Figure 4C shows expression level of all the 13 identified proteins.

In functional studies, we compared the protein sequences with the secondary (or derived) protein databases that contain information on motifs, signature, and protein domains. Highly significant hits against different protein databases allowed us to derive the biochemical functions of query protein. Table 2 summarizes the key findings related to bacterial processes responding to UV-B treatment based on the identification of representative proteins. All the 13 proteins identified in this study were sorted according to the functional categories defined by bacterial protein database. All the identified proteins were placed in eight different functional groups which included carbohydrate metabolism, energy metabolism, translation and transcription, fatty acid biosynthesis, nitrogen metabolism, antioxidative enzymes, membrane proteins and hypothetical (unknown) proteins (Table 2). There were two unknown proteins (hypothetical) namely, gi 779821175 and gi 503938301 among the 13 identified proteins.

Both the hypothetical proteins (gi 779821175 and gi 503938301) showed up-regulation under UV-B radiation stress. Till date, these two proteins have not been identified and characterized and their function is not known. The results of this study show that the protein gi 779821175 is encoded by a gene of E. cloacae located at locus AB285_13020 whereas gi 503938301 is encoded by a gene located at locus EcWSU1_04540. Domain analysis showed that gi 503938301 has a functional domain and shows similarity with FMN reductase superfamily. On the other hand, gi 779821175 has no functional domain and thus may be a structural protein. Physico-chemical properties of the protein gi 779821175 showed that it is rich in isoleucine (14.4%) and leucine (13.5%) whereas gi 503938301 is rich in leucine (10.1%) and glycine (9.6%). Details of functional domain analysis derived by InterProScan are presented in Table 2.

From the BLAST results, it was noted that gi 779821175 does not show significant similarity with the related sequences of known functional proteins of any bacteria available in the NCBI database. It showed maximum identity (27%) with the putative membrane protein of Burkholderia sp. ABCPW 1. However, protein gi 503938301 showed 99% similarity with the FMN reductase enzyme of Enterobacter. With a view to revealing homology and phylogenetic relationship of gi 779821175 and gi 503938301 with the proteins of other species available in the NCBI database, a phylogenetic tree was constructed using reference strain E. cloacae. The phylogenetic tree indicates that the protein gi 779821175 shows 99% bootstrap value with the hypothetical protein of Enterobacter hormaechei (Figure 6A) but only 28% bootstrap value with the putative membrane protein of Burkholderia pseudomallei PB08298010. In the case of protein gi 503938301, it showed 99% bootstrap value with FMN reductase enzyme of Enterobacter and 90% bootstrap value with the hypothetical protein of Citrobacter sp. 30_2 (Figure 6B).

An attempt was made to construct 3-D model of both the hypothetical proteins using sequence similarity and available structures in the NCBI PDB database. A 3-D model of the gi 503938301 was constructed (Figure 7A) by using sequence identity and positive values with available templates in the NCBI PDB database. Additionally, validity of the structure was checked by the Ramachandran plot (Figure 7B). Analysis of plot generated for gi 503938301 revealed 93.8% residues in the most favored region, 6.2% in additional allowed region and no amino acid in disallowed region (Figure 7B). The values obtained are well within normal range for a high quality model. The best validated structure of gi 503938301 has been deposited in the PMDB database with PMDB-ID PM0080456. A 3-D model of the protein gi 779821175 could not be constructed as no suitable template was available in the PDB.

To corroborate the findings of up-and/or down regulation of protein expression, transcript analysis of two hypothetical protein genes (cp011650 and cp002886) and three known protein genes namely, DNA gyrase inhibitor (sbmC), peroxiredoxin gene (ahpC), and outer membrane protein (ompC) was made after exposure of cultures to UV-B radiation for varying time periods. qRT-PCR results revealed significant differences in the level of expression of target genes in the control and UV-B treated cultures. It is evident that C_t values of each target gene from cultures exposed to UV-B for varying time periods differ, suggesting differences in the level of expression (Figure 8). Of the five genes, cp011650, cp002886, sbmC, and ahpC showed more than twofold increase in expression at 2 h of UV-B exposure as compared to untreated control. In fact, the level of expression of the above four genes was higher even after 3 h of continuous exposure of cultures to UV-B, the maximum level being in the case of ahpC followed by cp002886. However, the level of expression decreased significantly after 4 h of exposure excepting cp011650. Notably, the expression of OmpC, a membrane protein gene showed gradual decrease with increasing duration of UV-B exposure, the level being more than half after 3 h of UV-B exposure in comparison to untreated culture (Figure 8). As expected, there was no change in the level of expression of reference gene (16S rRNA) irrespective of increase in time period.