Samples of rhizosphere soil from Kalomegh (A. paniculata) fields were collected from 3 districts (Dhaka, Gazipur and Gaibandha) belonging to two different agroecological zones (AEZs) of Bangladesh. Among the three samples, two were collected from “Modhupur Tract” AEZ and another was from “Active Tista Floodplain” AEZ. Plant roots with adherent soil were taken from healthy plants, placed in sterile containers and transported to the laboratory as soon as possible. Before isolation, the roots were gently shaken to remove excess soil and vortexed for 10 min in 0.9% sterile saline. Samples were serially diluted with sterile saline. Aliquots from the diluted samples were plated onto nutrient agar. After incubation for 24 h at 30˚C, colonies formed on the nutrient agar plates were counted. The number of heterotrophic bacteria was calculated using the formula:

cfu/ml = (No. of colonies × dilution factor)/volume added on of culture plate.

Finally, the cfu/ml was converted to cfu/g of soil.

Morphologically dissimilar colonies were selected randomly from different plates for each sample and marked. A single touch of sterilized loop was taken and streaked on agar plate for pure culture. The streaked plates were incubated at 30˚C for 24 h. Well isolated single colony was selected for purification. The well isolated single colony was subjected to further purification through repeated plating by streaking method. Purified single colony was preserved and used for determining colony characteristics and identification.

The isolates were screened for pectinase activity by plate assay method. This was done by inoculating the organisms on modified MS medium [21] containing 0.3% KH₂PO₄, 0.6% Na₂HPO₄, 0.2% NH₄Cl, 0.5% NaCl, 0.01% MgSO₄・7H₂O, 1.5% agar and 0.2% pectin. The pectin containing MS medium was incubated at 30˚C for 48 h. Pectin utilization was detected by flooding the culture plates with freshly prepared Iodine-Potassium Iodide solution [22] . This solution forms a translucent halo zone where pectin depolymerization has occurred. A clear zone around the growth of the bacteria was indicated to pectinase activity. The enzyme activity was observed by measuring the diameter of clear zone around the colony in millimeter. Clear zone of each isolates were measured and recorded [23] .

The top 3 pectinase producing bacterial strains were provisionally identified based on morphological and biochemical characteristics. The colony characteristics (margin, elevation, form, surface, color, opacity and size) on the nutrient agar, growth pattern on selective media, microscopic characteristic after staining (cell shape, Gram reaction and spore formation) were recorded. A series of biochemical tests were performed to identify the bacteria [24] [25] [26] [27] .

The effect of pH for pectinase producing isolates was determined by culturing the bacterium in the media with different pH. The experiment was carried out at various pH 6, 7, 8 and 9. Optical density was measured at 600 nm by a Spectrophotometer (Specord 205, Analytik Jena, Germany). To assess the ability to tolerate higher salt concentration, overnight culture of test isolates were inoculated in nutrient broth containing 0%, 1%, 2%, 4%, 6% and 8% NaCl, respectively. The growth of isolated bacteria on different concentration of NaCl was measured by a Spectrophotometer at 600 nm.

Genomic DNA of the 3 most promising strains was isolated using the Maxwell DNA purification kit and Maxwell 16 extraction platform (Promega, USA). In brief, 50 μl of fresh bacterial sample of each strain was mixed with the lysis buffer and mixed thoroughly by vortexing. After incubation, the whole mix was transferred in the Maxwell 16 DNA purification cartridge. The cartridge was then loaded into Maxwell16 instrument and DNA was isolated following manufacturer’s instructions. Concentration of purified DNA samples was quantified by using NanoDrop (Thermo Scientific, USA).

The genomic DNA extracted from the bacterial strains was used to amplify the 16S rRNA gene. The primer pair 27F (5’-AGAGTTTGATCCTGGCTCAG-3’) and 1492R (5’-GGTTACCTTGTTACGACTT-3’) were used to amplify the gene by PCR. Amplification was carried out in an automatic thermocycler (Astec, Japan). Briefly, the 25 μl PCR reaction mixture contained 12.5 μl 2× GoTaq G2 hot start master mix, 0.1 μl of each primer together with 1.0 μl of DNA template. The PCR conditions were as follows: Initial denaturation at 95˚C for 4 min followed by 30 thermal cycling consisting of denaturation at 95˚C for 30 s, annealing at 49˚C for 30 s, extension at 72˚C for 90 s and a final extension at 72˚C for 5 min.

PCR products were loaded onto 0.8% agarose gel containing ethidium bromide solution (Promega, USA). DNA ladder (Promega, USA) was used as a DNA size marker. Electrophoresis was done at constant voltage of 100 V for 1.5 hour in a horizontal electrophoresis system (CBS Scientific, USA) following manufacturer’s protocol. After electrophoresis the gel was carefully removed and the DNA bands were observed under UV-transilluminator (Alpha Imager, USA) to confirm the presence of amplicon. Following electrophoresis, DNA band of desired size was excised from gel and placed in a 1.5 ml microcentrifuge tube. The PCR product was purified using the SV mini column according to the manufacturer’s instruction. The sequence of PCR products were determined by Sanger dideoxy chain termination method using an ABI 3700 Genetic Analyzer (1^st Base Laboratory, Malaysia).

The resulting DNA sequence was compared by basic local alignment search tool [28] . Based on the highest degree of similarity, identity for the bacterial isolates was assigned. For phylogenetic analysis, the 16S rDNA sequences were aligned with the Clustal W program [29] and the tree was constructed with the maximum likelihood method based on the Tamura-Nei model [30] integrated in the MEGA7 software [31] . The phylogenetic tree was tested with 1000 bootstrap replicates.

In the present study, bacteria were isolated from the rhizosphere of A. paniculata growing in three different districts of Bangladesh. The total heterotrophic bacterial counts were 1.56 × 10⁹, 1.53 × 10⁹ and 2.52 × 10⁹ cfu/g in the rhizosphere soil sample AP1, AP2 and AP3, respectively. The variation in the count of heterotrophic bacteria due to difference in location was not statistically significant. The total heterotrophic bacterial count observed in this study corroborates with previous report that the bacterial population in the rhizosphere ranges from 10⁸ to 10⁹ per gram of rhizosphere soil [32] .

From the nutrient agar plates 65 bacterial colonies were purified for further study. The colony characteristics of the purified isolates varied in margin, elevation, form, surface appearance, color, opacity and size. The isolates also varied in their growth pattern in nutrient agar slant. Out of the 65 isolates, 52 (80%) were Gram positive and 13 (20%) were Gram negative bacteria. Among the isolates 24 (36.9%) were spore forming and 41 isolates (63.1%) were non-spore forming bacteria. Appearance of diverse type of bacterial colonies on the nutrient agar plate indicates that the rhizosphere soil of A. paniculata promotes the survival and proliferation of both Gram positive and Gram negative as well as both spore forming and non-spore forming bacteria.

The isolates were also tested for their ability to grow on different types of selective and differential media (MacConkey agar, Mannitol Salt agar, Cetramide agar, Triple sugar agar etc.). Furthermore, isolates were subjected to different type of biochemical tests (Catalase test, MR-VP test, citrate utilization test, sugar fermentation test, nitrate reduction test etc.) for identification. Based on the morphological, cultural and biochemical characteristics isolates were provisionally identified as Bacillus, Pseudomonas and Micrococcus.

All of the 65 isolates were tested for their pectin degradation ability by plate assay method in the modified MS medium containing 0.2% pectin as sole source of carbon (Figure 1). Clear zone in plate were measured and recorded (Figure 2). Pectinase producing bacteria were present in all three soil samples. Among the 65 isolates of bacteria evaluated in this study, 62 (95.38%) isolates were pectinase enzyme producer. Tariq and Latif [33] isolated 62 bacterial isolates from soil, water, rotten fruit and vegetables of which 16 were pectinase producer. The findings of the present research indicate that in addition to fruit and vegetables, rhizosphere soil of A. paniculata also harbors a good number of pectinase producing bacteria. This also justifies the selection of A. paniculata rhizosphere soil for isolating pectinase producing bacteria for the present study.

The pectinase producing isolates were classified based on [34] as very good pectinase producer (clear zone ≥ 15 mm), good pectinase producer (clear zone ≥ 10 mm), weak pectinase producer (clear zone ≥ 5 mm) and poor pectinase producer (clear zone ≤ 5 mm). Among the 65 isolates, 41 isolates were very good pectinase producer indicating that not only the pectin rich fruit peels but also the rhizosphere soil of A. paniculata harbors good number of pectinase producer (Table 1). Among the very good pectinase producers, the top three strains (JBST22, JBST27 and JBST36) were selected for further study.

The top three pectinase producing strains were inoculated in nutrient broth containing different concentrations of NaCl (1%, 2%, 4%, 6% and 8%) to test their ability to survive under different osmotic pressure. All the three strains

were able to grow at 8% NaCl concentration. Among the three isolates, JBST36 grew better at pH 8.0 and JBST22 grew better in a more alkaline condition (pH 9.0). All the three strains were able to survive well at 40˚C temperature.

The 16S rRNA gene is a segment of prokaryotic DNA which codes for an rRNA, and this rRNA in turn makes up part of the small subunit of a ribosome. The 16S

rRNA gene sequence is about 1.5 kb long and is composed of both variable and conserved regions. This gene has enough interspecific polymorphisms to be used as a molecular marker to identify bacteria. Universal primers are usually chosen as complementary to the conserved regions at the beginning of the gene and at end of the whole sequence and the sequence of the variable region in between is used for the comparative taxonomy [35] [36] . 16S rDNA sequencing has played an important role in the accurate identification of bacterial isolates and discovery of many novel bacteria [37] .

As expected, the PCR amplification of the 16S rRNA gene yielded a single amplicon of about 1.5 kb for each of the three selected strains used in this study (Figure 3). The obtained DNA sequences were edited and preliminary identification was carried out by comparing the sequences with those in the nucleotide database of National Center for Biotechnology Information (NCBI) using BLAST. The partial 16S rRNA gene sequences obtained for JBST22, JBST27 and JBST36 were 100% identical to the published sequence for Bacillus flexus strain ML-27 isolated from Populus euphriatica, Bacillus subtilis strain PARZ2 isolated from Prosopis julifora and Bacillus subtilis strain CS10 isolated from wild rice, respectively. Based on the morphological, biochemical and molecular data, JBST22, JBST27 and JBST36 were identified and designated as Bacillus flexus strain JBST22, Bacillus subtilis strain JBST27 and Bacillus subtilis strain JBST36, respectively. The sequence of these 3 strains were submitted to the GenBank of NCBI and the accession numbers KY407785 (Bacillus flexus JBST22), MK085080 (Bacillus subtilis JBST27) and MK085082 (Bacillus subtilis JBST36) were obtained. The study reveals that there are at least 66 base differences between the 16S rRNA gene sequences of B. flexus JBST22 and B. subtilis JBST36 (Table 2). Among these 66 differences, 65 were base substitution. However, no such difference was present in the 16S rRNA gene sequences of B. subtilis JBST27 and B. subtilis JBST36. To confirm the position of these 3 strains in phylogeny, a number of 16S rRNA gene sequences of some representative Bacillus species were selected from the GenBank database for the construction of phylogenetic tree. As shown in Figure 4, the phylogenetic tree indicated that the strain JBST22 and B. flexus share one cluster whereas the other two strains (JBST27 and JBST36) and

Bacillus subtilis share another cluster.