Diversity analysis of Populus euphratica endophytic bacteria in Tarim River Basin, China

Sampling

The sampling sites where the material was collected were the P. euphratica forests at the disused Achik River located in Shaya County (N40°56′, E82°15′) and at the disused Ugan River located in Yuli County (N41°00′, E85°04′), Xinjiang Province, China. These forests are characterized by high density, flourishing growth, and low degradation degree and are sparsely populated. They are rare natural P. euphratica forests not affected by human activities. A total of eight P. euphratica trees in the two sample sites were selected to obtain samples. At each sampling site, four individual P. euphratica trees (P. euphratica tree height ≥6 m and DBH ≥50 cm) were randomly selected. Tree height was measured using a tree height gauge (Vertex-IV, Haglof, Dalarna, Sweden) and a tape measure for diameter at breast height. A sterilized annular ring drill was used to drill holes in the sterilized P. euphratica trunk, the hole approximately 1.6 m above the ground. The P. euphratica stem storage liquid was collected in sterilized centrifuge tubes, which were stored at 4 °C for further analysis. The borehole on the P. euphratica trunk was filled with specially prepared P. euphratica stakes to prevent the P. euphratica tree from dying (Khayir et al., 2011).

Isolation and purification of culturable strains

Culturable strains were isolated using three types of culture media: LB, King B, and NA. The Luria-Bertani medium (LB) contained 1% peptone, 1% NaCl, 0.5% yeast extract, at a pH of 7.5–8.0. The King B medium (King B) contained 2% peptone, 0.03% K₂HPO₄, 0.15% MgSO₄·7H₂O, 1.5% (v/v) glycerin, at a pH of 7.5–8.0. Nutrient Agar medium (NA) containing 3% beef extract, 0.5% peptone, 0.5% NaCl, at a pH of 7.5–8.0.

The storage liquid collected from 8 P. euphratica trees was mixed in an equal amount, gradient diluted with a liquid medium. It was subsequently spread on the LB, King B, and NA solid medium surface, then cultured upside down for 3–7 days at 37 °C. Single colonies with significant differences in morphology were selected, repeating purification 2–3 times. The purified bacterial strains were collected in a liquid medium containing 40% (v/v) glycerin and stored at −80 °C.

Evaluation of culture medium predominant index and clone libraries

The culture medium predominant index was calculated by the formula: D = N/NT, where N was the number of endophytic bacteria species isolated from the medium, and NT was the total number of isolated strains (Lan et al., 2008).

The clone library was evaluated using two methods: (1) Coverage value. The formula C = 1−nl/N was used, where nl represents the number of OTUs that only appeared once in the clone library, and N represents the total number of clones in the library. The C value theoretically represents the proportion of microbial species in the clone library to all microbial species in the sample. (2) Rarefaction curves. Rarefaction curves were calculated using Estimates 8.0 software (http://viceroy.eeb.uconn.edu/estimates).

Total DNA extraction and bacterial 16S rRNA gene amplification

The genomic DNA of culturable bacterial strains was extracted using the sodium dodecyl sulfate-proteinase K-cetyltrimethylammonium bromide (CTAB) method (Allen et al., 2006). PCR amplifications were carried out targeting the 16S rRNA gene with the 27F (5′-AGAGTTTGATCACTGGCTCAG-3′) and 1492R (5′-TACGGYTACCTTGTTACGACTT-3′) primers (Zhang et al., 2021).

The PCR reaction mixture (50 µL) contained DNA template 100 ng, 1× Taq reaction buffer (R001A; TaKaRa, Beijing, China), 200 μmol dNTPs, 10 pmol of each P1 and P2 PCR primers, and 1.25 U Taq DNA polymerase (R001A; TaKaRa, Beijing, China). Thermal cycling conditions were as follows: an initial denaturation at 94 °C for 5 min, followed by 30 cycles of 94 °C for 30 s, annealing at 55 °C (27F and 1492R)/53 °C (799F and 1492R) for 30 s, 72 °C for 1 min, with a final extension at 72 °C for 8 min. PCR products were visualized by 1 % agarose gel electrophoresis and purified using the TIANgel Midi Purification Kit (DP209; Tiangen, Beijing, China) as described by the manufacturer. The purified PCR amplicons of culturable strains were sent to Sangon Biotech (Shanghai) Co., Ltd. for sequencing.

The purified PCR products from the total DNA extracted from the P. euphratica storage liquid were inserted into the pMD18-T vector by overnight incubation in a water bath at 16 °C. They were subsequently transformed into DH5α competent cells (CB101; Tiangen, Beijing, China). Positive clones were screened for standard blue and white screening. Colonies randomly picked were screened directly for inserts by performing colony PCR. All positive clones were collected in a Luria-Bertani liquid medium containing 40 % (v/v) glycerin and stored at −80 °C.

Culture-independent PCR-RFLP bacterial community analysis

For the culture-independent approach, the total DNA was extracted as previously reported with some modifications (Krsek & Wellington, 1999). The primer pair 799F (5′-AACAGGATTAGATACCCTG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) were used to amplify the DNA of P. euphratica endophytic bacteria (Sun et al., 2008). These primers do not amplify P. euphratica chloroplast DNA. PCR amplification with 799F and 1492R results in an amplicon of approximately 735 bp.

PCR-RFLP was performed to analyze the diversity of positive clones. The PCR reaction mixture (25 µL) contained 2 µL E. coli carrying the cloned insert DNA, 1× Taq reaction buffer (R001A; TaKaRa, Beijing, China), 200 μmol dNTP, 10 pmol of each M13–47 and M13–48 PCR primers, and 0.75 U Taq DNA polymerase (R001A; TaKaRa, Beijing, China). Thermal cycling conditions were as follows: an initial denaturation at 94 °C for 5 min, followed by 35 cycles of 94 °C for 30 s, annealing at 56 °C for 30 s, 72 °C for 1 min, with a final extension at 72 °C for 8 min.

The PCR product digestion was performed by modifying the Miao et al. (2013) method: 20 µL of the PCR products was digested using the restriction enzyme Hae III for 12 h at 37 °C, the restriction fragments were separated on a 2% agarose gel running in 1× TAE buffer at 100 V for approximately 40 min. Clones with different enzyme digestion patterns were selected and grouped into OTUs, as described by Aravena et al. (2020). Subsequently, the selected OTUs were sent to Sangon Biotech (Shanghai) Co., Ltd. for 16S rRNA gene sequencing.

Phylogenetic analysis and bacterial distribution analysis

The sequences obtained were compared and analyzed using EZTAXON (EzTaxon server 2.1) and BLAST (http://www.ncbi.nlm.nib.gov/blast/blast.cgi). The related species 16S rDNA sequences were retrieved from the NCBI nucleotide database. MEGA 7.0 software was used for the construction of phylogenetic trees. Bootstrap analysis was performed on 1,000 random samples taken from the multiple sequence alignment analysis and was carried out using Clustal W. Phylogenetic analysis was performed by the Neighbor-Joining method (Kumar, Stecher & Tamura, 2016).

The bacterial community structure was analyzed using the Venny 2.1 software (https://bioinfogp.cnb.csic.es/tools/venny/index.html).

Isolation characteristics of endophytic bacteria in the culture medium

The P. euphratica stem storage liquid endophytic bacteria isolated by the LB, NA and King B culture medium were different (Table 1). A total of 66 strains of bacteria were isolated by the culture-dependent approach. By observing colony characteristics and cell morphology, it was found that the LB medium had the largest number of colonies and rich colony morphology. In contrast, King B medium had a smaller colony number and a single colony morphology.

The culture medium significantly affected the number and type of bacterial strains isolated. A total of 29 strains of endophytic bacteria were isolated from LB medium, belonging to 10 genera. A total of 24 strains were isolated from the NA medium, belonging to six genera, and 13 strains were isolated from the King B medium, belonging to five genera. There were no common strains in the three medium, among which NA medium and LB medium had three common strains: Pseudomonas xinjiangensis, Brenneria salicis and Bacillus safensis. However, NA medium and King B medium had only one common strain, Erwinia toletana. LB medium and King B medium had no common strains.

Moreover, according to the dominance index in each of the three media, the dominance index of LB medium was 0.567, and that of NA medium and King B medium was 0.333 and 0.233, respectively. The above results indicated that the endophytic bacteria diversity isolated from the LB medium was the highest, while the King B medium had the lowest diversity.

Identification of endophytic bacteria in Populus euphratica based on 16S rDNA sequences

In this study, a total of 66 endophytic bacteria strains were obtained by the culture-dependent approach. Their 16S rDNA amplified fragments were sequenced, and the sequencing results were submitted to GenBank. The accession numbers of the isolates 16S rDNA sequences were JQ353770–JQ353835. The similarity comparison results are shown in Table 2. The sequencing data indicated that the 66 strains of culturable endophytic bacteria belonged to three major phylogenetic groups, Firmicutes, Actinobacteria, and Gamma-Proteobacteria, with 30 species belonging to 16 genera.

Among the 66 endophytic bacteria, 38 strains, or 57.58 % of the total, belonged to Gamma-Proteobacteria, thus being the most dominant group. These strains belonged to eight known genera. Pseudomonas was the dominant strain, with 18 strains accounting for 27.27 % of the total isolated strains. A total of 14 strains had a high similarity of sequences, ranging from 99.42% to 100%, with Pseudomonas xinjiangensis model strains. It indicated that these 14 strains were different mutants of the same strain, which showed these strains have more multi-directional dissimilation tendency and high detection frequency. They were the dominant species of culturable endophytic bacteria in this study. Interestingly, the taxonomic unit of Pseudomonas luteola represented by CBN-1 (JQ353770) was not previously identified in P. euphratica stem storage liquid samples (Khayir et al., 2011; Hormathan et al., 2014). The sequence similarity between CGM-17 (JQ353817) and the model strain Pantoea rwandensis (JF295055) was only 96.01%, indicating that CGM-17 was a potential new species.

There were 24 strains belonging to Firmicutes, the second most dominant group accounting for 36.36% of the total isolated strains. These strains belonged to five known genera. Bacillus was the second most dominant genus, accounting for 19.69 % of the isolated strains. A total of 13 Bacillus strains were isolated, belonging to 5 taxonomic units. The taxonomic unit of Bacillus licheniformis represented by CG-9 (JQ353819) was firstly found in the stem storage liquid samples of P. euphratica in our study, and its sequence similarity with model strain Bacillus licheniformis (JF798392.1) was only 96.29 %. In addition, the sequence similarity between CG-11 (JQ353820) and model strain Bacillus anthracis (MF101005.1) was only 96.19 %, speculating that CG-9 and CG-11 were potential new species.

Four endophytic bacteria strains belonging to four known genera of Actinobacteria, Nesterenkonia, Agrococcus, and Cellulomonas, were identified, accounting for only 6.06% of the total number of isolates. In addition, 14 other genera of endophytic bacteria were isolated, which fully reflected the community diversity of P. euphratica stem storage liquid samples. By comparing the previous study’s P. euphratica stem storage liquid samples (Khayir et al., 2011; Hormathan et al., 2014), it was found that among the 14 genera, Erwinia, Raoultella, Agrococcus, and Cellulomonas were isolated from stem storage liquid for the first time. Similarly, the species Pantoea wallisii, Pantoea rwandensis, and Oceanobacillus manasiensis were for the first time isolated from P. euphratica located in Tarim River Basin.

The bacterial clone library was obtained by the culture-independent method. A total of 48 OTUs (400 clones in total) were obtained after RFLP sequencing (Fig. 1). The sequence accession numbers were JQ941396–JQ941443, and the similarity comparison results are shown in Table 3.

Phylogenetic analysis and RDP classification indicated that the 48 OTUs isolated from the clone library belonged to Firmicutes, Actinobacteria, Alpha-Proteobacteria, Beta-Proteobacteria, Gamma-Proteobacteria, Bacteroidetes, and Verrucomicrobia. Proteobacteria was the dominant group, accounting for 59.75% of the total OTUs. Among the three subgroups, Alpha-Proteobacteria accounted for 6.5%, Beta-Proteobacteria accounted for 0.5%, and Gamma-Proteobacteria accounted for 52.75% of the total OTUs. Firmicutes, the second most dominant group, accounted for 28.25% of the total OTUs. The third most dominant group was Bacteroidetes, accounting for 10.25% of the total OTUs. The other phyla accounted for a smaller percent abundance (Table 3).

16S rDNA sequence phylogenetic analysis of endophytic bacteria in Populus euphratica

A phylogenetic tree fusing the partial 16S rDNA sequences was constructed using the neighbor-joining method (Fig. 2). Five phylogenetic groups, Firmicutes, Actinobacteria, Proteobacteria, Bacteroidetes, and Verrucomicrobia, covered all taxonomic units of the bacteria identified in P. euphratica stem storage liquid in the Tarim River Basin.

Among them, 66 endophytic bacteria obtained from the culture-dependent approach were clustered into Firmicutes, Actinobacteria, and Proteobacteria. Gamma-Proteobacteria was the most dominant group of culturable endophytic bacteria clustered in a large branch of the phylogenetic tree. Firmicutes were the second most dominant group, and the 24 isolates were clustered on a large branch. Bacillus (19.69%) was the second most dominant genus, in agreement with the reported endophytic bacteria literature (Khayir et al., 2011; Ju, Ouyang & Zhang, 2014). Only four endophytic bacteria strains belonged to Actinobacteria, and their evolutionary branch was divided further into three branches. CGL-11 (JQ353794) and clone xj-272 (JQ941435) were clustered in the same clade, while the sequence similarity between CGL-11 (JQ353794) and Nesterenkonia aethiopica (AY574575) was 98.10%.

Proteobacteria was the most dominant group in the bacterial clone library obtained by the culture-independent method and included 11 OTUs. Its evolutionary branches were divided into three clades. Alpha-Proteobacteria and Beta-Proteobacteria were unique to the culture-independent method, clustering on separate clades. The clones and isolates of Gamma-Proteobacteria, the dominant group in the phylum, were clustered on a large branch. The sequence similarity between the dominant clone xj-14 (JQ941404), accounting for 20.25% of the total clone library of the culture-independent method, and model strain Brenneria salicis (HM441252.1) was 98.47%. The clones belonging to the genus Erwinia and Pseudomonas followed in abundance, accounting for 19.5% and 7% of the total bacterial clone library, respectively.

Firmicutes were the second most dominant group in the bacterial clone library, accounting for 22 OTUs. The sequence similarity between the dominant clone xj-40 (JQ941411), which accounted for 5.75 % of the total bacterial clone library, and the model strain Alloiococcus otitis (AY957475.1) was 96.42%. Clone xj-323 (JQ941440) alone occupied a taxon alone and had 96.71% sequence similarity with Candidatus Syntropholuna (KU681304.1). On the other hand, the sequence similarity between clone xj-21 (JQ941408) and model strain Thermacetogenium phaeum (NR_074723.1) was only 90.99%.

Bacteroidetes and Verrucomicrobia were only found in the culture-independent method. Bacteroidetes was the third most dominant group, with 11 OTUs, accounting for 10.25% of the total bacterial clone library. Its evolutionary branch was further divided into three branches. The clone xj-213 (JQ941429), which formed its own branch, had a 96.77% sequence similarity with Litoribacter populi (MN209790.1).

Actinobacteria was the fourth most dominant group in the bacterial clone library, with three OTUs, accounting for 1.5% of the total bacterial clone library. The clone xj-107 (JQ941419) and the culturable strains CGM-7 (JQ353811) and CGM-16 (JQ353816) were clustered in the same clade, classified as Cellulomonas. Verrucomicrobia contained only one OTU. Clone xj-398 (JQ941443) occupied a taxon alone and had 100% sequence similarity with an Uncultured Verrucomicrobia bacterium (HQ857650.1).

Analysis of bacterial species distribution in the stem storage liquid of P. euphratica

According to sequencing results, 53 bacterial species were identified in P. euphratica stem storage liquid, of which 30 species were isolated by the culture-dependent method and 25 species were identified by the culture-independent method (Fig. 3). The two methods exhibited obvious differences in the distribution of bacterial community structure. Only Brenneria salicis and Pseudomonas xinjiangensis were isolated by both methods, while most of the bacterial species were uniquely identified in the respective methods. Among the 30 species of bacteria obtained by the culture-dependent, 28 species (52.8% of the total identified) were unique. Thus, among the 25 species of bacterial OTUs identified by the culture-independent method, 23 species (43.4% of the total identified) were unique. These results indicate that the different isolation and identification approaches could be combined and compensate each other for the accurate bacterial community structure determination in the stem storage liquid samples of P. euphratica.