Diversity and function of culturable actinobacteria in the root-associated of Salvia miltiorrhiza Bunge

Sample collection

Samples used in this study were collected from Sichuan Province (E: 104.54°, N: 30.94°), China, on March 20th, 2019. The samples were composed of roots and rhizosphere soil. During sampling, the roots of healthy plants were excavated, the clumpy soil on the roots’ surfaces was discarded. Soil particles closely adhering to the surface of the roots were collected as rhizosphere soil. The roots and rhizosphere soil were placed in sterile paper bags and brought into the laboratory at 4 °C. Endophytic and rhizosphere actinobacteria were isolated within 48 h.

Surface sterilization and isolation actinobacteria

The root surfaces sterilization protocol included three steps: the tissue surface was immersed in 1% (v/v) sodium hypochlorite for 3 min; soaked in 75% (v/v) ethanol for 30 s and finally rinsed five times with sterile distilled water. The surface sterilization effectiveness was evaluated by plating the final wash water on Gause’s No. l medium. The sterilized surface samples were cut into small fragments and homogenates. A little sterile distilled water was added and mixed, then 0.2 mL suspension was dispersed onto six actinobacteria-selective mediums: (A) Gause’s No. l medium (Zhang & Zhang, 2011); (B) Gauze’s agar medium No.2, modified (Yan et al., 2016); (C) HVA (Qin et al., 2009); (D) Trehalose-proline agar (Qin et al., 2009); (E) TWYE medium (Qin et al., 2009); and (F) S medium (Cao et al., 2004).

For rhizosphere soil samples, 10 g of soil was added to 90 mL of sterile water and shaken at 100 × g at 4 °C for 3 min. A 10⁻² soil diluent was prepared using the 10-fold serial dilution method. Then, the homogenization buffer or diluent was placed in the media (A), (C), and (G). All media were supplemented with potassium dichromate (K₂Cr₂O₇) (75 mg L⁻¹) and penicillin sodium (2 mg L⁻¹) and incubated at 28 °C for four weeks before being monitored every seven days for microbial growth. Colonies were observed and selected according to their characteristics and morphology. The purified actinobacterial isolates were stored on the Gause’s No. l medium at 4 °C and in 30% glycerol at −80 °C.

Isolates Identification

For each isolate, genomic DNA was extracted using the HiPure Bacterial DNA Kit per to the manufacturer’s instructions. The 16S rRNA gene was amplified using universal primers 27F (5′-CAGAGTTTGATCCTGGCT-3′) and 1492R (5′-AGGAGGTGATCCAGCCGCA-3′). The PCR mixture (25 µL) contained 22 µL 1 ×Taq PCR Master Mix, 1 µL of the DNA template, and 1 µL of each primer. We used the following amplification protocol: PCR was initialized by a denaturing step at 94 °C for 4 min followed by 34 amplification cycles with denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, extension 72 °C for 1min, and a final polymerization step at 72 °C for 10 min. PCR products were stored at 4 °C and sequenced by Beijing TSINGKE Biological Technology. The obtained forward and reverse DNA sequence reads were used to produce contig sequence using DNASTAR Lasergene v.7.1 software (Burland, 2000). The 16S rRNA gene sequences were compared with those of NCBI. A maximum likelihood phylogenetic tree of 16S rRNA gene sequences was constructed using the molecular evolutionary genetics analysis (MEGA) software, v.7.0 (Kumar, Stecher & Tamura, 2016).

Analysis of the actinobacteria root-associated function

For function analysis, metagenomics was performed on the DNA mixture of the root-associated actinobacteria. DNA was mixed at 50 ng/L per strain, and three replicates were set. The purity and integrity of DNA were analysed by agarose gel electrophoresis (AGE); then, the library was constructed. After library quantification, Illumina PE150 sequencing was performed. Preprocessing of the raw data obtained from the Illumina HiSeq sequencing platform using Readfq (v.8, https://github.com/cjfields/readfq) was conducted to acquire clean data for subsequent analysis. The clean data were assembled and analysed using SOAPdenovo software (Qin et al., 2010; Karlsson et al., 2012; Karlsson et al., 2013) (v.2.04, https://sourceforge.net/projects/soapdenovo2/). The scaftigs predicted the open reading frame (ORF) using the MetaGeneMark (v.2.10, http://topaz.gatech.edu/GeneMark/) software (Zhu, Lomsadze & Borodovsky, 2010; Karlsson et al., 2012; Karlsson et al., 2013; Mende et al., 2012; Nielsen et al., 2014; Oh et al., 2014). For ORF prediction, the CD-HIT software (Li & Godzik, 2006; Fu et al., 2012) (v.4.5.8, http://www.bioinformatics.org/cd-hit) was adopted for redundancy and to obtain a unique initial gene catalogue. The DIAMOND software (Buchfink, Xie & Huson, 2015) (v0.9.9.110, https://github.com/bbuchfink/diamond/) was used to blast Unigenes to the functional database, including the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (Kanehisa et al., 2006; Kanehisa et al., 2014) (v.2018-01-01, http://www.kegg.jp/kegg/), Evolutionary Genealogy of Genes: Non-supervised Orthologous Groups (eggNOG) database (Powell et al., 2013) (v.4.5, http://eggnogdb.embl.de/#/app/home), and Carbohydrate-Active enZYmes (CAZy) database (Cantarel et al., 2009) (v.201801, http://www.cazy.org/) for comparison, function annotation and abundance analysis.

Screening for antimicrobial activity

Antimicrobial assays of 72 actinobacterial strains against the bacterial and fungal strains were performed using the agar diffusion method (Nimaichand et al., 2015). Six microorganisms, including two Gram-positive bacteria: Staphylococcus aureus, Bacillus cereus; two Gram-negative bacteria: Escherichia coli and Pseudomonas aeruginosa; and a yeast, Candida albicans, the S. miltiorrhiza pathogen, Fusarium oxysporum, served as the test microorganisms. The test bacteria and fungi were activated on LB broth and PDB for 12 h and 2 d for 30 °C, respectively. They were adjusted to approximately 10⁸ and 10⁶ CFU/mL with sterile distilled water to prepare a microbial and spore suspension. A one mL bacterial/fungal solution was then added to 100 mL LB/PDA media, mixed gently, and then slowly poured on the Petri dish (90 mm), which served as the test plate. The actinobacterial disks (nine mm) were cultured on Gause’s No. l medium for seven days at 28 °C. They were transferred to media seeded with the test bacteria and fungi in triplicate and were incubated at 37 °C for 24 h and 28 °C for 2 d, respectively. The inhibition zone (mm) was observed and documented. All experiments were replicated three times.

Screening for plant growth-promoting activity

Seventy-two strains were tested for plant growth-promoting activities in vitro. Siderophore production was studied by the method described by Alexander & Zuberer (1991). Briefly, using chrome azurol agar (CAS), colonies that exhibited a yellow halo after proliferation were classified as positive for siderophore production. The strains’ phosphate solubilization capability was analyzed on Pikovskaya (PVK) medium (Gupta et al., 1994). Finally, the phosphate solubilization capacity was determined by a halo surrounding colonies on Pikovskaya medium. IAA production was checked using Salkowski reagent in a tryptophan-amended medium (Gordon & Weber, 1951). The development of pink coloration indicates IAA production. Optical density was taken at 530 nm, and compared to a standard curve of indole-3-acetic acid of 5 µg/mL to 100 µg/mL at 530 nm. Nitrogen fixation activity was tested using nitrogen-free Ashby media (Li et al., 2008). Strain growth proved to have the ability to fix nitrogen. All strains were incubated for seven days and replicated three times.

Isolation of the root-associated actinobacteria

We isolated 201 endophytic actinobacteria and 168 rhizosphere soil actinobacteria from the root of S. miltiorrhiza and its rhizosphere soil. These organisms’ identities were determined based on 16S rRNA gene sequence analysis. The sequences of 100% identity were clustered into one species. According to the analysis of sequencing results, many isolated strains had the same 16S rRNA sequence. We obtained 72 actinobacterial strains after merging the duplicated sequences (GenBank accession numbers MW642091 to MW642162). Based on the BLAST result and phylogenetic tree analysis, actinobacteria obtained in this study were identified and categorized into three families (Streptomycetaceae, Tsukamurellaceae, and Nocardiopsaceae), four genera (Streptomyces, Nocardiopsis, Tsukamurella, and Kitasatospora), and 46 species, among which Streptomyces was the dominant genus. In the 46 species of actinobacterial, 14 were unique to rhizosphere soil, and 15 were unique to the root compartment. Most of the strains exhibited more than 98.65% similarity to other published type species. Table S1 presents the percentages of 16S rRNA gene sequence similarities (98.22%–100%) of these isolates to the closest type strains. Figure 1 shows the phylogenetic tree.

Function analysis of the root-associated actinobacteria

We explored the root-associated actinobacterial functionalities using Illumina metagenome sequencing, focusing on the functional genes, enzymes, and pathways involved in PGP activities. The metagenomic contigs were annotated in eggNOG, CAZy, and KEGG databases. We found that the root-associated actinobacteria had potent functions.

First, in our study, annotation using KEGG database showed that the genes involved in Metabolic was the most predominant. There were 71 maps whose relative abundance was more than 0.1% in Metabolism, among which Carbohydrate metabolism accounts for the largest proportion, followed by Amino acid metabolism (Table S2). Second, through the analysis of CAZy reference sequences, we observed that Glycoside Hydrolases and Glycoside Hydrolases were the dominant CAZy families, followed by Glycoside Hydrolases, Carbohydrate Esterases, Polysaccharide Lyases, and Auxiliary Activities (Fig. 2, Table S3). Third, the functional annotations by eggNOG process categories indicated Transcription [K]was the dominant function among 23 categories, followed by Carbohydrate transport and Metabolism [G] (>25000), Amino acid transport and metabolism [E], Signal transduction mechanisms[T], and Energy production and conversion[C] (>17000). The lowest number of genes is assigned to Cell motility[N], Chromatin structure and dynamics[B], RNA processing and modification[A], Extracellular structures[W], and Cytoskeleton[Z] (Fig. 3, Table S4).

Besides, the genes and enzymes involved in plant growth-promoting traits (PGPTs) were also annotated (Figs. 4 and 5), including, enzymes related to IAA synthesis (EC1.13.12.3, EC3.5.1.4, EC1.4.3.4, EC1.2.1.3), genes involved in nitrogen metabolism (nifU, narG, narI, nirB, nirD, nirK, nxrB, norB, glnA, GLUL, gltB, gltD, nxrA, and gdhA), phosphorus metabolism (phoD, phnG, phnI, phnJ and phnA), and siderophore biosynthesis (pchB, dhbF, and entF). We have discovered the presence of chitinase and β-1,3-glucanase that help fight plant pathogens.

On the other hand, some antibiotic synthesis pathways have been annotated, including map00521(Streptomycin biosynthesis), map00524(Neomycin, kanamycin, and gentamicin biosynthesis). And some possible biosynthetic gene clusters involved in secondary metabolite production, such as NRPS, NRPS-like, T1PKS, and hgIE-KS, were also annotated using the AntiSMASH tool (Kai et al., 2019) (https://antismash.secondarymetabolites.org). The complete citric acid cycle was annotated in the root-associated actinobacteria, which can help microorganisms use insoluble inorganic phosphate salts (Fig. S1).

Antimicrobial activity

Here, seventeen isolates exhibited antimicrobial activity against indicator microorganisms. Most of the active isolates showed antimicrobial activity against Staphylococcus aureus and Fusarium oxysporum. Among these tested isolates, thirteen strains were shown antimicrobial activity against Staphylococcus aureus, and eight strains were showed antimicrobial activity against Fusarium oxysporum. There were three strains and one strain that had inhibitory effects on Candida albicans and Escherichia coli, respectively. The EA083 strain (the similarity with Streptomyces lavendulae NRRL B-1230 is 99.71%) showed a high and broad antimicrobial spectrum (Fig. 6).

Plant growth-promoting activity

To further characterize the root-associated actinobacterial isolates, their PGPTs were tested. These isolates showed different PGP activities (Table S5). Of 72 strains, 24 isolates had phosphate solubilizing properties, 52 isolates grew on an N-free medium. 37 isolates (51.3%) produced IAA, with an average content of 2.48–21.62 µg/mL, and SA047 had the highest IAA content. Of the total 72 isolates, 25 isolates showed siderophore production. 7 isolates showed all tested PGPTs (Fig. 7).

The species diversity of culturable root-associated actinobacteria of S. miltiorrhiza

Three hundred and sixty-nine root-associated actinobacterial of S. miltiorrhiza were isolated in this study, belonging to forty-six species in four genera, among which Streptomyces was the dominant genus. Streptomyces was widely present in various habitats. Many studies have shown that, in various plants, it is the dominant genus of endophytic and rhizosphere actinobacteria (Kim et al., 2012; Kaewkla & Franco, 2013; Dinesh et al., 2017; Bulgarelli et al., 2012). Li et al. (2010) and Yan et al. (2016) isolated Streptomyces from S. miltiorrhiza, and the same genus was obtained in this study. Moreover, different from previous studies, the genera Tsukamurella and Nocardiopsis were isolated. To our knowledge, this is the first study that reported Tsukamurella and Nocardiopsis isolation from the tissues of S. miltiorrhiza. Besides, the isolates of EA050, EA140, and SA079 shared low 16S rRNA gene sequence similarities (<98.65%, the threshold for differentiating two species) with validly described species based on the results of BLAST search in NCBI, indicating that these isolates might represent novel Streptomyces species (Kim et al., 2014). In this study, Streptomyces turgidiscabies and Streptomyces acidiscabies were isolated. They reportedly cause potato scabies (Santos-Cervantes et al., 2016). However, they did not cause tissue lesions S. miltiorrhiza, and the reason remains are unclear. We speculated that it might be related to the root-associated microbial system balance.

The functional diversity of culturable root-associated actinobacteria of S. miltiorrhiza

Metagenomics help explore and discover the genes and functions of microorganisms in natural environments that are unreachable through classical microbial cultivation techniques. Although culture-independent approaches clarify potential microbial roles, culture-based studies are still important for practical microbial use. Some scholars have used metagenomics to conduct functional analysis of culturable microorganisms (Costa et al., 2020; Diaz-Garcia, Bugg & Jimenez, 2020). This implies we can combine the strengths of both metagenomics and cultivation techniques to understand the functions of culturable microorganisms. This study found that the root-associated actinobacteria of S. miltiorrhiza have rich functional genes, and have the potential to promote plant growth, protect plant against pathogens, and synthesize a variety of biologically active ingredients.

Plant growth promotion

Actinobacteria can promote plant growth by making nutrients/substrates (e.g., phosphorous, nitrogen, and iron) accessible to the host plant and producing various plant hormones (Liu et al., 2017; Purushotham et al., 2018; Borah & Thakur, 2020). IAA is an important plant hormone beneficial to plant growth. Many studies have reported that Streptomyces could produce IAA, promoting plant growth and increasing seed germination rate (Dochhil, Dkhar & Barman, 2013; Lin & Xu, 2013; Myo et al., 2019). Nitrogen and phosphorus are indispensable nutrients for plant growth and development. There is evidence that the root-associated actinobacteria can increase plant nutrient utilization by fixing nitrogen and dissolving phosphorus, consequently promoting plant growth (Anita & Richau, 2013; Solans et al., 2019). Iron is an essential trace mineral for almost all organisms, and it influences various activities and cell metabolisms, such as respiration and photosynthesis. However, iron availability is often limited due to its poor solubility (Cao et al., 2020). Studies have found that some microorganisms can synthesize siderophores to help plants acquire iron in the soil (Crowley et al., 1991). Besides, they can also compete with pathogens for iron to produce antagonism, which is also an important mechanism for antagonizing pathogens (Kamil et al., 2018). As far as we know, among the 46 species of actinobacteria in this study, many species reportedly promote plant growth. For example, Nocardiopsis dassonoillei and Streptomyces violaceoruber can promote wheat and pepper seedling growth, respectively (Sang et al., 2013; Allali et al., 2019).

In our study, the enzymes related to IAA synthesis (EC1.13.12.3, EC3.5.1.4, EC1.4.3.4, EC1.2.1.3) and genes related to nitrogen fixation (nifU), nitrification (nxrB), denitrification (nirK, norB) were annotated. Moreover, genes coding alkaline phosphatase(phoD), C-P lyase (phnG, phnI, and phnJ), C-P hydrolase (phnA), and siderophores biosynthesis (pchB, dhbF, and entF), were also found. In particular, we found that the above-mentioned related genes and enzymes are abundant in the top ten abundance species (Fig. 5). Besides, in vitro experiments proved the correctness of function prediction, and 37 strains produced IAA in the presence of tryptophan, 52 isolates had nitrogen-fixing characteristics, and 24 strains had phosphorus-dissolving characteristics, and 25 strains had siderophores-producing characteristics. In summary, it is indicated that the root-associated actinobacteria of S. miltiorrhiza might participate in IAA biosynthesis, siderophores production, and nitrogen and phosphorus metabolism in plants and soil by producing associated enzymes.

Antagonists pathogens

Disease is the main biotic stress factor leading to crop productivity reduction, and S. miltiorrhiza is no exception. The incidence of root rot caused by Fusarium oxysporum on S. miltiorrhiza roots is high in China. It has been earlier reported that the root-associated actinobacteria could effectively inhibit plant pathogens, thus protecting the host plants from diseases (Al-Askar et al., 2015; Muangham, Pathom-aree & Duangmal, 2015; Soltanzadeh, Nejad & Bonjar, 2016). Several antibiotic synthesis pathways were annotated in this study, including map00521(Streptomycin biosynthesis), map00524(Neomycin, kanamycin, and gentamicin biosynthesis). Moreover, phenazine (map00405) and chitinase (EC3.2.1.14) biosynthetic genes/pathways have been noted. Antibiotics, phenazines, and chitinase reportedly have an antagonistic effect on plant pathogens (Singh & Gaur, 2016; Aamir et al., 2020). Actinobacteria, with the activity of synthesizing antibiotics and chitinase, are used as biological control agents against plant pathogens (Al-Askar et al., 2015; Abbasi et al., 2019). Furthermore, our study also revealed the presence of gab D and gab T known to be significant genes during γ-aminobutyric acid (GABA) production, which helps suppress plant pests and diseases (Belbahri et al., 2017). Simultaneously, of the 46 species isolated, many have been reported to exhibit antibiotic synthesis, chitin degradation, and plant pathogen antagonism, such as Tsukamurella tyrosinosolvens could produce antagonistic effects against Phytophthora Capsici (Sang et al., 2013).

Therefore, we speculate that the root-associated actinobacteria of S. miltiorrhiza exhibited plant-pathogens antagonism. Fortunately, antimicrobial experiments confirm our conjecture. Eight strains inhibited Fusarium oxysporum; among them, the EA083, EA166, and SA107 strains showed noticeable inhibitory effects (Fig. 6). This result shows that the root-associated actinobacteria might play a key role in Fusarium oxysporum antagonism by producing antimicrobial compounds and chitinase. In summary, the root-associated actinobacteria of S. miltiorrhiza have great potential as a biological control agent. Also, we found that the root-associated actinobacteria of S. miltiorrhiza exhibited inhibitory effects on human pathogenic bacteria, suggesting that it can be used in the agricultural and pharmaceutical industries.

Biosynthetic potential of isolates

Actinobacteria have always been one of the primary sources of bioactive ingredient production. As expected, we found that the culturable root-associated actinobacteria of S. miltiorrhiza have the potential to synthesize bioactive ingredients. Studies have reported that active metabolite synthesis was evaluated using specific primer pairs that target PKS-I, II, and NRPS (Katz & Baltz, 2016; Peng et al., 2020). In our study, we identified possible biosynthetic gene clusters involved in secondary metabolite production. We also found that related pathways (map01052 and map01056) were annotated. Besides, our study also noted the biosynthetic pathways of flavones (map00944, map00941, and map000943), and alkaloids (map00960, map00950). It is indicated that the root-associated actinobacteria S. miltiorrhiza. have the potential to synthesize different bioactive ingredients.

Surprisingly, we also found that it has the activity of degrading aromatic compounds, including polycyclic aromatic hydrocarbons, toluene, xylene, which are remarkably harmful organic pollutants in the soil (Peng et al., 2020). Compared to physical and chemical technology, the plant-related microbe can excellently remediate aromatic compound pollutants from contaminated sites due to its environmentally friendly advantages. In this study, the abundance of ko involved in degrading aromatic compounds was annotated. The average relative abundance of ko00362 (Benzoate degradation), ko00627 (Aminobenzoate degradation), and ko00625 (Chloroalkane and chloroalkene degradation) were greater than 0.1%. It is shown that the culturable root-associated actinobacteria of S. miltiorrhiza, as soil amendments, can degrade aromatic compounds.