Study on rhizosphere microbial communities of medicinal plant curcuma longa l. to enhance turmeric yield and quality

iotechnologically potential microorganisms and eventually to a sustainable production of the novel bioactive metabolites. CHAPTER 2. MATERIALS AND METHODS 2.1. Materials 2.1.1. Turmeric cultivar and crop This study was conducted at a turmeric growing area located in Dai-Tap commune, Khoai-Chau district of Hung-Yen province 5 (20°47′35″ N, 105°56′42″ E) using local seed rhizome of Curcuma longa. 2.1.2. Chemicals, oligonucleotides, media and microorganisms 2.2. Methods 2.2.1. Soil sampling and determination of soil physiochemical parameters Sampling: Soil samples were taken randomly from turmeric rhizome soil (10-15 cm depth from the top soil). Samples of each plot were then bulked, homogenized and grouped together to one sample set, followed by storing at 4 o C prior to DNA extraction. Determination of soil physiochemical parameters 2.2.2. Study on impacts of chemical N fertilizing rates to turmeric productivity Experimental design Field study was conducted from April to December 2016 and replicated in 2017. The experiment was laid out in a randomized complete block design and three replicates. Experimental units consist of 10 m 2 plots each with one fertilizer regime, resulting in a total of 16 plots in a total area of 160 m 2 . The treatments were four N fertilizer rates (0, 150, 350 and 500 kgN.ha -1 .y -1 ) incorporated with K and P fertilizers (400:200 kg.ha -1 .y -1 ), resulting in 5 fertilizer regimes: N0, N150, N350 and N500, respectively. Soil samples were taken randomly from turmeric rhizome soil (10-15 cm depth from the top soil) at five points of each plot following a W-pattern (Thomas 1985) [115]. Plant growth and productivity parameters: Plant height (cm); Number of leaves/plant; Fresh rhizome yield (kg/ha); Fresh rhizome yield/dry rhizome yield ratio (%); Curcuminoids content. 6 After harvesting, turmeric rhizomes were sliced and dried. Curcuminoids from turmeric samples were extracted by an ultrasound assisted extraction method using ethanol/water (70:30, v/v) solvent as described by Mandal et al. (2009) [116]. The quantification of curcuminoids content was performed using high performance liquid chromatography (HPLC) following the method of Jayaprakasha et al. (2006) [46]. 2.2.3. Study on impacts of N fertilizing rates to diversity of turmeric microbial community Total DNA extraction from turmeric rhizosphere soil samples Total DNAs were extracted by using PowerSoil ® DNA Isolation kit (Mo Bio Laboratories, Qiagen, USA) and quantified by Nano drop (Nanodrop 2000c, Thermo Fisher Scientific, USA) in combination with electrophoresis in gel agarose 1% and stored at -20 o C. Metagenome amplicons sequencing Sequencing libraries were prepared from the PCR products using TruSeq® DNA PCR-Free Sample Preparation Kit (Illumina, USA). The quality of libraries was assessed on Bioanalyzer 2100 system (Agilent, USA) before sequencing on Illumina HiSeq 2500 platform (Illumina, USA). Bioinformatic analysis The metagenome databases were analyzed following Qiime2 analyzing pipeline (https://qiime2.org/) [118]. The analysis process comprises 3 main steps, namely (i) Preprocessing; (ii) Taxonomy; and (iii) Diversity analysis and visualization.  Preprocessing: Using quality control tools of Qiime (V1.7.0, [146] and 7 UCHIME algorithm ( manual/uchime_algo.html) [148].  Taxonomy: Using Uparse software (Uparse v7.0.1001, [149], Unite database (https://unite.ut.ee/) [150] and Silva database (https://www.arb-silva.de/) [151].  Diversity analysis and visualization  Microbial diversity indices Determination of Shannon-Weaver (H), Simpson (D1), Chao1 and ACE indices was performed by Qiime 2.  PCA & PCoA Principle Component Analysis (PCA) and Principle Coordinate Analysis (PCoA) were conducted by applying R software (v 3.1.2, R Core Team 2014).  Rarefaction curve  Statistical Analysis: Using ANOVA followed by Tukey’s Honest Significant Difference (HSD) post hoc tests. 2.2.4. Isolation of PGPR and plant growth promoting assays Isolation of PGPR strains with phosphate solubilizing ability: Rhizosphere soil samples were diluted and spread on IPA agar plates for bacterial colonies formation [138]. IAA producing assay: IAA assay was conducted following Salkowski’s method [139]. Antagonism to test pathogenic microorganisms: Agar diffusion test as described by Ahmad & cs. (1998) [140]. Determination of biochemical and physiological characteristics: According to Bergey’s Manual of Systematic Bacteriology [141, 142]. 8 Phylogenetic identification using partial 16S rDNA gene sequences: The partial 16S rDNA gene sequences of bacterial isolates were amplified using PCR with primers Pr16F-Pr16R and compared to published sequence in GenBank using BLASTn tool, and analyzed by BioEdit 7.0 [144] and MEGA X [145] [146] softwares. 2.2.5. Isolation and characterization of AMF Isolation of AMF spores from turmeric rhizophere soil samples: AMF spores from rhizosphere soil samples were isolated using wet sieving and decanting method (Gerdemann, Nicolson, 1963) [147]. Partial 18S rRNA gene amplification and phylogenetic inference Fragments of partial small subunit (SSU) rRNA gene from extracted genomic DNA samples were amplified using universal eukaryotic forward primer NS31 and reverse primers mixture AM containing AM1, AM2 and AM3 [148, 149, 150] to amplify AM fungal SSU sequences. Clones from each sample were tested for the PCR amplicons and sequenced on an ABI PRISM® 3100 Avant Genetic Analyzer (Applied Biosystems, USA) sequencer. 2.2.6. Preparation of biopfertilizer for turmeric and case study on turmeric productivity Preparation of biofertilizer from isolated turmeric rhizosphere effective microbial strains  Safety test: Safety tests for microbial strains were conducted in BALB/c mice as described by Carter et al. [151].  AMF AMF spores were preserved and inoculated in pot cultures of Plantago lanceolata (supplied by Institute of Seed and Biotechnology - Vietnam Academy of Forest Science). 9  PGPR fermentation and biomass harvesting  AMF spores harvesting  Preparation of PGPR and AMF inocula: Biomass mixture of PGPR strains and spore mixture of AMFs were blended in 1:1 ratio (w/w). Case study on the effect of biofertilizer in turmeric plant  Experimental design Experiment was conducted at the Ministry of Health’s Botanical garden in Thanhtri, Hanoi from May to November 2018. The 12 m 2 units were designed in a total area of 30 m 2 belonging to 2 experiment groups: (i) Microbial inocula applied turmeric (symbolized as CP), and (ii) Control (symbolized as DC).  Determination of growth and productivity parameters The growth parameters are monitored periodically. Productivity was determined based on fresh biomass at the end of the experiment. III. RESULTS AND DISCUSSION 3.1. Investigation of relationship between nitrogen fertilizer regimes and turmeric productivity Environmental parameters of turmeric farming site Initial edaphic conditions prior to experimental plotting were determined by standard methods according to TCVN 7373:2004, 7374:2004 and 7375:2004. Results showed that the parameters corresponded sandy loam characteristics with relatively good soil quality. 10 Turmeric productivity and curcuminoids content in response to N fertilizer rates Turmeric productivity of different fertilizing regimes was determined in terms of fresh rhizome yield and curcuminoids content. The averaged fresh rhizome yield after harvesting was recorded in form of mean±SD as following: 21038±5013 kg.ha -1 .y -1 (N0), 27003±4703 kg.ha -1 .y -1 (N150), 30902±1642 kg.ha -1 .y -1 (N350); 20578±2306 kg.ha -1 .y -1 (N500). From these results, it is obvious that obtained turmeric yield in regimes N0 and N500 was lower than in N150 and N350. Most likely, either excessive or inadequate nitrogen inputs had negative effect on turmeric plant’s vigor and nutrient accumulation. Farming regimes with nitrogen fertilization levels ranging from 150 to 350 kg.ha -1 were considered optimal for fresh rhizome yield of turmeric. Figure 3. 2. Averaged fresh turmeric rhizome yield and curcuminoids content in experimental regimes. Curcuminoids content in response to N fertilizer rates The averaged curcuminoids content of each farming regime was calculated from available data after determination of dry rhizome weight and analysis of rhizome extracts. Results revealed 11 the highest curcuminoids yield in the experimental plots of N150 farming regime. Figure 3.2 summarizes averaged turmeric yield and curcuminoids content in relation to N fertilizer inputs. As illustrated in the figure, with N fertilizers ranging from 150 to 350 kg.ha -1 .y -1 , the turmeric yield and curcuminoid content were both recorded at higher levels than in either non-fertilized N0 or over- fertilized N500. Effects of N, P and K chemical fertilizers on turmeric yield and curcumin content have been intensively investigated in various geoecological areas of the global, but research results were inconsistent. The present results have confirmed the hypothesis of N fertilizer’s effect on yield and curcumin content in C. longa. On the other hand, the research has supported fundamental data for designation of appropriated fertilizing practices for turmeric plant in Vietnam. 3.2. Investigation of effective microbial groups in turmeric rhizosphere The investigation of effective microbial groups in turmeric rhizosphere was oriented by published research results and reviews concerning symbiotic microorganisms of C. longa. Accordingly, we focused on two main rhizosphere microbial groups, namely growth promoting rhizobacteria (PGPR) and arbuscular mycorrhizal fungi (AMF), thereby aiming to screening for a collection of microbial candidates for preparation of effective biofertilizer. 3.2.1. Investigation of plant growth promoting rhizobacteria (PGPR) in turmeric rhizosphere 3.2.1.1. Isolation and screening of PGPR The isolation of PGPR strains from turmeric rhizosphere was performed with following criteria: (i) Dissolve inorganic phosphate; 12 (ii) Production of indole aceteic acid (IAA); (iii) Nitrogen fixation ability and (iv) Antagonism to test pathogenic microorganisms. Screening results for PGPR strains from turmeric rhizosphere soil samples are shown in Table 3.6. Table 3. 6. Phosphate solubilizing ability, development in nitrogen- free medium and IAA production of PGPR strains. N o. Strain Phosphate solubilizing ability (D-d, mm) Development in nitrogen-free medium IAA production (ppm) 1 PGP-V2 3 + 24.80±2.21 2 PGP-V4 5 - 9.66±1.72 3 PGP-V5 3 + 67.51±2.11 4 PGP-V15 5 + 35.47±3.05 5 PGP-V18 3 + 26.82±1.46 6 PGP-V20 4 + 77.87±2.78 7 PGP-V21 6 + 63.11±2.09 8 PGP-V22 4 + 11.61±1.85 9 PGP-V24 4 - 45.81±2.89 * D: diameter of halos around bacterial colony; d: diameter of bacterial colony on agar plate. The indirect plant growth promoting effect of PGPR strains was evaluated basing on in vitro antagonism to plant pathogenic fungi Aspegillus niger and Fusarium oxysporum on agar plates. The result revealed inhibitory activity against both tested fungi A. niger and F. oxysporum of Bacillus sp. PGP-V21 with halo diameters of 9 and 3 mm, respectively. PGP-V5, PGP-V20 and PGP-V21 emerged as promising effective bacteria with typical PGP characteristics in terms of relatively potent P solubility, nitrogen fixation and IAA production. The bacterial strains were taxonomical characterized by determining morphological and physio-biochemical characteristics, in combination to partial 16S rRNA gene sequences analysis. 3.2.1.2. Morphological and taxonomical characteristics 13 Figure 3.3. Phylogenetic tree of PGP-V5, PGP-V20, PGP-V21 and published bacterial species basing on parial 16S rRNA gene sequences (Maximum likelihood method, 100 bootstrap replicates, consensus tree). Methylobacterium populi AP014809 was the out group. Morphological characteristics of bacterial colonies on agar plates containing LB medium and under microscope (magnification x1000) were observed. The morphological, physiological and biochemical features of the strains were compared with Bergey's classification key [141] [142]. Based on published sequences in GenBank and subsequent sequence analysis, a classification tree of strains PGP-V5, PGP-V20 and PGP-V21 has been constructed (Figure 3.3). As a result, the PGP bacterial strains were determined as Bacillus sp. PGP-V5, Enterobacter sp. PGP-V20 and Bacillus sp. PGP-V21. 14 In conclusion, four PGPR strains were isolated and biologically characterized from rhizosphere of turmeric plant in the course of the investigation. 3.2.2. Investigation of arbuscular mycorrhizal fungi (AMF) in turmeric rhizosphere 3.2.2.1. Isolation of AMF Besides PGPR, AMF has been intensively mentioned as effective symbionts of turmeric rhizosphere in earlier reports, especially those in Indian turmeric varieties [160, 161], however, there has not been any similar research hitherto on the indigenous turmeric sample of Vietnam. By wet decanting and filtrating [147], the presence of AMF spores in rhizosphere soil samples of turmeric in the study area at three different investigated periods (2, 5 and 8 months after planting) was determined. The results showed that investigated AMF spore numbers increased from 23.6±5.5 spores/100 g soil two months after planting to 66.5±7.5 spores/100 g at 8 months old turmeric. 3.2.2.2. Morphological and taxonomical characterization of AMF strains isolated from turmeric rhizosphere Depending on AMF’s spores’ microscopic characteristics, including spore sizes, shapes and color, three most popular morphological groups were selected to preserve and culture, namely AM-N1, AM-N2 and AM-N3. The AMF groups were identified to belong to three distinct fungal groups of genus Glomus. 15 Figure 3. 5. Phylogenetic tree of AM-N1, AM-N2, AM-N3 and published fungal species basing on parial 18S rRNA gene sequences (Maximum likelihood method, 100 bootstrap replicates). Gigaspora margarita BEG152 was the out group. By analyzing partial 18S gene sequences after DNA isolation and Nested-PCR amplification, the AMF strains were taxonomically identified. The results unraveled relationship between selected AMF strains and published species on GenBank, most of which were found to belong to the genus Glomus. The phylogenetic tree of strains AM-N1, AM-N2 and AM-N3 was constructed and depicted in Figure 3.5. Taken these above results together, the isolated AMF strains from turmeric rhizosphere samples were identified as Glomus sp. AM-N1, Glomus intraradices AM-N2 and Glomus mosseae AM-N3. 16 3.3. Impact of N fertilizer rates to turmeric rhizosphere microbial communities 3.3.1. Abundance of turmeric rhizosphere bacterial communities under varied N fertilizer rates 3.3.1.1. Diversity indices of turmeric rhizosphere bacterial communities Achieved bacterial OTUs from turmeric rhizosphere samples of four fertilization regimes (N0, N150, N350 and N500) were analyzed using bioinformatic softwares and Unite database. The average number of fungal species in samples of the N0, N150, N350 and N500 regimes were 3105, 2286, 2760 and 2843, respectively (Table 3.16). Table 3. 1. 16S amplicons metagenome sequencing data and analyzed diversity indices. Sample name N0 (n=3) N150 (n=3) N350 (n=3) N500 (n=3) Averaged species number 3105 2286 2760 2843 Simpson index 0.703 ± 0.11 0.665 ± 0.08 0.712 ± 0.10 0.733 ± 0.12 ACE index 2012 ± 354 1893 ± 196 2068 ± 244 1992 ± 259 Chao1 index 1765 ± 103 1994 ± 262 1549 ± 306 1266 ± 220 Shannon-Weaver 3.77 ± 0.50 2.95 ± 0.25 4.04 ± 0.71 3.73 ± 0.56 Coordinate analysis (PCoA) based on biodiversity indices (Figure 3.10) allowed a separation of group L (blue) from group H (red). In other words, there was a difference between the bacterial communities from the high yielding group H (regimes N150 and N350) and the low yields L (regimes N0, N500). 3.3.1.2. Taxonomic composition of turmeric rhizosphere bacterial communities 17 The analysis results showed relatively uniform compositions of turmeric rhizosphere bacterial communities at phylum and class levels of groups L and H. Accordingly, predominant proportions of Alpha-proteobacteria belonging to Proteobacteria were observed in both groups, accounting for over 40% of the total identified OTUs (p <0.05). Similarly, class Gemm-1 of Gemmatimonadetes appeared most analyzed samples with over 5% of total OTUs (p<0.05). The analysed 16S amplicons metagenome data suggested on one hand an indigenous soil bacterial communities structure in the research area, on the other hand a specific composition of rhizosphere bacterial communities of turmeric plant. Noteworthy, the Gram negative bacteria of Proteobacteria, whose representers such as free-living nitrogen-fixing and poor nutritional conditions adapters [171], were abundantly found in rhizosphere samples of regime N0. Most likely, the N nutritional poverty condition in regime N0 inhibited other microbial groups, promoting competitive advantage and increased distribution of Proteobacteria, especially those belong to Alphaproteobacteria. In general, after analyzing NGS sequencing data of ITS and 16S metagenome amplicons of C. longa rhizosphere samples, a total of 4776 OTUs of bacteria and 1760 OTUs of fungi were obtained. Bacterial OTUs are identified mainly to belong to Proteobacteria (40-60%), Firmicutes (5-20%), Crenarchaeota (1-15%), Gemmatimonadetes (5-10%), Acidobacteria (5-10%), Actinobacteria (3-10%) and Chloroflexi (3-8%). The majority of identified fungal OTUs was Ascomycota (40-85%), Basidiomycota (10-60%) and about 1-12% of the remaining OTUs belonged to Mortierellomycota, Glomeromycota and Chytridiomycota. The obtained results suggested the dynamic of microbial taxonomical groups in rhizosphere soils under the influence of changing farming conditions (especially N fertilizers) and in addition proposed a relationship 18 between the structure of rhizosphere soil microbiota with productivity and quality of plants. 3.3.2. Abundance of turmeric rhizosphere fungal communities under varied N fertilizer rates 3.3.2.1. Diversity indices of turmeric rhizosphere fungal communities The fungal composition of turmeric rhizosphere at four fertilization regimes (N0, N150, N350 and N500) was determined by sequence analysis of ITS amplicons from total DNA samples. After analyzing the sequence using Uparse software, comparing with Unite database, OTUs of rhizosphere samples were identified (Table 3.17). Table 3. 2. ITS amplicons metagenome sequencing data and analyzed diversity indices. Sample name N0 (n=3) N150 (n=3) N350 (n=3) N500 (n=3) Averaged species number 736 588 718 688 Simpson index 0,927 ± 0,21 0,749 ± 0,09 0,823 ± 0,11 0,833 ± 0,16 ACE index 770 ± 113 623 ± 91 755 ± 106 783 ± 75 Chao1 index 761 ± 83 645 ± 136 754 ± 98 788 ± 177 Shannon-Weaver 5,76 ± 0,3 4,05 ± 0,70 4,43 ± 0,52 4,85 ± 0,48 The principle component analysis (PCA) for OTUs of rhizosphere soil samples showed a divergence of group H (N150 and N350) from low yield group L (N0, N500). This result suggested that the amount of N fertilizer inputs not only affect turmeric yield and curcuminoids content but also impact the rhizosphere fungal community structure. Chemical fertilizers had been determined to be an alteration factor to microbiological system in different agricultural soils [169] [170], however this is the first time its impact on rhizosphere fungal community of turmeric plant C. longa being structurally analyzed and reported. 19 3.3.2.2. Taxonomic composition of turmeric rhizosphere fungal communities The taxonomic composition at phylum and class levels of turmeric rhizosphere fungal communities in all investigated regimes appeared relatively uniformed. Accordingly, OTUs of Ascomycota were prevalent; especially in soil samples of regime N150 with 74.16±9.06% of total identified OTUs. Besides, the proportion of OTUs belonging to Basidiomycota tended to increase as the raising amount of N fertilizer. In addition, several fungal genera such as Zygomycota, Rozellomycota and Blastocladiomycota were found at higher rates in soil samples of L group (N0 and N500). However, statistical analysis showed no significant differences in abundance of Basidiomycota, Zygomycota, Rozellomycota and Blastocladiomycota between H and L groups (p>0.05). At class level, OTUs of Sordariomycetes were distributed more intensively in samples of regimes N350 and N500, while Eurotiomycetes appeared with a significantly higher proportion in samples of N0 regime (p<0.05). Agaricomycetes was found as the most dominant fungal class of Basidiomycota with significant increased abundance in samples of H group (p<0.05). Findings of dominant fungal classes in the microbiota of turmeric rhizosphere as well as their dynamics under varying N fertilizing rates were unprecedented. These results have contributed to specify a fundamental reference for extensive studies on optimal fertilization dosages for turmeric C. longa in Vietnam on worldwide scale. 3.3.3. Differences in composition of selected effective microbial groups in turmeric rhizosphere Based on characterized effective microorganisms in turmeric rhizosphere samples (Section 3.2), an oriented analysis in the rhizosphere microbial communities of turmeric was performed, 20 focusing on the difference in abundances of PGPR belonging to Bacilliaceae and AMF belonging to Glomeromycota in four N fertilization regimes. 3.3.3.1. Bacteria of Bacilliaceae Depending on the distribution of OTUs belonging to Bacilliaceae of turmeric rhizosphere samples, three main taxonomical groups were determined: genus Bacillus, genus Geobacillus and unclassified bacteria. Notably, the average proportion of OTUs belonging to Bacillus in the samples of N0 and N500 regimes (regimes of L group) accounted for 64 and 65% of the total OTUs of the Bacilliaceae family, respectively, and reached the peak at over 90% in samples under N150 and N350 regimes (H group). By statistical analysis, the difference in abundance of Bacillus-OTUs between two groups was determined to be significant, with p<0.05. These results suggest the role of some Bacillus bacteria in turmeric yield and quality. In particular, the finding may contribute to the orientation for preparation of biofertilizers from the PGPR strain belonging to Bacillus for turmeric farming in Vietnam. 3.3.3.2. Fungi of Glomeromycota Statistical analysis results in taxonomic composition of the turmeric rhizosphere of groups L and H showed no difference in proportion of fungi belonging to Glomeromycota at phylum level. However, the difference between two groups was significant when considering at the genus level (p<0.05). In detail, the OTUs belonging to genus Glomus in total OTUs of Glomeromycota accounted for 13 and 10% in N0 and N500 regimes of L group, respectively, and for 59 and 36% in N150 and N350 of H group, respectively. 21 In general, metagenome analysis at the genus level of Bacilliaceae and Glomeromycota revealed statistically significant differences between H and L groups, especially at the abundance of OTUs belonging to genera Bacillus and Glomus, of these effective representatives were isolated in the turmeric rhizosphere samples. These results have contributed confirming the prediction of effective microorganism groups correlating to turmeric yield and quality, thereby played a pivotal role for selection of indigenous microbial candidates for preparation of biofertilizer for turmeric plant C. longa. 3.4. Case study for effect of microbial inocula in turmeric As a combination of isolating result (section 3.2) and genetic analysis (section 3.3.3), a biofertilizer from strains of PGPR and AMF was prepared and tested in turmeric plants. The microbial inocula were composed of Bacillus sp. PGP-V21 and 3 AMF strains (Glomus sp. AM-N1, Glomus intraradices AM-N2 and Glomus mosseae AM-N3). These indigenous microbial strains were determined to be safe in animal models. Table 3. 3. Growth and productivity parameters of turmeric p