Plant-microbe rhizosphere interactions mediated by Rehmannia glutinosa root exudates under consecutive monoculture

Under consecutive monoculture, the biomass and quality of Rehmannia glutinosa declines significantly. Consecutive monoculture of R. glutinosa in a four-year field trial led to significant growth inhibition. Most phenolic acids in root exudates had cumulative effects over time under sterile conditions, but these effects were not observed in the rhizosphere under monoculture conditions. It suggested soil microbes might be involved in the degradation and conversion of phenolic acids from the monocultured plants. T-RFLP and qPCR analysis demonstrated differences in both soil bacterial and fungal communities during monoculture. Prolonged monoculture significantly increased levels of Fusarium oxysporum, but decreased levels of Pseudomonas spp. Abundance of beneficial Pseudomonas spp. with antagonistic activity against F. oxysporum was lower in extended monoculture soils. Phenolic acid mixture at a ratio similar to that found in the rhizosphere could promote mycelial growth, sporulation, and toxin (3-Acetyldeoxynivalenol, 15-O-Acetyl-4-deoxynivalenol) production of pathogenic F. oxysporum while inhibiting growth of the beneficial Pseudomonas sp. W12. This study demonstrates that extended monoculture can alter the microbial community of the rhizosphere, leading to relatively fewer beneficial microorganisms and relatively more pathogenic and toxin-producing microorganisms, which is mediated by the root exudates.

Scientific RepoRts | 5:15871 | DOi: 10.1038/srep15871 ( Supplementary Fig. S1a). However, only eight of the phenolic acids, excluding coumaric acid and salicylic acid, were found in the rhizospheric soil of R. glutinosa. The levels of most rhizosphere phenolic acids did not increase with increasing years of monoculture ( Supplementary Fig. S1b), suggesting that accumulation of root exudates and their possible autotoxicity might not be the direct cause for consecutive monoculture problems. We hypothesized that soil microbes might be involved in the degradation, utilization, and conversion of root exudates. For combined restriction enzymes, NMDS ordinations (relative abundance, logarithmic transformation, Bray-Curtis dissimilarity matrix) showed clear differences in both bacterial and fungal communities between the five different treatments (Fig. 2a,b). Kruskal's stress values were less than 0.1, suggesting a good representation of the relationship between points in the matrix. Cluster analysis (relative abundance, logarithmic transformation, Bray-Curtis dissimilarity matrix) of bacterial and fungal communities showed that the consecutively monocultured soils (SM, TM, and FOM) were more similar and considerably different from the newly planted soils (NP), indicating a shift in the rhizosphere microbial community following extended R. glutinosa monoculture (Fig. 2c,d).

Analysis of similarity (ANOSIM) and similarity percentage analysis (SIMPER).
Analysis of similarity (ANOSIM) of the T-RFLP data consisting of the relative abundance of T-RFs showed that both bacterial (ANOSIM Global R = 1.0, P = 0.001) and fungal (ANOSIM Global R = 1.0, P = 0.001) communities differed significantly among five different treatments. ANOSIM analysis based on the presence/absence of T-RFs also showed a significant difference among five treatments in both bacterial (ANOSIM Global R = 0.858, P = 0.001) and fungal (ANOSIM Global R = 1.0, P = 0.001) communities. When SM, TM, and FOM were combined as one sample and termed "consecutively monocultured soil (CM)", SIMPER analysis (relative abundance of peaks) showed pairwise dissimilarities between NP and CM were 23.9% for the bacterial community and 64.37% for the fungal community, respectively. The top T-RFs with 20% cumulative contribution to the dissimilarity between NP and CM were listed as Table 1.
For the bacterial community, consecutive monoculture produced a significant increase in the relative abundance of fragments including HaeIII 37 and 295 bp, AluI 252 bp and MspI 150 bp in the rhizosphere with the increasing years of monoculture, but a decrease in MspI 293 bp, HaeIII 205 bp, AluI 249 and 72 bp. Assignation from PAT has identified the fragment AluI 72 bp to be from Pseudomonas spp. in the class GammaProteobacteria. With regard to fungal community, T-RFs including AluI 428 and 588 bp, HinfIII 65 bp were significantly higher in newly planted soil than in consecutively monocultured soil while T-RFs including TaqI 251 and 357 bp, HinfIII 304 bp were significantly higher in consecutively monocultured soil (P < 0.05). The experimental T-RFs were determined by the in silico T-RFs deduced from the ITS sequences from GenBank and 50 fungal isolates. Several T-RFs (i.e. HinfIII 304 bp and TaqI 251 bp) constituting the highest proportion of the fungal population in two or three-year monoculture soils were assigned as Fusarium oxysporum ( Table 1).
Abundance of Pseudomonas and F. oxysporum by quantitative PCR. Quantitative PCR confirmed the presence of Pseudomonas and F. oxysporum in all five soil samples. The density of Pseudomonas was significantly lower in extended monoculture soils than in control and newly planted soils while the amount of F. oxysporum was significantly greater in extended monoculture soil than in control and newly planted soils (Fig. 3). The result was consistent with the T-RFLP analysis (Table 1). Potential pathogens with the highest isolation frequency from extended monoculture soil and infected R. glutinosa were identified as F. oxysporum or Fusarium sp. (44%) (Supplementary Table S1). The ratio of Pseudomonas to F. oxysporum was significantly greater in newly planted soil than in extended monoculture soils. In vitro antagonism assays showed that the percent of Pseudomonas spp. possessing antagonistic activity against F. oxysporum gradually decreased with increasing years of monoculture (Fig. 4). Most of the Pseudomonas strains with antagonistic activity against F. oxysporum were the same and named as Pseudomonas sp. W12 in this study.   that detected in rhizosphere soil, significantly promoted mycelial growth of F. oxysporum (Fig. 5a). Among the eight phenolic compounds identified in the soil, vanillic acid and ferulic acid had the highest stimulatory effect on the growth of F. oxysporum ( Supplementary Fig. S4a). The growth promoting effect of the   mixture was greater than that of single compounds. We also found that the phenolic acid mixture significantly promoted sporulation of F. oxysporum and the promoting effect increased as the concentration increased ( High pathogenicity of F. oxysporum and biocontrol effects of Pseudomonas sp. W12. The purified isolate of F. oxysporum rapidly caused wilt disease on the tissue culture seedlings of R. glutinosa. This occurred in the tissue culture vessels under sterile conditions and also in pots with sterilized soils (Fig. 6a,b). F. oxysporum invades roots and colonizes xylem vessels to cause wilt disease. When the pathogenic fungus was isolated from the infected plant parts it was verified to be the same as the inoculated strain based on morphological observation and ITS region sequencing. This demonstrated the high pathogenicity of F. oxysporum isolate to the R. glutinosa. The antagonistic effect of Pseudomonas sp. W12 against F. oxysporum was evaluated. Transplanted tissue culture seedlings of R. glutinosa treated with F. oxysporum (inoculated around 4 cm away from the seedling root) (CK) soon developed disease symptoms, withered, and died. However, using Pseudomonas sp. W12 (inoculated around 2 cm away from the seedling root) to inhibit F. oxysporum (inoculated around 4 cm away from the seedling root) (W12 treatment), the tissue culture seedling of R. glutinosa grew well and exhibited no disease symptoms during the entire experiment period (Fig. 6c). Quantitative PCR confirmed that the amount of Pseudomonas was significantly higher in the W12 treatment than in the control (CK) while the amount of F. oxysporum showed the opposite trend (Fig. 6d), which was consistent with the disease severity of the seedlings. These results indicate that the exogenous addition of antagonistic Pseudomonas can effectively decrease plant infection by F. oxysporum. The results also demonstrate that imbalances between these two strains (Pseudomonas sp. W12 and F. oxysporum) can lead to serious disease problems in extended monoculture. The results clearly show the biocontrol potential of the antagonistic Pseudomonas sp. W12.

Discussion
The consecutive monoculture problem, also known as replanting disease, is typically manifested as a dramatic reduction in plant growth, with a shortened production life, and a decline in biomass and yield 22 . Our four-year field experiment showed the typical growth inhibition effects caused by extended R. glutinosa monoculture (Fig. 1). Two-year consecutive monoculture can lead to complete loss of the root tubers. More attention is being given to the biological relationships that occur between plants and microorganisms in the root zone, which are crucial for plant growth and health 8,23 . In this study, T-RFLP results showed significant shifts in both bacterial and fungal communities in rhizosphere after extended monoculture of R. glutinosa (Fig. 2). Latz et al. 24 suggested that plant diversity improved protection against soil-borne pathogens by fostering antagonistic bacterial communities while plant monocultures had a negative impact on the abundance of antagonistic bacteria. Based on T-RFLP PAT assignation, we found that extended monoculture of R. glutinosa resulted in significant reduction of Pseudomonas populations (Table 1). Quantitative PCR assay confirmed the decline in Pseudomonas abundance (Fig. 3). Furthermore, plate counts showed that the relative abundance of Pseudomonas spp. with antagonistic activity against pathogenic F. oxysporum was significantly lower in extended monoculture soil than in newly planted soil (Fig. 4). Previous studies demonstrated that the increase of antagonistic Pseudomonas spp. that control specialized pathogens increased disease suppression in soils 25,26 . Disease-suppressive soil has been studied for take-all decline of wheat, which occurs after several consecutive years of wheat monoculture 26 . The decline of take-all disease resulted from the increase of antagonistic Pseudomonas populations that produce antifungal compounds such as 2,4-diacetylphloroglucinol [27][28][29] . Induction of disease suppression in soils after several years of crop monoculture has been reported for several diseases 8 . In our four-year field trial, we did not observe the development of disease suppression, perhaps due to the limited monoculture duration. However, we found the abundance of antagonistic Pseudomonas spp. decreased with duration of the monoculture indicating the importance of antagonistic Pseudomonas spp. in soil health and plant performance.
An enormous variety of potentially valuable low-molecular weight organic compounds are exuded by plant roots into the rhizosphere. These compounds in the rhizosphere have an ecological role in plant-microbe-soil interactions 30,31 . We identified pathogenic bacteria in the extended monoculture soil that might benefit from fresh carbon released by R. glutinosa roots. F. oxysporum or Fusarium sp. corresponding with T-RFs HinfIII 304 bp and TaqI 251 bp were significantly more abundant in extended monoculture soil than in newly planted soil (Table 1). Fusarium spp. are a well-known pathogens for many important crops. Culture-based study and quantitative PCR showed that F. oxysporum was a common pathogen in replanted soil and its density was significantly higher in extended monoculture soils (Fig. 3, Supplementary Table S1). Root exudates have pronounced selective effects on specific microbial populations in soil, which can deter one organism while attracting others 14,32 . Zhou et al. 33 found p-coumaric acid, an autotoxin of cucumber, increased F. oxysporum f.sp. cucumerinum Owen (a soil-borne pathogen of cucumber) population densities in soil and increased the severity of Fusarium wilt under field conditions. Sugars and amino acids in root exudates can promote spore germination and mycelial growth of the soil-borne pathogen F. oxysporum 34 . We found that the phenolic acid mixture, at the same ratio found in the R. glutinosa rhizosphere, could greatly promote the mycelial growth, sporulation, and toxin production of pathogenic F. oxysporum (Fig. 5). However, the test mixture could greatly inhibit the growth of beneficial Pseudomonas sp. W12 (Fig. 5). Bais et al. 35 found that rosmarinic acid, an antimicrobial compound released by Ocimum basilicum roots upon microbe attack, had a significant deleterious effects on Pseudomonas aeruginosa. P. aeruginosa is commonly used as a biocontrol agent for management of Fusarium wilt disease 36 .
Our results indicate that the biomass decline of R. glutinosa under four-year extended monoculture is influenced by at least two factors: i) the decline of beneficial bacteria with antagonistic activity against pathogens in the rhizosphere and ii) an increase in disease susceptibility due to poor growth of R. glutinosa at a time when pathogenic microbes have became dominant (Fig. 7). This knowledge is essential in developing methods to solve consecutive monoculture problems in R. glutinosa and other crops. Such methods might include the use of microbial fertilizer and organic amendment application. Additional work is needed to characterize aboveground plant performance and belowground microbial diversity after a long-term monoculture of R. glutinosa, The mechanisms by which antagonistic microbes control disease organisms would also be a useful research area.

Methods
Field experiment and soil sampling. R. glutinosa cultivar 'Wen 85-5' , common in the main production region, was used as the test plant material. R. glutinosa was typically planted on April 15 and harvested by October 30. After harvest, fields were kept fallow from October 31 to the following April 15. The experiment was conducted at Wenxian Agricultural Institute, Jiaozuo City, Henan Province (34°56′ N, 112°58'E). This area is the "geo-authentic" zone for R. glutinosa cultivation. It has a continental monsoon climate, annual mean temperature of 14.3 °C and an annual mean precipitation of 552 mm. A field previously cultivated with wheat was used for this experiment (sandy loam soil, total nitrogen 0.48 g·kg −1 , available nitrogen 21.34 mg·kg −1 , total phosphorus 1.36 g·kg −1 , available phosphorus 47.02 mg·kg −1 , total potassium 7.23 g·kg −1 , available potassium 251.34 mg·kg −1 , soil organic matter 10.32 g·kg −1 and pH 7.52). The treatments were: i) the newly planted (NP), ii) two-year consecutive Scientific RepoRts | 5:15871 | DOi: 10.1038/srep15871 monoculture (SM), iii) three-year consecutive monoculture (TM), iv) four-year consecutive monoculture (FOM) and v) control with no R. glutinosa cultivation (CK) ( Table 2). The above treatments were organized within a single field site with the same climatic conditions and closely adjacent to each other. Each treatment had three replicate plots and the study plots were completely randomized. Prior to planting, each replicate plot received four fertilizers: 1.6 kg N-P-K complex fertilizer, 1.25 kg (NH 4 ) 2 HPO 4 , 1.6 kg Ca(H 2 PO 4 ) 2 , and 0.8 kg K 2 SO 4 . All treatments were subjected to the same fertilization and field management during the entire experimental period.
On August 1, 2013 we collected soil samples from 5 random locations within each plot because of the significant difference in growth status between different treatments on this date (Fig. 1). Three plants were sampled at each location. Soil samples from all five treatments were collected at the same time. Fresh plants were carefully uprooted from the soil with a forked spade and slightly shaken to remove loosely attached soil. The rhizosphere soil that was tightly attached to roots and rhizomes was brushed off and collected. Plots without R. glutinosa served as a bulk soil control for contrast with rhizosphere communities under the influence of the host plant. Soil samples were sieved through 2 mm mesh to determine chemical properties (Supplementary Table S2), extraction of soil DNA, and identification of soil phenolic acids. Some soil samples were stored in 4 °C for up to a week for microbe isolation.
The extraction and quantification analysis of phenolic acids from root exudates and rhizosphere soil. A series of tissue culture seedlings of R. glutinosa were incubated under sterile condition for different lengths of time. Each vessel had 30 mL of Murashige-Skoog medium (supplemented with 0.2 mg/L 6-BA, 0.2 mg/L indole-3-butyric acid and 30 g/L sucrose) and three tissue culture seedlings. The root exudates in MS medium were first extracted by using 1 mol/L NaOH. The extract was adjusted to pH 2.5 with HCl and extracted five times with an equal volume of ethyl acetate. The pooled extracts were evaporated to dryness at 35 °C and dissolved in 5 mL methanol. Quantification analysis of phenolic acids in root exudates was carried out using high performance liquid chromatography (HPLC) (Shimadzu,   Japan) equipped with an ODS-C18 column (Inertsil ODS-SP, 4.6 × 250 mm, 5 μ m, Japan). The mobile phase was a mixture of 28% phase A (100% methanol) and 72% phase B (2% acetic acid) maintained at a constant flow of 0.8 mL/min. The column temperature was maintained at 30 °C and detection was performed at 280 nm. Identification and quantification of phenolic acids were performed based on the retention times and the addition of pure standards to the samples. The extraction and quantification analysis of phenolic acids from rhizosphere soil was the same as that of root exudates of tissue culture seedlings.
Purified fungal ITS fragments were digested with restriction endonucleases AluI and HinfI for 5 h at 37 °C, and with TaqI for 5 h at 65 °C. The length of terminal restriction fragments (T-RFs) was determined by the ABI 3730xl DNA sequencer (Applied Biosystems, Foster City, USA) in the GeneScan mode. Additional details on the experimental protocol of T-RFLP analysis were presented in Supplementary Material 1.
The T-RFLP Phylogenetic Assignment Tool (PAT) was used to find possible phylogenetic assignments of bacterial T-RF lengths 37 . For fungal T-RF identification, a database of ITS sequences from GenBank (3292 sequences downloaded using keyword 'ITS1F') and 50 fungal isolates obtained from the consecutive monoculture soil and infected Rehmannia plants (Supplementary Table S1) was established and analyzed for predicted T-RF lengths with the TRiFLe Program 38 . The T-RFs similar in size of ± 1-2 bp were summarized to operational taxonomic units (OTUs) 39 . Only phylogenetic assignments that matched the experimental T-RF lengths for all three restriction enzymes were used in this study.

Statistical analyses of T-RFLP data.
We analyzed the T-RFLP profiles by GeneMarker Version 1.2 (SoftGenetics LLC, PA, USA) according to manufacturer's instructions. The relative abundance of T-RFs was calculated as the peak area of the each T-RF divided by the total peak area of all T-RFs detected within a fragment length range between 35 and 700 bp.
Non-metric multidimensional scaling (NMDS) ordinations were used to examine relative similarities of microbial composition, which use an iterative algorithm to successively refine sample positions in the ordination with Kruskal's stress value used as a measure of goodness of fit. A stress value less than 0.1 indicates an accurate and reliable ordination with a low probability of misinterpretation. NMDS is more robust method for investigating the microbial community data derived from T-RFLP profiles than principal components analysis (PCA) when examining non-linear data 40 . Analysis of similarity (ANOSIM) was used to examine the statistical significance between samples. Similarity percentage analysis (SIMPER) was used to assess the relative contribution (%) of each T-RF to the dissimilarity between samples, typically on an abundance basis 40 . The above multivariate statistical analyses as well as clustering analysis based on the T-RFLP data were performed with the PRIMER V5 software package (PRIMER-E Ltd, Plymouth, UK).
Quantitative PCR for Pseudomonas and F. oxysporum. The quantitative PCR assay was used to quantify the genus Pseudomonas and F. oxysporum. Quantitative PCR was performed in 15 μ l reaction mixture containing 7.5 μ l 2 × SYBR green I SuperReal Premix (TIANGEN, Beijing, China), 0.5 μ l of each primer (10 μ M) and template DNA (20 ng of total soil DNA or a serial dilution of plasmid DNA for standard curves). The taxon-specific primer sets and their annealing temperatures are listed in Supplementary Table S3. Four independent quantitative PCR assays were performed for each treatment.
Isolation and counting of Pseudomonas spp. with antagonistic activity. Pseudomonas spp.
were isolated by Pseudomonas selective isolation agar (PSIA) 41 . Rhizosphere suspensions were prepared, serially diluted and plated onto PSIA. PSIA plates were incubated at 32 °C for 18 h. From each treatment, all colonies were purified and screened for in vitro antagonism towards F. oxysporum. For in vitro antagonism assays, the bacterial isolates were inoculated at the periphery of potato-dextrose-agar (PDA) plates and incubated for two days at 28 °C, after which the isolated F. oxysporum was transferred to the centre of the plates. After an additional three days of incubation at 28 °C, Pseudomonas spp. with antagonistic activity against F. oxysporum were recorded.
The effects of phenolic acids on the growth of Pseudomonas sp. W12 and F. oxysporum. Based on HPLC analysis, eight phenolic acids including protocatechuic acid, phthalic acid, p-hydroxybenzoic acid, vanillic acid, syringic acid, vanillin, ferulic acid, benzoic acid were identified in rhizosphere soil. These standard compounds were first dissolved in small amount of methanol and then prepared with distilled water into a mixture at the same ratio as detected in rhizosphere soil of R. glutinosa. The isolated F. oxysporum was inoculated onto the center of 9 cm diam Petri dishes filled with a 10-fold dilution of soil-extract agar medium (SEM) containing different final concentrations of single phenolic Scientific RepoRts | 5:15871 | DOi: 10.1038/srep15871 acids or mixtures (30,60,120,240,480 and 960 μ mol/L). The soil-extract medium added into the same amount of distilled water containing equal amount of methanol was used as the control. Each treatment had three replicates. After eight days of incubation at 28 °C in 0:24 L:D, colony diameter was measured. Sporulation of F. oxysporum was achieved by adding the test mixture to 8-fold dilution of potato-sugar-agar (PSA) broth medium. After seven days of incubation in a liquid shaker with 180 rpm at 28 °C, the spores were counted by hemocytometer and converted to the number of conidia in a liquid culture. Likewise, for toxin detection, the same amount of spore suspension was inoculated into a 10-fold dilution of SEM broth medium containing different concentrations of phenolic acid mixture and incubated in a liquid shaker with 180 rpm at 28 °C for ten days. At the end of the incubation period the fermentation broth was filtered through 0.22 μ m filter membrane and extracted with an equal volume of ethyl acetate in a shaker with 160 rpm for 8 h. After centrifugation, the organic phase in the upper layer was vacuum-dried and dissolved in distilled water. Three toxins in the extracts including deoxynivalenol (DON), 3-Acetyldeoxynivalenol (3A-DON) and 15-O-Acetyl-4-deoxynivalenol (15A-DON) were detected using the HPLC system with an ODS-C18 column (Inertsil ODS-SP, 4.6 × 250 mm, 5 μ m). The mobile phase consisted of 100% acetonitrile (A) and 0.005% phosphoric acid (B) with a gradient elution of 0 min; 5% A and 95% B → 9 min; 70% A and 30% B → 18 min; 100% A → 23 min; 100% A → 23.01 min; 5% A and 95% B → 45 min; 5% A and 95% B at a constant rate of 0.7 mL/min. The UV monitor was set to 224 nm. The column temperature was maintained at 35 °C. The toxins were identified by comparing the retention times with those of the corresponding standard compounds.
The effects of single phenolic acids and their mixture at the same ratio as detected in rhizosphere soil on the growth of isolated Pseudomonas sp. W12 were determined by adding the test compounds to an 8-fold dilution of LB broth medium. The LB broth medium added into the same amount of distilled water containing an equal amount of methanol was used as the control. After 8-10 h incubation in a liquid shaker at 200 rpm and 28 °C, bacterial culture density, expressed as optical density (OD) was determined at 600 nm with a plate reader (Thermo Scientific Multiskan MK3, Shanghai, China).
Assessment of the pathogenicity of F. oxysporum and the biocontrol potential of Pseudomonas sp. W12. Tissue culture seedlings of R. glutinosa were incubated on specific MS medium for 45 d and then transplanted to plastic pots filled with sterilized growing medium for 14 d. The pots were placed in a growth chamber at 25 °C with a photoperiod (L:D) of 16:8. Isolated F. oxysporum was inoculated in a liquid shaker at 180 rpm and 28 °C at (L:D) 0:24 for 2 d. Then spore suspension was added to the tissue culture vessels under sterile conditions and the transplanted seedlings to test the effects of F. oxysporum on the development of Fusarium wilt of R. glutinosa. Assessment of biocontrol potential of isolated Pseudomonas sp. W12 was conducted by exogenous addition of Pseudomonas sp. W12 (inoculated 2 cm away from the seedling root) and F. oxysporum (inoculated 4 cm away from the seedling root). The transplanted tissue culture seedlings were only treated by F. oxysporum (inoculated 4 cm away from the seedling root). Equal amounts of LB broth medium were used as the controls. Each treatment had three replicates.