Materials and Methods

Culture: E. crassipes plants were collected from a lake from the Western part of Guangdong University of Technology, Guangzhou, China in fall (September). The plants were cultured in nutrient solution (0.25 mM Ca(NO3)2·4H2O, 0.25 mM KNO3, 0.1 mM MgSO4·7H2O and 0.05 mM KH2PO4) for 20 days at 25 °C with 16 h of light (2000 Lx) and 8 h of darkness indoors. The solution was replaced completely every two days. After preculture, the E. crassipes plants were cultured in either 1.8 mM NH₄Cl solution, 1.8 mM KNO₃ solution, or distilled water for two weeks under the same conditions as the preculture.

Sample preparation: The first 1-2 cm of root tips (90 g) of E. crassipes mixed with 90 mL sterile water and some glass beads were vortexed (2 000 rpm) for 20 min and then allowed to rest for 20-30 s. The supernatant solution was collected. Some of it was stored in -80 °C for DNA extraction and some was used as a sample solution which was diluted in 10-fold steps with sterile water.

Determination of the number of culturable rhizosphere microbial communities: The densities of microbial communities in the plant rhizosphere were determined by a conventional culture method using the following media: salt beef extract tryptone medium for bacterial culture; Gao 1 synthetic medium for actinobacteria, with the addition of potassium dichromate (100 µg/mL) to inhibit bacteria and molds and salt Czapek Dox agar for mold culture. Diluted sample (100 µL) was spread onto plates and cultured at 37 °C for one or two days (for bacteria) or 28 ºC for a week (for actinobacteria and molds). The number of clones in each type of culture medium was determined. Each solution was assayed in triplicate to obtain the average (^{Matsuzawai, Tanaka, Tamaki, Kamagata, & Mori, 2010}).

Extraction of DNA and PCR amplification: DNA was extracted from rhizosphere samples prepared as described above using a soil DNA extraction kit (Hangzhou Haoji Biotechnology Ltd.) according to the manufacturer's instructions. Two rounds of PCR amplification were performed with primers listed in Table 1. The primers for this procedure were designed from the V3 region of bacterial 16s rRNA. The conditions for the first amplification were 94 °C for 2 min, 35 cycles (94 °C for 30 s, 58 °C for 30 s and 68 °C for 1.5 min). The conditions for the second amplification were 94 °C for 2 min, 30 cycles (94 °C for 30 s, 55 °C for 45 s and 68 °C for 30 s). PCR products were separated on 1 % agarose/EtBr gel. Amplification experiments were performed in triplicate on every sample (^{Mota et al., 2008}).

Table 1: Primer sequences for PCR 

Denaturing gradient gel electrophoresis (DGGE) analysis of microbial rRNA genes: PCR products were purified with a purification kit (Hangzhou Haoji Biotechnology Ltd.) and then separated by DGGE using the Decode Universal Mutation Detection System (BioRad, USA) with a run time of 12 h in Trisacetate-EDTA buffer heated to 60 °C. Bacterial 16s rRNA gene PCR products were electrophoresed in an acrylamide gel with a gradient of 35-60 % denaturant. After electrophoresis, the gel was stained with SYBR Green I for 30 min and the bands were examined using the Gel Doc 2000 gel imaging system (Bio-Rad, USA).

Statistical analysis of DGGE image: The DGGE image was scanned with the Quantity One program and the intensity of every band was reported. The evenness and diversity indexes were compared by one-way analysis of variance and Tukey HSD test for post-hoc analysis using SPSS (13.0). Diversity was calculated as the Shannon index of diversity (H) to compare changes in diversity of microbial communities between the three treatments using the following function: H=-∑(P_i) log₂P_i, where P_i=n_i/N, n_i is the band intensity and N is the sum of all band intensities in the curve. Simpson's diversity index (D) for infinite populations is equal to 1-sum (P_i * P_i). The DGGE evenness (E), a measure of how evenly DGGE bands are distributed in a given sample, is calculated as E=H/log₂ (S) where richness (S) refers to the number of bands detected in a given sample.

Sequencing and analysis of bright bands: Clear differential display or common display bands were cut out and redissolved in Tris buffer (pH 8.0). The supernatant was subjected to PCR (94 °C for 2 min, 30 cycles; each cycle included 94 °C for 30 s, 55 °C for 45 s and 68 °C for 30 s) with the primers BacV3-F2 and Bac-V3-R2 (listed in Table 1). PCR products were recycled, cloned and sequenced. All sequences were blasted against the nonredundant nucleotide database in NCBI to identify the most similar sequences in the database (http://www.ncbi.nlm.nih.gov). These findings were used to classify of our sequences.

Results

Determination of the total numbers of culturable bacteria, actinobacteria and molds: The numbers of rhizosphere bacteria, actinobacteria and molds are listed in Table 2. Under each culture condition, the numbers of bacteria were the highest, followed by actinobacteria, and less molds. The numbers of bacteria were largely similar across the three culture conditions, whereas the numbers of actinobacteria and molds from the rhizosphere of E. crassipes cultured in NH₄Cl solution were two orders of magnitude higher than those from the rhizospheres of plants cultured in distilled water and KNO₃ solution.

Table 2: The number of rhizosphere microbiology from Eichnornia crassipes cultured in three conditions (cfu/mL) 

PCR amplification of extracted DNA and DGGE analysis: The results of electrophoresis of the second round of PCR products are shown in Fig. 1. There was only one 230 bp band in every lane (with each sample run in two lanes). All three samples from the DGGE results showed many clear bands (Fig. 2). The Shannon index of diversity (H), Simpson's diversity index (D) and the evenness (E) of every sample are listed in Table 3. Theyindicate that all three samples had an abundance of diversity in the microbial community of the rhizosphere. But there was no significant difference (all P > 0.05) among the Shan non index of diversity (H), among Simpson'sdiversity index (D), or among the evenness (E) of three samples.

Fig. 1: Electrophoresis of the PCR products (lanes 1 and 2, samples cultured in KNO3 solution; M, marker; lanes 3 and 4, samples cultured in NH4Cl solution; lanes 5 and 6, samples cultured in distilled water). 

Fig. 2: Denaturing gradient gel electrophoresis result (lanes 1, 4, and 7, samples cultured in KNO3 solution; lanes 2, 5, and 8, samples cultured in NH4Cl solution; lanes 3, 6, and 9, samples cultured in distilled water). The short segment beside the lane indicates the bands that were sequenced. 

Table 3: The Shannon index of diversity (H), Simpson's diversity index (D) and the evenness (E) of every sample 

Sequence analysis: Clear differential display bands or clear common display bands were selected to be sequenced. The selected sequences were used for BLAST against the non-redundant nucleotide sequence database to search for the most similar sequences whose classifications were known. Among the eight sequences from the sample cultured in KNO₃ solution, two were similar to sequences of Pseudoxanthomonas and two were similar to those of Acidovorax. The other four sequences were similar to those of Devosia, Novosphingobium, Enterobacter and Reyranella. Among the eight sequences from the sample cultured in NH₄Cl solution, there were sequences similar to those of Acinetobacter, Flectobacillus, Cytophaga, Streptomyces, Devosia, Enterobacter, Rhizobium and Citrobacter, whereas among the nine sequences cultured in distilled water, there were three sequences similar to that of Novosphingobium and six sequences similar to those of Pseudomonas, Pseudacidovorax, Acidovorax, Brevundimonas, Rhodospirillum and Methylosinus. Pseudoxanthomonas, Enterobacter and Citrobacter were found in bands from all three samples. The genus Reyranella was found only in samples cultured in KNO₃ solution; Acinetobacter and Streptomyces were unique to samples cultured in NH₄Cl solution, and Pseudomonas, Pseudacidovorax and Methylosinus were found only in samples cultured in distilled water.

Discussion

It has been previously reported that the diversity and abundance of different components of rhizosphere microbial communities were influenced by plant nutrition and other environmental factors. ^{Hamonts et al. (2013}) reported that nitrogen limitation significantly influenced the structure and function of denitrifier communities in the rhizosphere of wheat. An investigation by ^{Marschner et al. (2004}) showed that bacterial community structure was influenced by soil pH and type of phosphorus fertilisation. ^{Söderberg and Bååth (2004}) found that bacterial activity in the barley rhizosphere decreased when ammonium was added.

Our study also found that not only the numbers of culturable microbial communities but also the compositions of the microbial communities of E. crassipes were influenced by different forms of nitrogen. Although Pseudoxanthomonas, Enterobacter and Citrobacter were found in all three samples, different treatment samples had unique genera and dominant genera. Pseudoxanthomonas and Acidovorax were the dominant genera in the rhizosphere microbial community of E. crassipes cultured in KNO₃ solution, whereas Novosphingobium was the dominant genus in the sample cultured in distilled water. The differences among the rhizosphere microbial communities were a response to environmental factors. Because individual components of the rhizosphere microbial community had different functions that allowed them to exploit specific environmental conditions. In the environment with plenty of ammonium, the special microbial communities of E. crassipes rhizosphere can effectively absorb or transform the ammonium; while in the environment with much nitrate, the microbial communities of rhizosphere will change into adapted compositions which can take in or transform the nitrate. Through all these microbial communities of rhizosphere, the nitrogen can be removed from water or transformed into other forms which can be utilized by other organisms. The study of the effects of various forms of nitrogen on these communities can provide a theoretical foundation for rhizosphere microbiology in phytoremediation, plant growth, plant protection and other areas. In the rhizosphere microbial community of E. crassipes, several bacterial species are common in soil and water. Some of them, including Acidovorax, Acinetobacter, Streptomyces, Enterobacter, Citrobacter, Novosphingobium and Pseudomonas, can decompose organic or macromolecular compounds. Some are nitrogen fixers, such as Rhizobium, Devosia, Pseudacidovorax and Rhodospirillum. Methylosinus is an example of a methanotroph. Many bacteria are very important in phytoremediation (^{Lu et al., 2010}; ^{Bai, Liang, Liu, Hu, & Qu, 2014}). But from the results of our study, we could not identify specific nitrifying or denitrifying bacteria in the NH₄Cl or KNO₃ treatment samples, probably because the removal of nitrogen in water with NH₄Cl or KNO₃ is a complex process that comprises not only nitrification and denitrification, but also absorption and use by bacteria and plants. Thus, nitrogen removal represents the integrated action of plants and bacteria in phytoremediation.

There are several likely reasons for the effects of different forms of nitrogen on rhizosphere microbial communities. First, the form of nitrogen present influences the pH of culture media (^{O'Donnell, Seasman, Macrae, Waite, & Davies, 2001}; ^{Kemmitt, Wright, Goulding, & Jones, 2006}). Nitrogen in the form of ammonium decreases the pH of the soil, while nitrogen in the form of nitrate increases the pH. Ma et al. (2008) reported that ammonium increased the numbers of actinobacteria and fungi in the rhizosphere of Yumai 49 (a medium-gluten wheat), and our study confirmed this result. The numbers of culturable actinobacteria and molds from the rhizosphere of E. crassipes cultured in NH₄Cl solution were more than two orders of magnitude higher than those cultured from plants in distilled water or KNO₃ solution. DGGE analysis also confirmed that NH₄Cl can promote the growth of actinobacteria such as Streptomyces. We are aware that NH₄Cl can reduce the pH of solutions. Molds prefer slight acidity and survive better in acidic solutions. In contrast, actinobacteria prefer slight alkalinity or neutrality (^{Su et al., 2013}), so the mechanism whereby NH₄Cl improved the growth of actinobacteria was not pH-dependent.

Another reason why different forms of nitrogen influenced the rhizosphere microbial community is that the form of nitrogen present can directly affect the growth of the community. Because of various preferences for the type of nitrogen, the presence of different forms of nitrogen alters the components of the rhizosphere microbial community. The experiments of ^{Zhu, Yao, Long, and Yang (2004}) indicated that ammonium ion could inhibit the hyphal growth of AM fungus. ^{Liu, Huang, and Kong (2005}) also confirmed that Aspergillus niger ML2 and ML4 exhibited high efficiency of phosphate solubilisation when cultured in a medium with ammonium as the only nitrogen source.

The final reason why different forms of nitrogen can influence the rhizosphere microbial community is that different forms of nitrogen can differently influence the growth and development of plants, particularly roots (^{Gao & Pan, 1998}; ^{Ma, Wang, Wang, & Wang, 2004}). The rhizosphere of plants provides a microenvironment for the growth of rhizosphere microbes. Respiration and other physiological processes of plant roots can affect this microenvironment. Rhizosphere microbiology is associated with the secretion and abscission of plant roots (^{Liljeroth, Bååth, Mathiasson, & Lundborg, 1990}), and different forms of nitrogen can influence these processes, thereby influencing the microbiology of the rhizosphere. Amino acids secreted from the roots of E. crassipes induced chemotaxis in its rhizosphere bacteria (^{Zhao & Zheng, 1996}). The root exudates from E. crassipes can stimulate the growth of phenol-degrading Enterobacter sp. nov (^{Wu & Zheng, 1993}). In our experiment, ammonium was found to stimulate the growth of actinobacteria; however, whether any secretion or abscission from the roots of E. crassipes can stimulate the growth of actinobacteria remains to be investigated.