Biocontrol of verticillium wilt and colonization of cotton plants by an endophytic bacterial isolate

Authors

  • C.-H. Li,

    1. College of Agronomy, Nanjing Agricultural University, Jiangsu, China
    2. Nanjing Sub-Center of National Rapeseed Development Center, Institute of Industrial Crops, Jiangsu Academy of Agricultural Sciences, Jiangsu, China
    Search for more papers by this author
  • L. Shi,

    1. College of Life Sciences, Key Laboratory of Microbiological Engineering of the Agricultural Environment, Ministry of Agriculture, Nanjing Agricultural University, Jiangsu, China
    Search for more papers by this author
  • Q. Han,

    1. College of Agronomy, Nanjing Agricultural University, Jiangsu, China
    Search for more papers by this author
  • H.-L. Hu,

    1. College of Agronomy, Nanjing Agricultural University, Jiangsu, China
    Search for more papers by this author
  • M.-W. Zhao,

    Corresponding author
    1. College of Life Sciences, Key Laboratory of Microbiological Engineering of the Agricultural Environment, Ministry of Agriculture, Nanjing Agricultural University, Jiangsu, China
    • College of Agronomy, Nanjing Agricultural University, Jiangsu, China
    Search for more papers by this author
  • C.-M. Tang,

    Corresponding author
    • College of Agronomy, Nanjing Agricultural University, Jiangsu, China
    Search for more papers by this author
  • S.-P. Li

    1. College of Life Sciences, Key Laboratory of Microbiological Engineering of the Agricultural Environment, Ministry of Agriculture, Nanjing Agricultural University, Jiangsu, China
    Search for more papers by this author

Correspondence

Ming-Wen Zhao, College of Life Sciences, Nanjing Agricultural University, Key Laboratory of Microbiological Engineering of the Agricultural Environment, Ministry of Agriculture, Nanjing 210095, Jiangsu, China. E-mail: mwzhao@njau.edu.cn

and Can-Ming Tang, College of Agronomy, Nanjing Agricultural University, Nanjing 210095, Jiangsu, PR China. E-mail: cmtang@yahoo.cn

Abstract

Aims

To explore biocontrol potential of 39 DAEB isolates (doubly antagonistic towards both Verticillium dahliae Kleb and Fusarium oxysporum) against verticillium wilt of cotton and to elucidate colonization and category characteristics of an endophytic bacterium with significant biocontrol activity.

Methods and Results

Thirty-nine antagonistic endophytic bacteria strains were tested for their ability to control verticillium wilt in cotton plants caused by a defoliating pathotype of V. dahliae 107 in cotton under controlled conditions. The biocontrol trial revealed that an endophytic bacterium, designated HA02, showed a significant biocontrol activity to V. dahliae 107. After cotton seedlings were inoculated with a gfp gene-tagged HA02 (HA02-gfp), HA02-gfp populations were higher in the root than in the stem; in addition, the HA02-gfp was distributed in the maturation zone of cotton root. Furthermore, HA02-gfp also colonized seedlings of maize, rape and soybean after the bacteria inoculation. Phylogenetic trees based on 16S rDNA sequences combined with morphological, physiological and identification showed that the bacterium belongs to the Enterobacter genus.

Conclusions

Our results showed that only 1 of 39 DAEB isolates demonstrated more efficient biocontrol potential towards V. dahliae 107 in greenhouse and field trials. HA02-gfp mainly colonized cotton in roots. In addition, we quantitatively observed HA02 colonization in other hosts. HA02 belongs to the Enterobacter genus.

Significance and Impact of the Study

This is the first study on biocontrol to defoliating pathotype of V. dahliae Kleb by endophytic bacteria. The HA02 showed effective biocontrol to V. dahliae 107 in greenhouse and field trials. Furthermore, we assessed the quantitative and qualitative colonization of HA02 in cotton seedlings. Our study provides basic information to further explore managing strategies to control this critical disease.

Introduction

Verticillium wilt was first reported in 1914 in Virginia, USA (Carpenter 1914). Since then, this disease was found not only throughout all cotton-producing regions of the USA but also through other cotton-producing countries in the world. Verticillium wilt is caused by a soil-inhabiting fungus, Verticillium dahliae Kleb, and in cotton, it is now considered to be the most important disease. This pathogen can be divided into two types: defoliating and nondefoliating, according to their virulence (Chang et al. 2008). While the nondefoliating pathotype is widespread and only causes mild wilt and no defoliation (Pérez-Artés et al. 2000), the defoliating pathotype develops earlier, faster and induces severe yield and quality losses compared with the nondefoliating one. Verticillium wilt caused by the defoliating pathotype has spread to a number of countries, including the USA, China, Spain, Turkey, Israel, mid-Asia countries, etc. (Li and Yang 2007; Korolev et al. 2008). Moreover, the disease has progressively increased in many regions since 1990 and has become a serious obstacle for cotton production in China. Selection of resistant cultivars is considered the most effective and economical method of disease control. However, because of the lack of immune or highly resistant upland cotton germplasm against the defoliating pathotype, little progress has been made towards this potential solution (Chang et al. 2008). Xiao et al. (1998) found that crop rotation could be a successful practice for managing cotton verticillium wilt, but it was not widely adopted in practice. In addition, no fungicides are currently registered for controlling this disease in cotton (Göre et al. 2009). Therefore, it is necessary to develop alternative methods to manage this disease. The use of biological control agents has been increasing worldwide and is a promising alternative for controlling soil-borne diseases in sustainable and organic agriculture. In the past, rhizosphere bacteria, such as Pseudomonas spp. and Serratia plymutica, have been shown to be effective antagonists towards verticillium wilt (Mercado-Blanco et al., 2004; López-Escudero and Mercado-Blanco, 2011; Erdogan and Benlioglu 2010). More recent studies have indicated that endophytic bacteria colonize the internal tissues of plants can even improve plant growth and plant health (Schulz et al. 2006; Tiwari et al., 2010; Jalgaonwala et al. 2011). Internal plant tissues provide a protective environment for endophytic bacteria, which colonize an ecological niche similar to that of phytopathogens. Therefore, endophytic bacteria are suitable as biocontrol agents to control plant pathogens (Hallmann et al. 1997; Berg and Hallmann 2006). In some cases, endophytic bacteria can significantly improve seed germination and plant growth under adverse conditions (Berg and Hallmann 2006). In addition, endophytic bacteria constitute an environmentally sound alternative to protect plants against the attack of fungal pathogens (Bloemberg and Lugtenberg 2001; Verma et al. 2004). Recently, we systematically studied endophytic bacteria against important fungal pathogens, V. dahliae Kleb and Fusarium oxysporum, in cotton. We found that a considerable population of antagonistic endophytic bacteria is present in cotton roots and that populations fluctuate depending on pathogen genotypes, cotton genotypes and growth stages. The species of antagonistic endophytic bacteria isolates towards both V. dahliae Kleb and F. oxysporum pathogens are diverse, and some have growth-promoting potential (Li et al. 2010). However, these antagonistic endophytic bacterial as biocontrol agents still need to be identified by further tests.

Here, we explore biocontrol potential of the antagonistic endophytic bacteria that inhibit the defoliating pathotype with greenhouse and field trials. Similarly, we investigated the colonization of HA02 with more efficient biocontrol potential in cotton and other hosts in vivo to understand the interaction between HA02 and cotton.

Materials and methods

Bacterial strains, pathogens and cotton cultivar

Thirty-nine endophytic bacteria isolates, doubly antagonistic towards both V. dahliae Kleb and F. oxysporum (DAEB), were isolated by Li et al. (2010) from cotton roots (Table 1). One V. dahliae Kleb isolate, V. dahliae 107, which is a highly virulent defoliating pathotype, was provided by the Institute of Plant Protection of the Jiangsu Academy of Agriculture Sciences. The MINI (Ptet-gfp) plasmid, which contains one copy of constitutively expressed gfp and chloromycetin resistance genes (downstream of a constitutive promoter, Ptet) in tandem, was provided by professor Jun Zhu (Nanjing Agricultural University, China).

Table 1. Sources and DAEB isolate strains in this study
Cotton cultivarCode
  1. DAEB, endophytic bacteria isolates, doubly antagonistic towards Verticillium dahliae Kleb and Fusarium oxysporum, which were isolated by Li et al. (2010) from cotton roots.

Haidao20 (Gossypium barbadense L.)HA01, HA02, HA03, HA04, HA05
Haidao16 (G. barbadense L.)HB01, HB02, HB03, HB04
Cao7005 (G. herbaceum L.)C01, C02,. C03, C04, C05,. C06, C07, C08, C09, C10, C11, C12
Sumian16 (Gossypium hirsutum L.)S01, S01
Changkan (G. hirsutum L.)CK01, CK02, CK03, CK04, CK05, CK06, CK07, CK08, CK09, CK10, CK11, CK12, CK13, CK14
Ya7113 (Gossypium arboreum L.)Y01, Y02

A cotton cultivar, sumian9 (Gossypium hirsutum L.), highly sensitive to verticillium wilt was provided by the Xingyang Experimental Station of Agriculture (Jiangsu, China). Similarly, a wheat cultivar (Triticum aestivum L., Yangmai 158), a corn cultivar (Zea mays L., Suyu24), a rice cultivar (Oryza sativa L., Gudao6), a rape cultivar (Brassica napus L., Huanza1) and a soybean cultivar (Glycine max L., Tongdo5) were provided by Mintian seed Ltd (Jiansu, China).

Colonies of the V. dahliae 107 were subcultured from glycerol stocks on potato dextrose agar plates and incubated at 25°C for 10 days. Spores from the potato dextrose agar plate were inoculated into a czapek medium containing 2 g NaNO3, 1 g K2HPO4, 1 gMgSO4·7H2O, 1 g KCl, 2 mg FeSO4·7H2O and 30 g l−1 sucrose. Spores were incubated for 5–8 days until the concentration reached approximately 1·0 × 107 spores ml−1. To inoculate the cotton plants with the pathogen, an aqueous suspension of spores was filtered through several layers of cloth and adjusted to a concentration of 1·0 × 106 spores ml−1 with sterile distilled water (Qu et al. 2005). Likewise, V107 spores were grown in potato dextrose medium at 25°C in the dark for 7 days. Subsequently, an aqueous suspension of spores was transferred to glass flasks (loosely closed with cotton wool) containing humid and autoclaved wheat. The flasks were incubated at 25°C for 30 days and shaken by hand for 5 min twice a day. The wheat inoculum of V107 was air-dried before inoculating the soil.

Screening of DAEB strains against verticillium wilt in a greenhouse trial

DAEB isolates were firstly screened for biocontrol potential against verticillium wilt caused by the V. dahliae 107 strain in a greenhouse. Each treated experimental unit contained a DAEB isolate and the V. dahliae 107 treatment, while the control only had the V. dahliae 107 treatment; to simplify the experiment, the DAEB isolates are randomly divided into two groups (one group comprised 19 isolates, while the other had 20 isolates),. The experiment was set with a completely randomized design with four replications per treatment (each pot was a replication).

To remove surface pathogens, cotton seeds were first delinted by sulphuric acid (12 g sulphuric acid/100 g cottonseeds, shaken together for 2–3 min) and then washed with autoclaved water. The surface-sterilized upland cotton seeds were sown in plastic pots filled with sterile vermiculite. These pots were placed in a greenhouse (20°C/28°C, night/day) for 7 days, and the plants were thinned to 20–30 and were irrigated and fertilized every 5 days (Huang et al. 2006). When the seedlings had 2–3 emerged leaves, a challenge inoculation was carried out as described by Chen et al. (1995). As verticillium wilt in cotton plants reached the peak, severity of verticillium wilt was evaluated. A scale from 0 to 4 was used to classify plants according to the percentage of plant tissue affected by chlorosis, leaf necrosis or defoliation (0: healthy plant or plant without symptoms; 1: plant affected by 1–33%; 2 = 34–66%; 3: 67–99%; 4: dead plant) (Huang et al. 2006). The leaf wilt index (LWI) and percentage of protection was calculated as follows:

  • Leaf wilt index = Σ (number of plants with a disease scale × value of the corresponding scale)/(total plants rated × the highest scale value)

  • Protection% = 100–Leaf wilt index (treatment)/Leaf wilt index (control) × 100

Ten strains selected in the preliminary screen were further tested in another greenhouse pot trial. The experiment was arranged in a completely randomized design with three replications. Pot soil was mixed with the V. dahliae 107 wheat inoculum at a concentration of 1% (w/w). Nine 10-l plastic pots per treatment were filled with the mix. Surface-sterilized cotton seeds were dipped 18 h in the DAEB suspension with approximately 108 CFU ml−1 (OD595 = 0·3) according to Kavino et al. (2007); likewise, nonbacterized seeds were used as controls. Cotton seeds were sown in each pot, which were placed in a greenhouse (20°C/28°C, night/day), and the plants were thinned to 14–30 after seedling, irrigated and fertilized in the same way as the challenge inoculation described above. When the first leaf of seedlings emerged, 10 ml DAEB suspension with approximately 109 CFU ml−1 was drenched to every plant. Plant roots below the soil line from 2 to 5 cm were slightly punctured with 7-cm needle after 2–3 leaves emerged. Finally, the degree of leaf wilt was evaluated as the challenge inoculation described above.

Biocontrol efficacy of DAEB strain HA02 and CK06 in a field trial

The strains HA02 and CK06 from the pot trial were selected for a field trial. The field experiment was conducted at the Longpao site of cotton seed production, Nanjing, China (32°03′N, 118°46′E). The soil has a heavy loam texture (constituent of soil particles: 28%, <0·002 mm; 41%, 0·002–0·02 mm; 31%, 0·02–2 mm. pH 6·82; organic matter, 12·33 g kg−1; total N, 1·15 g kg−1; available P, 0·34 g kg−1; and available K, 4·37 cmol kg−1). The experiment was set as a randomized block design with three replicates per treatment, and each replication was 25 m long, and four-row plots, with a row spacing of 80 cm and a plant density of 4·5 plants per m2. The field had been continually planted with cotton for 2 years without verticillium wilt. The V. dahliae 107 wheat inoculum was distributed in the field with 50 g per row.

Surface-sterilized cotton seeds were sown in bowls with nutrition soil (Fan et al. 2007) on 3 April 2011; all the seedlings were thinned to one plant in every bowl. Seedlings with four leaves were transplanted to the experimental plots on 18 May. The treatments were carried out every 15 days from 18 May to 28 August and were consistently exposed to the following two procedures: A, each plant was drenched with 20 ml DAEB suspension (109 CFU ml−1); and B, each plant was drenched with 20 ml DAEB suspension (109 CFU ml−1) and inoculated by a stem puncture with a sterile syringe about 2 cm above the soil line with two separate droplets of 20 μl DAEB suspension (109 CFU ml−1), and control plants were drenched with water. Cotton cultivation was managed according to local agronomic practices. When verticillium wilt in cotton plants reached the peak, the LWI and percentage of protection were estimated at flowering stage. Similarly, we surveyed relative traits of cotton yield, such as height, branch number, boll number, boll weight and lint percentage at harvest stage. The experiment condition was strictly controlled in both greenhouse and field trials above. In addition, disease severity of verticillium wilt was assessed using a stem cut rating system at the end of the harvest stage (Dong et al. 2006; Huang et al. 2006). Briefly, each plant was scored for vascular discoloration rated on a 0–4 scale: 0: no discoloration; 1: one-fourth of the cross-section showing discoloration; 2: one-half of the cross-section showing discoloration; 3: three-fourth of the cross-section showing discoloration; 4: full cross-section showing discoloration. A vascular discoloration index was expressed as a

  • Vascular discoloration index = Σ (number of plants with a disease scale × value of the corresponding scale)/(total plants rated × the highest scale value)

  • Protection% = 100 − Vascular discoloration index (treatment)/Vascular discoloration index (control) × 100

Construction of gfp-labelled HA02

MINI (Ptet-gfp) plasmids were introduced into HA02 by electroporation (Elbeltagy et al., 2001).; Newman et al. 2003). Then, HA02 bacteria were cultured in 10 ml of nutrient broth medium (Difco). Samples (1 ml) of subculture were transferred to 2000-ml flasks containing 400 ml of nutrient medium growing at 30°C shaken at 160 rev min−1 to an OD650 of 0·6. Cells were chilled on ice separately and harvested by centrifugation at 200  g for 10 min at 4°C. Each pellet was washed four times with cold, sterilized distilled water and, finally, re-suspended in cold, sterile 10% glycerol–water. The cell suspension was dispensed in 200 μl aliquots and stored at −70°C for future use. Competent cells (200 μl) were thawed on ice, and 50 ng of plasmid was added and quickly mixed. The mixture was incubated on ice for 10 min and transferred to a sterile, prechilled cuvette (0·2 cm of interelectrode gap). The plasmids were electroporated into the cells by an electroporation system (Bio-Rad, Hercules, CA) set at 2·5 kV cm−1, 25 μF, and 200 Ω. Following the pulse, the cells were immediately diluted with 1 ml of nutrient broth medium (Difco), transferred to the sterilized tubes and incubated at 37°C for 2 h, with shaking at 120 rev min−1, and then, they were plated on a selective medium containing 20 μg ml−1 chloromycetin. Transformants that emitted green fluorescence were screened with confocal laser scanning microscope (with excitation wavelength of 488 nm). The plasmid stability was examined as described by Humberto et al. (2002). The transformant growth curve in nutrient broth culture was determined by the method of Humberto et al. (2002). The biocontrol efficacy of gfp-labelled HA02 against verticillium wilt in greenhouse was conducted as the trial of ‘Screening of DAEB strains against verticillium wilt in a greenhouse trial’ above.

Population dynamics of HA02 in cotton

Surface-sterilized cotton seeds were germinated in a nutrient agar medium for 2 days at 37°C; seeds that did not exhibit fungal or bacterial contamination were selected and dipped for 2 h in the HA02-gfp suspension with approximately 108 CFU ml−1 (OD595 = 0·3) according to Kavino et al. (2007). Then the seeds were transferred to a plastic pot (20 cm × 20 cm) with sterilized soil; each pot contained 30 seedlings and was cultured at 28°C and 12 h of light per day in a growth chamber. Seeds with no inoculation were used as negative controls. Seedlings were sampled every 5 days from the 1st to the 40th day after inoculation and were surface-sterilized in 0·05% sodium hypochlorite for 15–30 min, and then, they were quickly washed with sterilized distilled water. The surface-sterilized stems and roots were excised, weighed and macerated separately in a 0·8% saline solution. Each sample was completely triturated with a sterile mortar and pestle in 9·9 ml of the final buffer wash and allowed to stand at room temperature for 10 min. Then, 100 μl of a 10-fold serial dilution of the suspension in sterile phosphate buffer was plated in triplicate on nutrient agar plates containing 20 μg of chloromycetin per ml. The plates were incubated at 28°C for 48–72 h. Green fluorescent colonies that appeared on the plates after incubation at 30°C were counted with the aid of an LG ps2 fluorescence stereomicroscope (Olympus, Tokyo, Japan). The experiment was set with three replicates.

Colonization of HA02 in cotton roots

Surface-sterilized cotton seeds were germinated in a nutrient agar medium for 2 days at 37°C; those germinations that did not exhibit fungal or bacterial contamination appeared were selected and dipped for 2 h in the HA02-gfp suspension with approximately 108 CFU ml−1. Then they were transferred to sterile flasks (10 cm in diameter and 15 cm in height) with 500 ml sterilized culture medium containing 8 g of agar per litre of tap water (i.e. 0·8% agar; pH 6·5) and cultured at 28°C and 12 h of light per day in a growth chamber. Plants with no inoculation were used as negative controls. After 7, 12 and 17 days of inoculation, the seedling roots were removed and lightly rinsed with sterilized distilled water. These intact seedling roots were subsequently observed, photographed and excited with blue light with a fluorescence stereomicroscope.

Population dynamics of HA02 in other hosts

Seeds of wheat, maize, rice, rape and soybean were surface-sterilized, inoculated and sowed, and endophytic inoculant of seedlings roots and stems was extracted and counted as the trial of ‘Population dynamics of HA02 in cotton’ above.

Identification of HA02

The partial 16S rDNA sequence of HA02 was sequenced elsewhere (Li et al. 2010). Our nucleotide sequence was compared with that of the type strain in EzTaxon server 2.1. A phylogenetic tree was constructed using Mega program Bootstrap analyses (Chun et al. 2007).

The morphological and physiological characteristics of HA02 were assessed according to Bergey's Manual of Systematic Bacteriology (Breed and Gibbons 1984) and Bergey's Manual of Determinative Bacteriology (Holt et al. 1994).

Data analysis

The data were statistically analyzed with the help of the computer software package spss ver. 15. The colonization number of HA02-gfp was converted to log10 CFU g−1 of fresh weight (CFU g fw−1) prior to analysis of variance (anova), and Student–Newman–Keuls test was used for testing the mean difference among the treatments. P < 0·05 was considered as significant.

Results

Screening potential biocontrol agents in the greenhouse trial

Verticillium wilt in cotton plants reached its peak after 15 days from the challenging inoculation with V107. Similarly, the LWI of plants bacterized with the 39 DAEB strains ranged from 15·9 to 66·9 in group 1 and from 17·3 to 67·5 in group 2, while the LWI of the control was 74·2 in group 1 and 77·3 in group 2 (Table 2). Thus, all the DAEB strains had biocontrol potential against verticillium wilt to a certain extent.

Table 2. Ability of 39 DAEB strains to control verticillium wilt was evaluated in the challenge inoculation trial in the greenhousea
Group1Group2
StrainLWIProtection%StrainLWIProtection%
  1. DAEB, endophytic bacteria isolates, doubly antagonistic towards Verticillium dahliae Kleb and Fusarium oxysporum.

  2. Values are the mean ± SE of four replicates that followed by the same letter are not significantly different at the 0·05 probability level according to Student–Newman–Keuls tests.

  3. a

    The seedlings had 2–3 leaves; a challenge inoculation was carried out as described by Chen et al. (1995). When verticillium wilt in cotton plants reached the peak, the leaf wilt index (LWI) and percentage of protection was estimated.

HA0515·9 ± 1·0h78·6 ± 1·3CK0617·3 ± 1·1h77·6 ± 1·4
HA0216·4 ± 1·9h77·9 ± 2·5CK1325·6 ± 1·5g66·9 ± 2·0
C0828·9 ± 1·2g61·1 ± 1·6HB0228·9 ± 1·5fg62·7 ± 1·9
HA0130·1 ± 1·0g59·4 ± 1·3HB0431·0 ± 1·0f59·9 ± 1·3
HA0335·2 ± 2·3f52·5 ± 3·0Y0236·3 ± 2·0e53·1 ± 2·6
C0440·9 ± 2·0ef44·9 ± 2·7CK0337·2 ± 1·1e51·9 ± 1·4
S0141·1 ± 1·8ef44·6 ± 2·4CK0738·0 ± 1·2e50·9 ± 1·6
C1243·2 ± 1·8e41·7 ± 2·5HB0140·0 ± 1·1e48·3 ± 1·4
C0246·2 ± 0·9de37·7 ± 1·2CK1140·2 ± 0·8e48·0 ± 1·0
C0746·6 ± 1·5de37·2 ± 2·0CK1447·3 ± 1·1d38·8 ± 1·5
C0551·1 ± 1·8cd31·1 ± 2·5HB0350·0 ± 0·7d35·3 ± 1·0
C0953·0 ± 1·4c28·6 ± 1·8CK0551·1 ± 0·7d33·9 ± 0·9
S0253·1 ± 1·7cf28·4 ± 2·3CK0256·3 ± 0·9c27·3 ± 1·1
C0353·9 ± 2·5c27·3 ± 3·4CK1256·4 ± 1·4c27·0 ± 1·9
C0654·1 ± 2·6c27·1 ± 3·5Y0162·1 ± 0·8b19·7 ± 1·1
C1161·1 ± 1·2b17·7 ± 1·6CK0464·1 ± 1·8b17·2 ± 2·3
C1061·6 ± 1·3b17·0 ± 1·8CK0864·3 ± 2·2b16·8 ± 2·8
HA0462·6 ± 2·3bc15·6 ± 3·1CK1065·0 ± 1·3b15·9 ± 1·7
C0166·9 ± 1·5b9·8 ± 2·0CK0967·0 ± 1·3b13·4 ± 1·7
Control74·2 ± 1·6a CK0167·5 ± 1·4b12·8 ± 1·9
Control77·3 ± 1·4a

On the basis of the biocontrol efficacy of DAEB isolates in the challenging inoculation trial and seedling vigour index in the germination experiment (Li et al. 2010), the 11 strains with biocontrol potential and without harmful seedling vigour were selected to perform the pot trial (Table 3). Strains HA02, CK06 and HA03 had better biocontrol potential against the disease verticillium wilt at the peak of verticillium, with a protection effect of 72·4, 70·3 and 51·4%, respectively. In contrast, the other strains showed a lower or no biocontrol potential against the disease.

Table 3. Ability of 10 DAEB a strains to control verticillium wilt was evaluated in the pot trial in the greenhouse
StrainLWIProtection%StrainLWIProtection%
  1. DAEB, the endophytic bacteria isolates, doubly antagonistic towards Verticillium dahliae Kleb and Fusarium oxysporum.

  2. a

    Ten DAEB strains were selected from the challenge inoculation trial (Table 2). Pot trial was treated with dipping the DAEB suspension to cotton seeds. When verticillium wilt in cotton plants reached the peak, the leaf wilt index (LWI) and percentage of protection were estimated. Values are the mean ± SE of three replicates that followed by the same letter are not significantly different at the 0·05 probability level according to Student–Newman–Keuls tests.

HA0220·0 ± 2·6d72·4 ± 3·6CK0369·7 ± 1·0a3·9 ± 1·3
CK0621·5 ± 1·6d70·3 ± 2·2CK1373·0 ± 2·8a−0·7 ± 3·8
HA0335·3 ± 2·3c51·4 ± 3·2Y0276·3 ± 1·9a−5·2 ± 2·6
CK0754·1 ± 1·9b25·4 ± 2·6HB0478·0 ± 2·2a −7·6 ± 3·0
HA0160·1 ± 4·9b17·1 ± 6·7Control72·5 ± 2·0a
HB0267·9 ± 0·6a6·3 ± 0·8

Biocontrol efficacy of DAEB strains HA02 and CK06 in the field trial

The strains HA02 and CK06 screened in the greenhouse trial were selected for the field trial (Table 4). The LWI was 33·3 with HA02 drenching treatment and 31·8 with HA02 drenching plus puncturing treatment in the 14th week after the transplant of seedlings, which provided a protection effect of 45·8 and 48·3%, respectively. The vascular discoloration indices were 66·7 and 68·0 by the two treatments at the end of the harvest, which resulted in a protection effect of 16·3 and 14·8%, respectively. However, CK06 provided little protection of cotton plants against verticillium wilt in the two treatments (Table 4). At the end of the harvest, both treatments with HA02 significantly increased traits related to yield (Table 5), such as plant height, numbers of branch, numbers of boll and weight of boll. As a result, cotton yields were increased in both treatments. In contrast, these two treatments with CK06 showed lower efficacy.

Table 4. Ability of HA02 and CK06 strains to control verticillium wilt was evaluated in the field trial
Treatmenta LWIProtection%VDIProtection%
  1. Values are the mean ± SE of three replicates that followed by the same letter are not significantly different at the 0·05 probability level according to Student–Newman–Keuls tests.

  2. a

    CK06 and HA02 strains were selected from the pot trial (Table 3). Field trial was treated by drenching (Dr) and drenching plus puncturing(Dr + P) with the bacterial suspension to the cotton plant. When verticillium wilt in cotton plants reached the peak (in the 14th week after the transplant of seedlings), the leaf wilt index (LWI) and percentage of protection was estimated. In addition, the vascular discoloration index (VDI) was assessed at the end of the harvest stage.

CK06, Dr57·1 ± 2·0a7·2 ± 2·680·3 ± 1·2a−0·7 ± 1·2
CK06, Dr + P56·0 ± 2·3a8·9 ± 3·078·2 ± 2·5a2·0 ± 2·6
HA02, Dr33·3 ± 1·4b45·9 ± 1·966·7 ± 1·4b16·3 ± 1·5
HA02, Dr + P31·8 ± 1·2b48·3 ± 1·668·0 ± 0·4b14·8 ± 0·4
Control61·5 ± 2·1a 79·7 ± 1·3a
Table 5. Influence of HA02 and CK06 strains to plant characteristics related to yield was evaluated in the field trial
Treatmenta Height (cm)Number of branchesNumber of bollsBoll weight (g)Lint%Cotton yield
(kg per 75 m2)CK±%
  1. Values are the mean ± SE of three replicates that followed by the same letter are not significantly different at the 0·05, probability level according to Student–Newman–Keuls tests.

  2. a

    CK06 and HA02 strains were selected from the pot trial (Table 3). Field trial was treated by drenching (Dr) and drenching plus puncturing (Dr + P) with the bacterial suspension to the cotton plant. We surveyed the plant characteristics related to yield, such as height, branch number, boll number, boll weight and lint percentage at harvest stage.

CK06, Dr103·0 ± 1·8b15·5 ± 0·3b16·0 ± 0·2b4·7 ± 0·1b39·7 ± 0·5a16·0 ± 0·5b3·2
CK06, Dr + P100·7 ± 2·0b15·7 ± 0·1b15·7 ± 0·3b4·8 ± 0·1b39·4 ± 0·4a15·8 ± 0·2b1·9
HA02, Dr111·1 ± 1·6a16·4 ± 0·2a19·5 ± 0·2a5·4 ± 0·0a39·2 ± 0·6a21·4 ± 0·6a38·1
HA02, Dr + P110·8 ± 2·1a16·4 ± 0·2a19·6 ± 0·2a5·4 ± 0·1a39·0 ± 0·3a20·9 ± 0·7a34·8
Control101·6 ± 1·7b15·6 ± 0·3b15·6 ± 0·4b4·7 ± 0·2b39·4 ± 0·6a15·5 ± 0·4b

Qualitative and quantitative colonization of HA02 in cotton

We selected a gfp-labelled HA02 by electroporation, designated as HA02-gfp, which showed a similar growth curve in nutrient broth medium and biocontrol efficacy against verticillium wilt as the parental HA02 strain. After 10 continuous culture cycles in the absence of antibiotic pressure, 98% of HA02-gfp was chloromycetin resistant and showed green fluorescence. Therefore, we evaluated the colonization of HA02-gfp instead of the parental HA02 colonization (Fig. 1). The population tended to increase from the 1st to the 10th day, and then, it tended to decrease; subsequently, it showed a slight increase at the 30th day. When the seedlings were 40 days old, they still had log10 2·74 CFU g fw−1. The HA02-gfp populations in stem had the same trend as those in root, but were significantly lower than in roots.

Figure 1.

After cotton seedlings were inoculated with the gfp-tagged bacteria (HA02-gfp), the fluctuation of HA02-gfp populations in root were the same as that in stem, but there were more the bacterial populations in root than stem. Populations (log10 CFU g fw−1) of HA02-gfp in roots (▲) and stems (■). The error bars are SE.

Because HA02-gfp mainly colonized cotton roots, we qualitatively studied colonization in this organ (Fig. 2). HA02-gfp was distributed in the maturation zone of primary roots and lateral roots of 7-day-old seedlings observed with a stereoscopic fluorescence microscope. However, HA02-gfp could not be found in root tips and elongation zone. At 12–17 days after inoculation, HA02-gfp was not discovered in the whole root.

Figure 2.

Observation of HA02-gfp colonization under stereoscopic fluorescence microscopy. A, HA02-gfp showed strong green fluorescence distributed in maturation zone of primary root and lateral roots after 7 days of seedling treatment; CK, the control showed no green fluorescence (×12·6).

Quantitative colonization of HA02-gfp in other hosts

HA02-gfp could be detected in 1- to 30-day-old seedlings of all the hosts except rice after the host seeds were treated with the HA02-gfp suspension (Table 6). The HA02-gfp populations were different for the factors of the cultivars and tissues.

Table 6. Root and stem colonization number (log10 CFU g fw−1) of the gfp-tagged bacteria (HA02-gfp) in maize, wheat, rice, rape and soybean seedlings were evaluated after these host seedlings were inoculated with HA02-gfp
Species (cultivar)TissueSeedling (day)
15102030
  1. Values are the mean ± SE of three replicates.

Maize (Suyu24)Root4·49 ± 0·124·76 ± 0·086·3 ± 0·165·04 ± 0·075·90 ± 0·20
Stem3·55 ± 0·093·74 ± 0·094·00 ± 0·054·30 ± 0·132·90 ± 0·01
Wheat (Yangmai 158)Root5·50 ± 0·145·57 ± 0·144·22 ± 0·033·53 ± 0·080·00 ± 0·00
Stem4·95 ± 0·054·08 ± 0·043·00 ± 0·040·00 ± 0·000·00 ± 0·00
Rice (Gudao6)Root0·00 ± 0·000·00 ± 0·000·00 ± 0·000·00 ± 0·000·00 ± 0·00
Stem0·00 ± 0·000·00 ± 0·000·00 ± 0·000·00 ± 0·000·00 ± 0·00
Rape (Huanza1)Root4·37 ± 0·053·80 ± 0·033·00 ± 0·020·00 ± 0·000·00 ± 0·00
Stem3·53 ± 0·044·63 ± 0·062·48 ± 0·010·00 ± 0·000·00 ± 0·00
Soybean (Tongdo5)Root2·60 ± 0·034·26 ± 0·073·10 ± 0·052·48 ± 0·020·00 ± 0·00
Stem4·08 ± 0·074·29 ± 0·103·27 ± 0·012·78 ± 0·022·65 ± 0·03

Classification of HA02

Partial 16S rDNA sequences of HA02 had 1306 bp, and the sequences were deposited in the GenBank under accession no. FJ205654 and shared 99·9% homology with Enterobacter cancerogenus LMG 2693 (z96078). A phylogenetic tree was constructed with partial 16S rDNA sequences of HA02 and closely related strains using MEGA software. HA02, Ent. cancerogenus LMG 2693 (z96078) and Enterobacter asburiae JCM 6051T (AB004744) sequences form a branch in the phylogenetic tree (Fig. 3).

Figure 3.

Phylogenetic tree based on 16S rDNA sequence homology of HA02. Numbers in parentheses represent the sequence accession number in GenBank. Numbers in each branch point denote the percentages of support by bootstrap (1000). Bar: 0·1% of sequence divergence.

On lysogeny broth medium, HA02 had circular colonies with a creamy colour, and its cells were rod shaped and measured approximately 0·45 μm by 1·50 μm. Vegetative cells were straight rods and were peritrichous with 3–6 flagella; they did not have endospores and measured 1·50 μm in length and 0·45 μm in width; they were gram-negative. Physiochemical traits showed negative oxidase reaction, methyl red reaction, indole production and H2S production, while there was a positive VP reaction, gelatine reaction, lactose assimilation, utilization of citrate, malonate and acetate as sole carbon source. The test of glucose fermentation showed acid and gas production at 30°C; however, it showed no gas production at 44·5°C. These results suggest that morphological and physiological identification of HA02 are in agreement with Enterobacter sp. characteristics (Table 7). Therefore, based on these results, we believe that HA02 belongs to Enterobacter sp.

Table 7. Morphological and physiochemical traits of HA02
TestTraitsTestTraits
  1. ‘+’ and ‘−’ represent ‘Positive’ and ‘Negative’, respectively.

Colony colourIvory whiteH2S production
Cell shapeRod, peritrichous flagella (3–6)Lactose assimilation+
Vegetative cell size (μm)1·50 × 0·45As sole carbon source
Gram stainCitrate+
EndosporesMalonate+
Motility+Acetate+
Oxidase reaction30°C glucose ferment
Methyl red reactionAcid production+
VP reaction+Gas production+
Gelatine reaction+44·5°C Glucose ferment
Indole productionGas production

Discussion

To our knowledge, this is the first study on biocontrol to defoliating pathotype of verticillium wilt by endophytic bacteria. For screening endophytic bacteria as biocontrol agents against V107, trials in greenhouse and field were carried out. To explore interaction between HA02 and cotton, we assessed the quantitative and qualitative colonization of HA02 in cotton seedlings.

The field trial demonstrated HA02 had better biocontrol efficacy, while CK06 lacked biocontrol efficacy against V. dahliae 107. The different biocontrol efficacy between HA02 and CK06 could be due to their adaptability and competition in different environments (Berg and Hallmann 2006), but this yet to be researched further. The internal population size of endophytic bacteria with biocontrol potential is one of the most important factors for controlling plant pathogens (Press et al. 2001; Kloepper and Ryu 2006). Therefore, to obtain high internal inoculum population in hosts, we designed an inoculating treatment by drenching plus puncturing cotton-plant roots in the field trial. However, there were no significant differences in biocontrol efficacy between drenching and drenching plus puncturing treatment. We presumed that, under field conditions, roots are susceptible to be wounded by wind, rain, pathogens, pests, farming practices and other abiotic and biotic factors. Wounds and lateral roots are the main entry for endophytic bacteria into plant tissues (Hurek et al. 1994; Liu et al. 2006), so the inoculum could effortlessly entered into plant tissues and reach its quantity equilibrium by drenching or drenching plus puncturing methods. Drenching the cotton plant with HA02 was an efficient inoculation method to suppress verticillium wilt.

The populations of HA02-gfp in roots were about 100 times higher than those in stems, and the different colonization populations of antagonistic bacteria between organ have also been reported by Nga et al. (2010), who discovered that a Pseudomonas aeruginosa 231-1 systemically protect watermelon against gummy stem blight, mainly colonized in the hypocotyls of watermelon but not found in the true leaf after seed soaking and subsequently soil drenching to 7-day-old seedlings. Seven days after HA02-gfp inoculation, HA02-gfp was mainly distributed in the maturation zone of cotton roots, which had similar root-colonizing patterns of Pseudomonas fluorescens A6RI in tomato (Gamalero et al. 2004), Azospirillum brasilense in rice (Ramos et al. 2002) and Bacillus megaterium C4 in maize (Liu et al. 2006). It is possible that high concentration exudates in the maturation zone caused chemotaxis (Gamalero et al. 2004) and that lateral roots in the maturation zone were the main entry of endophytic bacteria (Hallmann et al. 1997). Twelve to 17 days after HA02-gfp inoculation, HA02-gfp could not be observed under the stereoscopic fluorescence microscope. One possible explanation is that roots became opaque with lignin. Alternatively, HA02-gfp may have colonized the inner layers of cotton roots.

Currently, it is believed that the mechanism of virulence of V. dahliae Kleb in cotton may be the blocking of the vascular bundle by mycelium and spores of V. dahliae Kleb and calluses produced by cotton plants. Alternatively, the symptoms may be the result of wilt toxins secreted by V. dahliae Kleb (Fradin and Thomma 2006). Antagonistic mechanisms of endophytic bacteria towards fungal pathogens are mainly related to antibiosis, competition, lysis, induction of plant defenses and plant growth (Berg and Hallmann 2006). HA02 could excrete siderophores and showed protease activity (Li et al. 2010). Therefore, biocontrol mechanisms of HA02 towards V. dahliae 107 may be due to that HA02 competed against V107 by excreted siderophores, which inhibited V. dahliae 107 invading and spreading (O'Sullivan and O'Gara 1992); or HA02 produce lysis proteases of wilt toxin or antibiosis, which reduced the wilt extent. Press et al. (2001) reported that Serratia marcescens strain 90–166 colonized roots internally and has been shown to elicit induced systemic resistance against various diseases of cucumber. Owing to HA02-colonizing cotton root with quite numbers, the biocontrol mechanisms of HA02 towards V. dahliae 107 may be due to that HA02 induced systemic plant resistance against the pathogen (Kloepper and Ryu 2006).

Acknowledgements

This work was supported by a grant from Natural Science Foundation of China (no. 31071459).

Ancillary

Advertisement