OUP user menu

Isolation and characterization of endophytic streptomycete antagonists of fusarium wilt pathogen from surface-sterilized banana roots

Lixiang Cao , Zhiqi Qiu , Jianlan You , Hongming Tan , Shining Zhou
DOI: http://dx.doi.org/10.1016/j.femsle.2005.05.006 147-152 First published online: 1 June 2005


A total of 131 endophytic actinomycete strains were successfully isolated from surface-sterilized banana roots. These isolates belonged to Streptomyces (n= 99), Streptoverticillium (n= 28), and Streptosporangium (n= 2) spp. The remaining 2 isolates were not identified. About 18.3% of the isolates inhibited the growth of pathogenic Fusarium oxysporum f. sp. cubense on banana tissue extract medium. The most frequently isolated Streptomyces sp. strain S96 was similar to Streptomyces griseorubiginosus. About 37.5% of the S. griseorubiginosus strains were antagonistic to F. oxysporum f. sp. cubense. The antagonism of strain S96 was lost when FeCl3 was introduced into the inhibition zone. In vivo biocontrol assays showed that the disease severity index (DSI) was significantly (P= 0.05) reduced and mean fresh weight increased (P= 0.001) in plantlets treated with strain S96 compared to those grown in the absence of the biocontrol strain. These findings indicate the potential of developing siderophore-producing Streptomyces endophytes for the biological control of fusarium wilt disease of banana.

  • Banana
  • Endophytic Streptomyces
  • Fusarium oxysporum f. sp. cubense
  • Fusarium wilt
  • Siderophores

1 Introduction

Ecologically, actinobacteria and, particularly, the Streptomyces spp. are generally saprophytic, soil-dwelling microorganisms that spend the majority of their life cycle as spores. It has also been demonstrated that actinomycetes are important microorganisms in the rhizosphere and their antagonism to phytopathogenic fungi has been demonstrated [[]. Some actinobacteria are also known to inhabit the tissues of healthy plants. By cultivation-independent techniques based on 16S rDNA genes, the actinomycetes were found to colonize inside the roots of barley [[] and stems of potato [[]. The actinomycetes that resided in healthy plant tissues without causing symptoms of disease were defined as endophytic actinomycetes. The first endophytic actinomycetes studied were Frankia strains isolated from non-leguminous plants [[]. Other endophytic actinomycetes such as Streptomyces, Streptoverticillium, Nocardia, Micromonospora, Microbispora, and Streptosporangium strains were isolated from surface-sterilized roots of different plant species in Italy [[] and of maize in Brazil [[]. The use of endophytic actinomycetes as biological control agents of soil-borne root disease is of interest through their ability to colonize healthy plant tissue and produce antibiotics in situ [[]. Indeed, endophytic actinomycetes isolated from wheat roots have shown the ability to reduce the impact of “take-all” disease on wheat by up to 70% in glasshouse trials using naturally infested soils (http://www.abc.net.au/science/news).

Banana is one of the main fruits cultivated in subtropical and tropical regions. Panama disease, which is also known as fusarium wilt, is regarded as one of the most destructive diseases of banana production in those regions [[]. The pathogenic fungus (Fusarium oxysporum f. sp. cubense, Foc) enters through the roots and blocks the vascular system causing the plant to wilt, followed by death of the whole plant [[]. Several existing disease management strategies such as crop rotation with rice, and injection of rhizomes with 2% carbendazim, are tedious. A cost-effective measure of control for the disease is still not available. A complementary approach for managing fusarium wilt is biological control and the search for antagonistic microorganisms has identified several antagonistic fungi and bacteria with high activity [[,[0]. But the banana plant is a giant perennial herb, so the introduction of suppressive soils and antagonistic bacteria around the rhizosphere will be time- and labor-consuming. The introduction of endophytic Streptomyces strains into plantlets at an early stage of growth is facile in application for banana plants derived from tissue culture planting material have fewer microorganisms and endophytes could grow in the banana plant tissue. However, reports of endophytic streptomycete antagonists of the fusarium wilt pathogen of banana are rare.

Previously, during a search for novel biological control agents against fusarium wilt, endophytic Streptomyces sp. strain S76, assigned to the Streptomyces griseorubiginosus clade, was isolated from healthy banana plants (soil pH 7.3) collected from Panyu Town [[1]. In this study, strain S96, assigned to the S. griseorubiginosus clade, was also isolated from healthy banana roots (soil pH 5.1) and its effects on Foc race 4 and banana plantlets were evaluated in vitro and in vivo.

2 Materials and methods

2.1 Media and strains

S medium [[2] was used for actinobacterial isolation. This was prepared by dissolving 10 g dextrose, 4.0 g casein hydrolysate, 0.5 g K2HPO4, 0.2 g MgSO4· 7H2O, 0.1 g CaCl2· 2H2O, 10 mg ferric citrate, 0.01 mg CoSO4· 7H2O, 0.1 mg CuSO4· 5H2O, 1.5 mg H3BO3, 0.8 mg MnSO4· H2O, 0.2 mg (NH4)6Mo7O24· 4H2O, 0.6 mg ZnSO4· 7H2O in 1000 ml deionized water. Banana Tissue Extract medium (BTE medium) was used for the antibiosis assay; it was prepared by chopping 1 kg banana pseudostems into small pieces and submerging them in 1000 ml boiling tap water for 1 h, then 15 g agar was added to filtrates before autoclaving at 121°C for 20 min.

F. oxysporum f. sp. cubense (Foc) race 4 was isolated from wilting rhizomes of banana under field conditions in Guangzhou, China. The pathogen was cultured on potato dextrose agar (PDA) and incubated at 28°C under white fluorescent light for 7 days. The aerial cultures of Foc were suspended in sterile tap water and filtered through double-layered cheesecloth. The inoculum concentration (106 spores ml−1) was prepared with the aid of a hemocytometer.

2.2 Sample collection

Healthy banana plants (Musa sp., AAA, Cavendish subgroup, cv. Williams) were collected from banana plantations (soil pH 5.1) located in the suburbs of Guangzhou, China. There were no Panama disease symptoms occurring historically in the plantations. A total of five banana plantlets were dug out carefully to ensure that maximal amounts of root material were collected, the samples were placed in plastic bags and taken to the laboratory and processed within 4 h of collection.

2.3 Isolation of endophytic Streptomyces from banana root interior and their identification

The isolation and identification of endophytic Streptomyces from the banana root interior were performed as described previously [[1,[3].

2.4 Antibiosis assay

The isolates were inoculated at the margin of Petri dishes containing BTE medium. After incubation at 25°C for 3 days, the Foc race 4 was inoculated at the center of the Petri plate. After 4 days further incubation, the distance between the margins of the streptomycete colony and the fungal colony was evaluated for evidence of inhibition.

In order to evaluate antagonism between the endophytic streptomycete and the pathogenic fungus on autoclaved banana pseudostem tissue, the endophytic streptomycete was inoculated on one side of the autoclaved banana tissue (0.5 × 4 cm) and incubated at 26°C for 3 days, then the pathogenic fungus was inoculated on the other side. The distribution of fungal mycelia was observed by light microscopy after staining with lactophenol blue solution.

2.5 Effect of iron on antagonism of streptomycete

0.01 g FeCl3 was added onto a sterile filter paper disc which was placed in the inhibition zone on plates, or 0.05% FeCl3 was added to the BTE medium, and antagonism was then evaluated in comparison to controls. Detection of siderophores using chrome azurol S (CAS) reagents and ferric perchlorate assay was undertaken according to Payne [[4]. BTE broth was inoculated with strain S96 and incubated at 26°C for 5 days; CAS reagents were added into the broth for detecting siderophores produced by S96, absorbance (A630) of CAS reagents after the addition of banana tissue extracts was measured to as a blank and tap water plus CAS reagents as a reference.

2.6 In vivo biocontrol assay

An in vivo biocontrol assay was performed to test the ability of strain S96 to control the severity of wilt in banana plantlets (Musa sp., AAA, Cavendish subgroup, cv. Williams) caused by Foc race 4. Sixty plantlets from tissue culture (1-month old) were grouped into three equal groups (20 plantlets each) and subjected to different treatments. The soil comprising a mixture of sand and clay (4:1, w/w) was used in the experiments. For treatment A, roots of the test plantlets were immersed in a suspension of strain S96 (106 cfu ml−1) for about 1 h before replanting. In treatment B and C, roots were immersed in sterile tap water for about 1 h before replanting. The plantlets were grown in pots (17 cm high × 20 cm of diameter, one plantlet per pot), each containing 3 kg soil, and distributed at random in a glass greenhouse under natural lighting and day/night temperature of 30/20°C. The plantlets were watered once every two days from the second day onwards and fertilizer was not added to the soil. After three weeks, the soil contained in each pot in treatments A and B was inoculated by pouring onto the soil surface the same amount of spores (104 spores g−1 soil) of Foc race 4. After three months of planting, the plants were removed from their pots and their fresh weights were recorded. Disease evaluation was based on external examination for chlorosis of leaves graded according to the wilt index (WI) scale and internal examination for the extent of rhizome discoloration graded according to the vascular discoloration index (VDI), as reported by Saravanan et al. [[].

2.7 Statistical analysis

Data were subjected to analysis of variance (ANOVA) and the mean values of plant fresh weight and disease index were statistically analyzed using Student's t-test. Differences were considered significant when the probability was less than 0.05.

3 Results

3.1 Identification and communities of endophytic actinomycetes

After incubation at 26°C for 3 days, endophytic actinobacterial cultures were found to grow out of banana root tissues. A total of 131 endophytic actinomycete pure cultures were isolated. They belonged to Streptomyces (99 isolates), Streptoverticillium (28 isolates) and Streptosporangium (2 isolates) genera. The remaining 2 isolates never bore reproductive structures and were not identified. Two morphologically different Streptosporangium strains and thirteen Streptoverticillium morphologically different strains were isolated. Most isolates were identified as Streptomyces species and 38 morphologically different ‘species’ were isolated. Streptomyces species were classified into eleven groups according to color of substrate and aerial hyphae on S medium (Table 1). About 40% of isolates from banana roots belonged to group A of Streptomyces and were most frequently isolated. Twenty-four strains (61.5%) of group A were similar to S. griseorubiginosus on the basis of morphological criteria. The most frequently isolated strain S96 contains ll-diaminopimelic acid based on the chemical analysis of whole cell extracts and diagnostic sugars were not detected. Analysis of the 16S rDNA gene sequences (accession no. AJ519793) by BLASTN confirmed that the isolates could be grouped in the Streptomyces clade. The highest similarity of 99% was shown between the strain S96 and S. griseorubiginosus. The results suggested that the strain S96 might belong to a member of the genus Streptomyces. Compared to other endophytic actinomycete species, S. griseorubiginosus was the most frequently isolated Streptomyces species from healthy banana roots.

View this table:

Number and characteristics of endophytic streptomycete isolated from banana roots

GroupsColor ofNumber of strainsNumber of species
Aerial hyphaeSubstrate hyphae
AGrayRed, orange or violet399
BPinkRed or orange205
DGreen or blueViolet, orange or colorless88
FGrayish blue shadesOrange, red or colorless43

3.2 Antifungal activities of endophytic actinomycetes in vitro

The antibiosis assay showed that a total of 24 strains (18.3%) were antagonistic to F. oxysporum f. sp. cubense, the pathogen of Panama disease. Most of the strains antagonistic to Fusarium strains belonged to group A in Streptomyces, while no strains belonging to groups B, D, F, G, H, I and J were antagonistic to F. oxysporum f. sp. cubense strain (Table 1). About 37.5% of S. griseorubiginosus showed activity against F. oxysporum f. sp. cubense, a proportion larger than other actinomycete taxa. The antibiosis assay on autoclaved banana pseudostem tissue showed that the distribution of mycelia of pathogenic F. oxysporum f. sp. cubense decreased in one site near the colony of endophytic Streptomyces strain in banana tissue.

3.3 Effect of iron on antagonism of streptomycete

When 0.01 g of solid FeCl3 was added onto a filter paper disc placed near the colony of strain S96, the fungi could grow towards the inhibition zone (Fig. 1), and antagonism was not observed when the Fusarium strain and endophytic streptomycete were inoculated onto BTE medium containing more than 0.05% FeCl3. But in the same medium containing less than 0.05% FeCl3, the antagonism was still observed. A weak CAS reaction (A630, −0.26) was measured when banana tissue extracts were added to the CAS reagent. This showed that Fe3+ of CAS reagents bound to the phytosiderophore in banana tissue extracts, resulting in a decline in absorbance at 630 nm. The banana tissue extracts turned blue when FeCl3 or FeSO4 were added into banana tissue extracts, suggesting that some iron complex was formed. Ferric perchlorate assays indicated that the hydrooxamate-type siderophores were formed by strain S96 incubated in BTE broth.


Effect of iron on antagonism of the endophytic Streptomyces sp. strain S96 to Fusarium oxysporum f. sp. cubense on banana tissue extracts plates. The fungi grew towards the colony of strain S96 when an excess of FeCl3 was introduced into the inhibition zone. A, Colony of strain S96 (left, the inhibition zone between the margins of the streptomycete colony and the fungal colony is distinct). B, Colony of strain S96 +0.01 g FeCl3 on filter paper disc (upper, fungal mycelia could grow throughout the inhibition zone between the margins of the streptomycete and fungal colonies). C, Filter paper disc soaked with 50 μmol l−1 FeCl3 (below) D, Filter paper disc soaked with 0.1% EDTA (right).

3.4 Biocontrol assay

Four weeks after inoculation, some visual external wilt symptoms (yellowing of leaves) of plantlets in treatment B (plantlets inoculated with Foc only) were first observed; leaf chlorosis began with older leaves and progressed to younger leaves. After six weeks, five plantlets died while the others showed mild to severe leaf symptoms. In contrast, the plantlets in treatment A (treated with strain S96 and inoculated with Foc) and treatment C (control plantlets not treated with S96 and Foc) remained healthy, without any sign of leaf chlorosis. Leaf symptoms in plantlets treated with strain S96 were seen clearly only in the sixth week; none of the plantlets died after six weeks. After eight weeks, although one plantlet in the control treatment showed some browning on the margin of lower leaves, the vascular discoloration was not observed upon cutting of suckers horizontally. It was observed that the disease severity index (DSI) for WI and VDI were significantly (P= 0.05) reduced (both about 46%) in plantlets treated with strain S96 (treatment A) compared to those grown in the absence of the biocontrol agent (treatment B) (Table 2). This showed that strain S96 could slow disease development and the expression of symptoms were delayed by about two weeks compared with the control treatment B. The mean fresh weight of plantlets in treatment A was higher compared to those in treatment C (P < 0.001). The mean fresh weight of plantlets in treatment B was lower compared to those in treatment C (P < 0.01), showing that banana plantlet growth was affected by fusarium wilt disease. The plantlets with severe wilt symptoms were stunted and eight of them died (Table 2).

View this table:

Disease severity indices (DSI) and mean fresh weight of plantlets grown in the presence and absence of endophytic Streptomyces sp. strain S96 after three months in the greenhouse

TreatmentaFresh weight (g)bDSINumber of plantlets inoculated with Foc-4Number of infected plantletsNumber of dead plantlets
A603 ± 152.12.020101
B382 ± 93.93.720198
C409 ± 121.11000
  • Foc-4, Fusarium oxysporum f. sp. cubense race 4; WI, wilt index; VDI, vascular discoloration index.

  • a A: Roots immersed in 106 cfu ml−1Streptomyces sp. strain S96 before planting and later inoculated with 104 Foc-4 spores g−1 soil in pots.

  • B: Roots immersed in sterile tap water before planting and later inoculated with 104 Foc-4 spores g−1 soil in pots.

  • C:Roots immersed in sterile tap water before planting and without Foc-4 inoculation.

  • bAverage fresh weight of five plantlets for each treatment (mean ± SD).

4 Discussion

The Streptomyces spp. identified by morphological characteristics were the most frequent isolates among the actinobacterial endoflora of banana roots in the present study. Similar results were found in field crops and Italian native plants [[], wheat roots from a range of sites across South Australia [[5] and tomato roots in South China [[3]. The majority of antagonistic streptomycetes found by the antibiosis assay have similar morphological characteristics (similar to S. griseorubiginosus), results consistent with our previous findings from the other banana plantation [[1]. Our results suggested that these S. griseorubiginosus strains are important for the resistance of healthy banana plantlets to fusarium wilt. But the proportion (18%) of antagonistic strains is lower than our previous findings (50%) (P < 0.05). It has been pointed out that disease suppressiveness of soils studied so far is essentially microbiological in nature, i.e. it results from more or less complex microbial interactions, rather than from a direct effect of soil physicochemical factors on the pathogen [[6]. So it is probable that this statistically significant difference between the proportions of antagonistic strains is related to the different soil physicochemical factors of the two banana plantations that were sampled. The soil pH of banana plantations is 5.1 in this study and 7.3 in our previous study. A principal factor mediating the equilibrium of Fe in soil is the pH; the more alkaline the soil, the less Fe3+ is available [[6]. It was hypothesized that the availability of Fe3+ is related to the antagonism of endophytic Streptomyces, and our antibiosis assay in vitro confirmed the hypothesis by adding FeCl3 to the inhibition zone of strain S96.

In this study, CAS reaction was observed when banana tissue extracts were added to the CAS reagent. This indicated that banana tissue extracts contain free phytosiderophores, so the concentration of free iron is very low. Our BTE medium is iron-restricted. Some Streptomyces strains could produce substantial amounts of antibiotics under iron-restricted conditions [[7,[8]. But the activity of a few antibiotics, such as tetracycline, is known to be influenced slightly by the presence of an excess of iron [[9,[0]. In this study, the antagonism was not observed after an excess of iron was added to the inhibition zone, so it is probable that siderophores produced by endophytic Streptomyces are involved in the antagonism and that antibiotics are not the major antifungal metabolites here. Siderophores exhibit considerable structural variability and affinity for iron, which determines the growth of a microbe under competitive conditions when availability is a limiting factor. The role of siderophores produced by pseudomonads has received more attention since Pseudomonas siderophores were found to correlate with the biocontrol of disease-suppressive soils [[1]. Some plant growth-promoting rhizobacteria promoted growth by supplying the plant with sequestered iron [[2]. Soil Streptomyces were also reported to produce certain kinds of siderophores such as enterobactin and coelichelin in iron-limited media [[3,[4]. But little work was been done on endophytic streptomycetes. Our work showed that the mean fresh weight of banana plantlets inoculated with siderophore-producing endophytic Streptomyces increased significantly compared to the control plantlets (P < 0.001). It is probable that endophytic Streptomyces strains promote the growth of banana plantlets by supplying sequestered iron to them. Previous studies on plant–microbe rhizosphere interactions involving Streptomyces lydicus and Pisum sativum indicated that root and nodule colonization by Streptomyces lead to an increase in the average size of nodules and improves the vigor of bacteroids by enhancing nodule assimilation of iron and possibly other soil nutrients [[5]. The results are consistent with our results.

Previous research work has suggested that limiting available iron in the soil could favor soil suppressiveness to fusarium wilt [[6,[7]. Siderophores produced by endophytic Streptomyces strains may induce soil suppressiveness when Fe3+ is in relatively low concentration in the soil for their stability constants were higher than the Fe-acquiring facilities of the pathogen. Furthermore, endophytic Streptomyces strains can form spores that survive adverse conditions and produce filamentous mycelia that improve colonization within the plants. So biocontrol agents for field application will be screened from siderophore-producing streptomycete endophytes. The efficacy of Streptomyces sp. strain S96 against banana fusarium wilt will be further studied in greenhouse and field trials.


We thank Prof. Y.C. Zhong for good advice in writing the manuscript. This work was supported by grants from the Department of Science and Technology, Guangdong Province, China (Grant No. 2004060225), the Youth Faculty Foundation of Zhongshan University, and the Chinese National Natural Science Fund (Grant No. 30370030).


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
View Abstract