Control of pathogenic Vibrio spp. by Bacillus subtilis BT23, a possible probiotic treatment for black tiger shrimp Penaeus monodon


Correspondence to: Dr B. Vaseeharan, Genetics and Biotechnology Division, Central Institute of Brackish Water Aquaculture, 75 Santhome High Road, R.A. Puram, Chennai 600 028, India (e-mail:


Aims: The present study evaluated the in vitro and in vivo antagonistic effect of Bacillus against the pathogenic vibrios.

Methods and Results: Cell-free extracts of Bacillus subtilis BT23 showed greater inhibitory effects against the growth of Vibrio harveyi isolated by agar antagonism assay from Penaeus monodon with black gill disease. The probiotic effect of Bacillus was tested by exposing shrimp to B. subtilis BT23 at a density of 106−108 cfu ml−1 for 6 d before a challenge with V. harveyi at 103−104 cfu ml−1 for 1 h infection. The combined results of long- and short-term probiotic treatment of B. subtilis BT23 showed a 90% reduction in accumulated mortality.

Conclusions: This study reports that pathogenic vibrios were controlled by Bacillus under in vitro and in vivo conditions.

Significance and Impact of the Study: Results indicated that probiotic treatment offers a promising alternative to the use of antibiotics in shrimp aquaculture.


Aquaculture is the world's fastest growing food production sector, with cultured shrimp and prawn growing at an annual rate of 16·8% between 1984 and 1995 (Subasinghe et al. 1998). However, disease outbreaks have caused serious economic losses in several countries. According to a recent World Bank report, global losses due to shrimp diseases are around US$ 3 billion (Lundin 1996). Vibrio species occur as the dominant flora in various developmental stages of Penaeus monodon and have been described as the causal pathogens (Lightner 1996; Sung et al. 2001). Potential negative consequences of using antibiotics in aquaculture for the prophylactic treatment of diseases are the development of drug-resistant bacteria and reduced efficacy of antibiotic treatment for human and animal diseases (Moriarty 1997). Increased concern about antibiotic-resistant micro-organisms (Amabile et al. 1995) has led to suggestions of alternative disease prevention methods, including the use of non-pathogenic bacteria as probiotic biocontrol agents. (Austin et al. 1995; Moriarty 1997). Lactic acid bacteria have been tested as probiotics in warm-blooded animals and attempts have been made to use lactic acid bacteria as antagonists (probiotics) of shrimp pathogens (Gatesoupe 1999; Skjermo and Vadstein 1999). Bacillus spores have been used as biocontrol agents to reduce vibrios in shrimp culture facilities (Skjermo and Vadstein 1999; Rengipipat et al. 2000). Bacillus constitutes a large part of the microflora of the gills, skin and intestinal tracts of shrimps (Sharmila et al. 1996). Bacillus spp. are often antagonistic against other micro-organisms, including fish and shellfish pathogenic bacteria (Gatesoupe 1999; Rengipipat et al. 2000). The present study investigated the inhibitory activity of Bacillus subtilis BT23, isolated from shrimp culture ponds, against pathogenic Vibrio harveyi under in vitro and in vivo conditions.

Materials and methods

Bacterial strains

A virulent strain of V. harveyi, obtained from P. monodon with black gill disease (LD50 value 105 cfu ml−1 under experimental infection of P. monodon juveniles), was used as a pathogenic strain and B. subtilis BT23, obtained from shrimp culture ponds, was used as an antagonistic strain. Bacillus BT21, Bacillus BT22 and B. subtilis BT23 were isolated from shrimp culture ponds. The bacteria used in this study were identified using standard morphological, physiological and biochemical plate and tube tests (Holt et al. 1994). All strains were taken from the stock culture collection of our laboratory and had been stored in Luria-Bertani (LB) broth cultures with sterile glycerol (15% v/v).

Antagonism assay

The initial screening of antagonism by Bacillus BT21, Bacillus BT22 and B. subtilis BT23 was by the agar well diffusion plate assay method (Bauer et al. 1966). Vibrio harveyi, V. anguillarum, V. vulnificus and V. damsela were precultured in LB broth incubated at 28 °C for 2 d and 50 µl of this culture were spread over the agar plates. Bacillus spp. culture (3 d old) was centrifuged at 9600 rev min−1 for 15 min and the supernatant fluid filtered through a 0·22-µm membrane filter (Sartorius, Bedford, MA, USA) to obtain cell-free extracts (culture supernatant fluid). A volume (100 µl) of Bacillus cell-free extracts was introduced into the wells of the agar medium and incubated for a period of 24–48 h at 28 °C. Antibacterial activity was defined as the diameter (mm) of the clear inhibitory zone formed around the well.

Co-culture experiments

Bacillus subtilis BT23 and V. harveyi were precultured separately in LB broth at 28 °C for 3 d. Vibrio harveyi was inoculated into LB broth at an initial cell density of approx. 103 cfu ml−1, whereas the initial levels of B. subtilis BT23 were 105, 107, 108 and 109 cfu ml−1. All combinations were performed in triplicate. The co-culture plates were incubated at 28 °C and samples withdrawn daily for the determination of V. harveyi densities. The numbers of V. harveyi were estimated by preparing 10-fold serial dilutions and 0·1 ml from each dilution was inoculated into thiosulphate citrate bile salts sucrose agar plates.

Effect of Bacillus subtilis BT23 cell-free extracts

Bacillus subtilis BT23 was precultured in LB broth and then used to inoculate 50 ml of LB broth in the same four combinations at an initial cell density of 104−108 cfu ml−1. The flasks were incubated at 28 °C with agitation (200 rev min−1) and samples withdrawn daily and the number of cfu determined. Sterile filtered supernatant fluid (2 ml) was tested by adding 1 ml of supernatant fluid to 1 ml of fresh LB broth in test tubes and inoculating it with 100 µl of V. harveyi in LB broth, yielding approx. 104 cfu ml−1. Controls were made by inoculating V. harveyi (100 µl) in 2 ml of LB broth without B. subtilis BT23 cell-free extracts. Each combination was tested in triplicate and the growth of the V. harveyi monitored by recording the optical density at 600 nm with a spectrophotometer.

Experimental infection of shrimp and Bacillus subtilis BT23 treatment

Bacillus subtilis BT23 was grown for 3 d at 28 °C (150 rev min−1) in LB broth and V. harveyi was grown for 24 h in tryptone soy broth. Two hundred P. monodon, each approx. 5–6 g, were divided equally into eight groups, each housed in a 300-l tank. Four of the tanks were treated with B. subtilis BT23 for 5 d at a level of 106 cfu ml−1 at 28 °C (long-term treatment) by adding the bacteria to the water. After 5 d, shrimp in all of the four tanks were infected with V. harveyi (104−106 cfu ml−1) for 1 h and two of the tanks were again treated with B. subtilis BT23 (106 cfu ml−1) for 1 h along with V. harveyi (combined treatment).Of the remaining four tanks, two were treated with B. subtilis BT23 (106 cfu ml−1) for 1 h and the animals were then transferred to a tank in which V. harveyi (104−106 cfu ml−1) was maintained for 1 h (short-term treatment); the other two tanks were infected with V. harveyi (104−106 cfu ml−1) alone for positive control. The cumulative mortality of the shrimp was recorded and analysed using analysis of variance.


Antagonism assay

The cell-free extract of Bacillus BT21, Bacillus BT22 and B. subtilis BT23 showed inhibitory activity against Vibrio spp. Of these, B. subtilis BT23 showed a higher inhibitory activity than the other two Bacillus spp. tested. BT23 showed inhibitory activity against 112 Vibrio spp., V. harveyi (39 isolates), V. anguillarum (24 isolates), V. vulnificus (30 isolates) and V. damsela (19 isolates) obtained from P. monodon culture hatcheries and ponds (Table 1). The diameters of the inhibitory zones around the growth of Vibrio spp. were about 3–6 mm (Fig. 1).

Table 1.  Inhibitory activity of Bacillus subtilis BT23 against Vibrio harveyi, V. anguillarum, V. vulnificus and V. damsela obtained from Penaeus monodon culture hatcheries and ponds
 B. subtilis with inhibitory effects against
Source of Vibrio spp.V. harveyi isolatesV. anguillarum isolatesV. vulnificus isolatesV. damsela isolates
  1. MBV, Monodon baculovirus; WSSV, white spot syndrome virus.

Vibriosis-infected post-larval P. monodon73124
MBV-infected post-larval P. monodon96118
Black gill-diseased P. monodon juveniles12743
WSSV-infected P. monodon gills10834
Figure 1.

Petri dishes containing cell-free extract of (a) Bacillus BT21, (b) Bacillus BT22 and (d) Bacillus subtilis BT23 showed inhibitory zones against the growth of (A) Vibrio harveyi, (B) V. anguillarum and (C) V. damsela. No inhibitory zone was found in the control (c). Note the Bacillus subtilis BT23 (d) showing the greatest inhibitory zones. The strain was identified as B. subtilis BT23 and used for further in vivo and in vitro studies

Co-culture experiments

The growth of pathogenic V. harveyi was inhibited by B. subtilis BT23 culture inoculated at an initial level of 105−109 cfu ml−1(Fig. 2). Lower concentrations of B. subtilis BT23 (105 and 107 cfu ml−1) allowed initial growth of V. harveyi, but cfu densities never reached the level of the control. High concentrations (109 cfu ml−1) of B. subtilis BT23 allowed an initial increase of V. harveyi followed by a decrease in the total viable counts (Fig. 2). Co-culture experiment results showed that, when the concentration of B. subtilis BT23 increased, the growth of V. harveyi was controlled under in vitro conditions.

Figure 2.

Growth pattern of Vibrio harveyi at 28 °C with and without Bacillus subtilis BT23 at different initial concentrations (colony-forming units; cfu). ◆, Without B. subtilis; ▮, B. subtilis 105 cfu ml−1; ▴, B. subtilis 107 cfu ml−1; ○, B. subtilis 108 cfu ml−1 and •, B. subtilis 109 cfu ml−1

Effect of cell-free extracts of Bacillus subtilis BT23

Cell-free extracts of B. subtilis BT23 inhibited the growth of V. harveyi in liquid culture under aerobic conditions. The inhibitory efficiency was high in B. subtilis BT23 cell-free extracts of 108 cfu ml−1 and low in 104 cfu ml−1. Bacillus subtilis BT23 cell-free extracts did not restrict the growth of V. harveyi for 2 d and after that the growth was remarkably controlled (Fig. 3) when compared with the growth of V. harveyi without B. subtilis BT23.

Figure 3.

Growth of Vibrio harveyi at 28 °C with cell-free extracts of Bacillus subtilis BT23 extracted by different cell densities. With B. subtilis cell-free extracts of: ◆, 104 cfu ml−1; ▮, 105 cfu ml−1; ▴, 106 cfu ml−1; ○, 107 cfu ml−1 and •, 108 cfu ml−1

Experimental infection of shrimp and probiotic treatment

The studies on the probiotic treatment and infection of shrimp revealed that the mortality of shrimp by V. harveyi infection was reduced by B. subtilis BT23 strains under in vivo conditions. The cumulative mortality of infected P. monodon not treated with B. subtilis BT23 reached 50% on the 9th day after infection with V. harveyi and 100% on the 17th day. However, in the case of probiotic treatment groups, the mortality levelled off after 5 d when the cumulative mortality was 10% in the combined treatment (Fig. 4). Both the long- and short-term treatments with B. subtilis BT23 caused a decrease in cumulative mortality, to 32 and 60%, respectively. No mortality was found in control tanks which were not exposed to V. harveyi. The effect of probiotic treatment was most pronounced during the first day of infection.

Figure 4.

Cumulative mortality of Penaeus monodon juveniles infected with Vibrio harveyi with and without probiotic treatment of Bacillus subtilis BT23. ○, Control; ▴, long-term treatment; ▮, short-term treatment and ◆, combined treatment


The present study showed that the growth of pathogenic V. harveyi was controlled by non-pathogenic B. subtilis BT23 under in vivo and in vitro conditions. The control of fish and shellfish pathogenic Vibrio, particularly using non-pathogenic bacterial strains and disease prevention, has received much attention during the last decade (Sugita et al. 1998; Rengipipat et al. 2000). Fuller (1989) defined a probiotic as a live microbial feed supplement which benefits the host animal by improving its intestinal microbial balance. Co-culture experiments showed that the inhibitory activity of B. subtilis BT23 increased with increasing density of the antagonist. A high concentration of B. subtilis BT23 (antagonist) was required to inhibit V. harveyi in the co-culture experiments. The present study showed that the antagonist must be present at significantly higher levels than the pathogen and the degree of inhibition increased with the level of antagonist. During the co-culture, 107−109 cfu ml−1 were required to inhibit the growth of the pathogen V. harveyi. Therefore, a potential probiotic co-culture must either be supplied on a regular basis or be able to colonize and multiply on or in the host. The Bacillus spp. used as probiotics for terrestrial livestock are of telluric origin and are not autochthonous in the gastrointestinal tract but they may be active during intestinal transit (Gouthier et al. 1994). Kennedy et al. (1998) isolated a strain of B. subtilis from common snook (Centropomus undecimalis). The inoculation of this strain into the rearing water resulted in the apparent elimination of Vibrio spp. from the snook larvae. Smith and Davey (1993) reported that Pseudomonas fluorescensreduced diseases caused by Aeromonas solmonicida in fish. Austin et al. (1995) also observed a similar phenomenon, that V. alginolyticus, used as a probiotic strain, reduced the diseases caused by Aerom. solmonicida, V. anguillarum and V. ordalli in P. monodon. Maedo and Liao (1992) reported the use of a soil bacterial strain, PM-4, that promoted the growth of P. monodon nauplius, probably acting as a food source. This strain also showed an in vitro inhibitory effect against V. anguillarum.Rengipipat et al. (1998) reported that inoculation of Bacillus S11, a saprophytic strain, resulted in greater survival of the post-larval P. monodon that were challenged by pathogenic luminescent bacterial culture. These works strongly suggest the effective control of microflora in fish and shellfish in culture environments by antibiotic-producing bacteria. Purification and characterization of the antibacterial substance would help to understand the mechanism of antibacterial activity of Bacillus strains.

Probiotic treatment offers a very promising alternative to the use of antibiotics in fish and shrimp aquaculture. Further study is needed to elucidate the exact mode of action of the observed beneficial effects and to understand the possibilities and limitations of microbial control in aquaculture.


The authors thank Dr Junda Lin (Florida Institute of Technology, Melbourne, FL, USA) for his comments on an earlier version of the manuscript.