A case study on chemical defense based on quorum sensing: antibacterial activity of sponge-associated bacterium Pseudoalteromonas sp. NJ6-3-1 induced by quorum sensing mechanisms

Abstract

A case study to investigate the relationship between antibacterial activity and quorum sensing mechanisms was carried out on a sponge-associated bacterium with remarkable biological activities: Pseudoalteromonas sp. NJ6-3-1. The dependence of active substance production on cell density was studied under various growth conditions. Bacteria NJ6-3-1 was found to start producing antibacterial compounds only when cell density reached the threshold value of OD₆₃₀ = 0.4. To simulate the competitive real marine environment, NJ6-3-1 at low cell density (OD₆₃₀ value below the required threshold value) was co-cultured with the terrestrial bacterium Staphylococcus aureus. Antibacterial activity assays indicated the existence of some signal molecules in the metabolites of S. aureus that could induce NJ6-3-1 to produce antibacterial substances even at low cell density. Three diketopiperazines (DKPs) as metabolites and potential autoinducers of NJ6-3-1 were synthesized and co-cultured with low density NJ6-3-1. The antibacterial activity assay showed that one of these DKPs—cyclo-(l-Phe-l-Val)—was the autoinducer and could indeed induce NJ6-3-1 to produce antibacterial substances under low cell density. Our results thus provide preliminary support to the hypothesis that the antibacterial activity of NJ6-3-1 is controlled by the quorum sensing system in both an intra-species and an inter-species manner.

Introduction

Marine microorganisms are the focus of much current marine biochemical research because of their wide diversity in terms of bioactivity and resource sustainability, which makes them ideal candidates for marine drug exploration and development. For such development, the two main bottlenecks are the low content of bioactive compounds in marine organisms, and the limited supply of bioresources. As work in this field intensifies, researchers are paying increasing attention to the biological mechanisms of production of bioactive metabolites by microorganisms. This is due to recognition of the fact that a better understanding of the purpose of bioactive metabolite production will be the most effective way to improve the efficiency of bioactivity screening, and to enhance the yield of active substances from marine microorganisms (Bultel-Ponce et al. 1999; Holler et al. 2000; Kelecom 2002; Isnansetyo and Kamei 2003). Owing to the complex and special living circumstances in which they reside, marine microorganisms usually coexist in the marine environment. Many scientists have proposed that the need for a chemical defense mechanism is the main condition inducing marine microorganisms to produce bioactive substances (Dushenkov and Raskin 2008; Huang and Li 2006; Burgess et al. 1999).

Quorum sensing phenomena have been gradually revealed with the development of microbiology (Fuqua et al. 1994; Bassler 2002; Federle and Bassler 2003), and more and more scholars have proposed the hypothesis that chemical defense mechanisms in microorganism, especially bacteria, are regulated by quorum sensing (Hentzer and Givskov 2003; Rasmussen et al. 2005).

Quorum sensing is a metabolic regulation and control process that enables bacterial cells to sense changes in cell density and induce the expression of specific genes (Fuqua et al. 1994). This regulatory process is mediated through sensing changes in the concentration of specific signal molecules being released either by the bacteria themselves, or from external sources. When the bacteria grow to a high cell density, the concentration of autoinducers (AI) they release could reach a threshold value. The bacteria could then sense these AI, with the subsequent appearance of certain specific physiological characters, including the production of antibiotics, virulence expression, biofilm formation and so on (Winzer and Williams 2001; Shih and Huang 2002; Parsek and Greenberg 2005).

To date, two kinds of quorum sensing systems, namely, intraspecies and interspecies systems, have been discovered. The AI-mediated quorum sensing system has been proposed to be primarily an intraspecific communication system in which bacteria sense the release of AI by themselves, and then express certain physiological characters (Williams 2006). The AI-2-mediated quorum sensing system is proposed to be an interspecific communication system in which bacteria sense interspecies signal molecules released by exogenous microorganisms that are then used to express some specific physiological characteristic (Bassler 2002; McDougald et al. 2003; Williams 2006).

Regarding signal molecules, acylated homoserine lactones (AHLs) were first confirmed two decades ago to be AI signal molecules produced by Gram negative bacteria (Swift et al. 1994). Besides AHLs, the oligopeptides known as autoinducing peptides(AIPs) released by Gram-positive bacteria, and some aromatic alcohols released by fungi are also known to be AI intraspecies quorum sensing signals (Voloshin and Kaprelyants 2004; Nickerson et al. 2006; Chen and Fink 2006). The AI-2 signal molecule, which is predicted to be a furanosyl borate diester, has been proposed to serve as a universal signal for interspecies communication, and is found both in Gram-negative and Gram-positive bacteria (Federle and Bassler 2003; March and Bentley 2004). Other than AI and AI-2 signal molecules, many new signal molecules have also been found recently, with some examples including 2-heptyl-3-hydroxy-4-quinolone (PQS), diketopiperazines (DKPs), and methyl dodecenoic acid (DSF) (Holden et al. 2000; Mitova et al. 2004; Guo and Chen 2007).

In our previous studies, a sponge-associated marine bacterium, Pseudoalteromonas sp. NJ6-3-1, attracted our interest because of its remarkable biological activities. Three DKPs—cyclo-(ΔVal-l-Val), cyclo-(l-Phe-l-Val) and cyclo-(l-Pro-l-Leu)—were isolated from NJ6-3-1. It was discovered that the antibacterial and cytotoxic activities of NJ6-3-1 could be enhanced when it was co-cultured with exogenous bacteria (Zheng et al. 2005b). These results suggested the likely presence of a quorum sensing system in NJ6-3-1. In this study, NJ6-3-1 was chosen as a case study to verify our proposed hypothesis that chemical defense of marine bacteria, namely the production of antibacterial substances, is regulated by quorum sensing mechanism. An experiment to study the dependence of the production of active substances on cell density was carried out. The antibacterial activity of NJ6-3-1 induced by exogenous bacteria and potential autoinducer DKPs under low density conditions was also investigated.

Materials and methods

Bacterial strains and culture conditions

The marine bacterium NJ6-3-1(AY621063) used in this study was isolated from the sponge Hymeniacidon perleve and identified as Pseudoalteromonas sp. based on its 16 S rRNA sequence analysis. Its secondary metabolites show remarkable antimicrobial and cytotoxic activity (Zheng et al. 2005a). It was grown in Marine Broth (MB, peptone 5 g, yeast extract 1 g and FePO₄ 0.1 g, dissolved in 1 L seawater, pH 7.2–7.6).

The co-cultured bacterium was Staphylococcus aureus CMCC (B) 26001, which was cultured in nutrient medium (peptone 5 g, yeast extract 5 g, beef extract 5 g, glucose 5 g and NaCl 5 g, dissolved in 1 L water, pH 7.0).

As two different AHL monitor strains,Chromobacterium violaceum CV026 and Agrobacterium tumefaciens A136 (pCF218)(pCF372) (McLean et al. 2004) were used to detect AHL-production. These monitor strains were grown in LB supplemented with appropriate antibiotics (Ravn et al. 2001), and incubated at 28°C for 24 h. Chromobacterium violaceum 31532 and A. tumefaciens KYC6 were used as positive controls in the experiments while the monitor strains themselves acted as negative controls.

Low cell density determination experiments

Bacterial culture samples each containing 5 ml of suspension of marine bacterium NJ6-3-1 (OD₆₃₀ = 0.6) in 300 ml MB medium were prepared in 500 ml Erlenmeyer flasks. These culture samples were inoculated individually under nine different conditions (listed in Table 1). Cell density was determined by periodically monitoring the absorbance of OD₆₃₀ by Microplate Reader RT-2100C. The monitoring process ended when the bacteria entered the decline phase.

Table 1 Antibacterial activity of metabolites and cell density of bacterium NJ6-3-1 in different culture conditions. MB Marine broth

All the crude metabolites of strain NJ6-3-1 cultured under different conditions were extracted and their antibacterial activity against S. aureus was investigated. Low cell density was determined down to the threshold cell density of the antimicrobial metabolites produced, meaning that strain NJ6-3-1 produces antibacterial metabolites only when its cell density equals or exceeds this threshold value. The culture condition of strain NJ6-3-1 at low cell density was used in all subsequent experiments.

Interspecies induction experiments

Interspecies induction experiments were carried out in order to determine whether strain NJ6-3-1 at low cell density could induce the production of antibacterial metabolites through co-culture with the terrestrial bacterium S. aureus.

Staphylococcus aureus was grown in 150 ml nutrient medium in 250 ml Erlenmeyer flasks at 37°C for 2 days. Then, 50 ml culture liquid was centrifuged at 6,000 g for 15 min and the supernatant removed. The cell residue mixed with 50 ml 1/5 MB medium was termed S. aureus living suspension. Another 50 ml S. aureus living suspension sterilized by autoclaving at 121°C for 15 min was termed S. aureus dead suspension. Supernatant (50 ml) was filtered through a 0.22 μm filter membrane and the filtrate was regarded as S. aureus sterile metabolites. All samples were stored at 4°C.

A 300 ml culture of marine bacterium NJ6-3-1 was grown under conditions of low cell density pre-determined as described above. When strain NJ6-3-1 reached the exponential growth phase, the above-described samples were added individually to the culture liquid of strain NI6-3-1. The co-cultured liquids were then incubated for 4 days at 25°C until the bacteria reached the maximal growth rate. The antibacterial activity of crude metabolites from strain NJ6-3-1 was then investigated.

Intraspecies induction experiments

Three DKP metabolites from NJ6-3-1 as potential AI were synthesized by GL Biochem. Six milligrams of synthesized cyclo-(ΔVal-l-Val), cyclo-(l-Phe-l-Val), cyclo-(l-Pro-l-Leu) were dissolved individually in 1 ml methanol. They were then subjected to ultrasound to ensure complete dissolution.

Samples of 300 ml marine bacterium NJ6-3-1 were cultured under conditions of low cell density. When strain NJ6-3-1 reached the exponential growth phase, three DKPs and 1 ml methanol as the control were added individually into the culture liquid of strain NI6-3-1. The co-cultured liquids were then incubated at 25°C for 4 days until the bacteria reached the maximal growth rate. The antibacterial activity of crude metabolites from strain NJ6-3-1 was then investigated.

Screening for AHL production

Screening for AHL production was performed according to Pinto et al. (2007). The tested strains were streaked in parallel to the monitor strains on LB agar plates. Chromobacterium violaceum CV026 produces purple pigment only in the presence of exogenous AHL. Thus, pigment production indicated a positive result in response to AHL produced by the tested strain. Agrobacterium tumefaciens A136 produces a blue pigment in response to AHL when the medium is supplemented with 50 μg/ml 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal). All plates were incubated at 28°C for 24 h, except those streaked with A. tumefaciens A136, which were incubated for up to 3 days. The monitor strain was then streaked in parallel with the positive strains, and the plates were re-incubated under the same conditions. All experiments were repeated at least twice.

Preparation of crude metabolite extract

After bacterium NJ6-3-1 reached stationary phase, the culture liquid was first centrifuged at 6,000 g for 15 min to remove the cells, and then extracted three times with ethyl acetate (EtOAc) (100 ml × 3). After solvent removal under reduced pressure at 37°C, the extracts were used as the crude metabolites samples for bioactivity assays.

Antimicrobial activity of the crude metabolites from strain NJ6-3-1

The terrestrial bacterium S. aureus was used as the test microorganism to determine the antibacterial activity of the marine bacterium NJ6-3-1. Antibacterial activity was assayed in duplicate using a standard paper disc assay (Mearns-Spragg et al. 1998). The dried crude extracts were dissolved in EtOAc to a concentration of 100 mg ml⁻¹. The samples (20 μl) were used to saturate the antimicrobial assay paper disks (6 mm) with a drying period between each application. The disks were placed onto the agar surface containing the test microorganism, and incubated at 37°C for 24 h after a diffusion process at 4°C for 10 h. The diameters of any inhibition zones formed around the paper disks were then measured.

Thin-layer chromatography autobiography overlay assay

The crude metabolite extracts of marine bacterium NJ6-3-1 were used in a thin-layer chromatography (TLC) autobiography overlay assay. Crude extracts were dissolved in EtOAc and made up to a concentration of 100 mg ml⁻¹. The solution (2 μl) was submitted to TLC analysis on a 5 × 5 cm silica gel plate (TLC aluminium sheets, 20 × 20 cm, Silica Gel 60F₂₅₄, Merck, Whitehouse Station, NJ) using a mixture of dichloromethane/methanol (10:1, v/v) as the mobile phase. The developed TLC plates were sterilized under a UV lamp for 30 min before being encased in a base of nutrient agar in a Petri dish (9 mm). This was then covered by molten nutrient agar (45°C) containing the test microorganism S. aureus. After a 10-h diffusion process at 4°C, the plate was incubated at 37°C for 24 h and the upper agar was sprayed with 5 mg ml⁻¹ methylthiazoletetrazolium (Sigma, M5655, St. Louis, MO), which is converted to a formazan dye by the test microorganism. Inhibition zones were observed as clear spots against a purple background and their R _f values were calculated.

Results and discussion

Growth of marine bacterium NJ6-3-1 under different culture conditions

The maximum growth density of NJ6-3-1 differed under different culture conditions (Fig. 1). The cell density of strain NJ6-3-1 was always lowest in 1/5 MB at different temperatures of 20, 25 and 30°C (Fig.1a–c), suggesting the existence of an intimate relationship between the growth of strain NJ6-3-1 and the nutrient status of the medium.

Fig. 1

a–f Growth curves of bacterium NJ6-3-1 in different culture conditions. a–c Growth of NJ6-3-1 at different temperatures: a 20°C, b 25 °C, c 30 °C; d–f growth of NJ6-3-1 in the different strength marine broth (MB) medium: d MB, e 1/2 MB, f 1/5 MB

In the same medium, the higher the temperature, the greater the cell density of strain NJ6-3-1, and the shorter the time required to reach maximum growth density. Similar results were found in all types of media tested (Fig.1d–f), indicating the strong influence of temperature on the growth of bacteria NJ6-3-1.

Relationship between cell density and antibacterial activity of metabolites produced by strain NJ6-3-1

The crude metabolites of strain NJ6-3-1 grown under different conditions were extracted and their respective antibacterial activity was determined by standard paper disc assay. The results are shown in Table 1. Production of antibacterial substances was not universal among all metabolite extracts, but varied depending on the cell density of strain NJ6-3-1. Antibacterial activity was highest when cultured in MB and 1/2 MB, at temperatures of 25 and 30°C, respectively. At a low cell density of 1/5 MB, the strain NJ6-3-1 did not produce detectable activity at all the temperatures tested, but did produce activity in the other conditions.

The relationship between antibacterial activity and cell density is shown in Fig. 2. It is evident that strain NJ6-3-1 could produce antibacterial substances only when the OD₆₃₀ value was above the threshold value of 0.4. Significantly, it was noted that strain NJ6-3-1 did not demonstrate antibacterial activity when incubated in 1/5 MB medium. These results again illustrate that the production of antibacterial substances of NJ6-3-1 depended strongly on cell density, and the threshold value of cell density was OD₆₃₀ = 0.4. In our experiment, the stationary phase was either too long or too short when culturing strain NJ6-3-1 at a temperature of 20 or 30°C, making it difficult to control the threshold value of cell density to allow for production of antibacterial activity at these two temperatures. Thus, the low cell density condition in the cell-dependent experiment was carried out under the chosen conditions of 1/5 MB and 25°C.

Fig. 2

Antibacterial activity and cell density of bacterium NJ6-3-1 in different culture conditions. Culture conditions: 1–9 indicate different culture conditions as defined in Table 1. Antibacterial activity: + Inhibition zone 1–3 mm, ++ inhibition zone 3–5 mm, +++ inhibition zone ≥ 5mm

The cell density-dependent experiment indicated that the production of antibacterial metabolites by strain NJ6-3-1 was most probably regulated primarily by quorum sensing systems. Increasing evidence is now available showing the importance of quorum sensing in regulating the metabolism of antibacterial substances in microorganisms. Notable examples include the antibacterial polyketide metabolized by Pseudomonas fluorescens NCIMB (Pirhonen et al. 1993), the bacteriocin produced by lactic acid bacteria (De Kievit and Iglewski 2000), and the biosynthesis of antibiotic β-lactam carbapenems by the plant soft rot pathogen Erwinia carotovora (Byers et al. 2002). Kievit and Byers also found that antibiotic production by marine bacteria Pseudomonas aureofaciens and Erwinia carotovora is controlled by quorum sensing (El-Sayed et al. 2001; Quadri 2002; Whitehead et al. 2002). The results of the present study are consistent with these examples.

Antibacterial activity produced by NJ6-3-1 in low density culture condition after adding different substances from Staphylococcus aureus

The presence of S. aureus in the growth medium has been reported to induce production of antimicrobial compounds by, and enhancement of antibacterial activity in, marine bacteria (Mearns-Spragg et al. 1998). In addition, we found in a previous study (Zheng et al. 2005b) that the antibacterial and cytotoxic activities of strain NJ6-3-1 could be enhanced when co-cultured with the exogenous bacteria S. aureus. Staphylococcus aureus was therefore chosen as the exogenous bacteria to co-culture with strain NJ6-3-1 in interspecies induction experiments.

The marine bacteria NJ6-3-1 at low cell density was co-cultured with three different types of exogenous samples after the bacterium entered the exponential growth phase. The antibacterial activity of strain NJ6-3-1 was then determined using an autobiographic overlay assay after strain NJ6-3-1 had reached the stationary phase (Fig. 3).

Fig. 3

Thin layer chromatography (TLC) bioautographic overlay assay of metabolites produced by bacterium NJ6-3-1 when co-cultured with different substances from Staphylococcus aureus under low cell density conditions. 1 50 ml S. aureus living suspension, 2 S. aureus dead suspension, 3 S. aureus sterile metabolite. TheR _f values of the main antibacterial compounds in samples 2 and 3 are 0.031 and 0.041, respectively

Results from the autobiographic overlay assay (Fig. 3) showed the absence of antibacterial activity by strain NJ6-3-1 co-cultured with the S. aureus dead suspension. However, remarkable antibacterial activity was demonstrated by strain NJ6-3-1 when it was co-cultured with S. aureus living suspension and S. aureus sterile filtrate during the course of cultivation. Also, a new antibacterial substance with an R _f value of 0.1 appeared in the TLC aluminium sheets.

Pathogenic S. aureus employs AHLs and peptides, respectively, to control the expression of multiple virulence genes in concert with cell population density (Winzer and Williams 2001). Furthermore, S. aureus possesses a functional luxS gene and is capable of producing the signal molecule AI-2 in addition to AIPs (Doherty et al. 2006). The AI-2 so produced could also be employed for interspecies communication (Surette and Bassler 1998). Our experiment indicated that signal molecules most probably exist in the metabolites of S. aureus. These metabolites could then induce strain NJ6-3-1 to produce antibacterial substances under low cell density conditions. This supports our hypothesis that the metabolism of antibacterial substances in strain NJ6-3-1 is most probably regulated by an interspecies quorum sensing system.

Antibacterial activity produced by strain NJ6-3-1 in low density culture conditions with different DKPs

In the intraspecies induction experiments, marine bacterium NJ6-3-1 of low cell density were co-cultured with three kinds of DKPs according to the procedure described in Materials and methods. Methanol (1 ml) co-cultured under identical conditions was used as the control (Fig. 4).

Fig. 4

TLC bioautographic overlay assay of antibacterial activity against S. aureus of metabolites produced by bacterium NJ6-3-1 when co-cultured with different diketopiperazines (DKPs) under low cell density conditions. 1 1 ml methanol (control), 2 cyclo-(ΔVal-l-Val), 3 cyclo-(l-Phe-l-Val), 4 cyclo-(l-Pro-l-Leu)

Figure 4 illustrates that strain NJ6-3-1 showed no antibacterial activity when it was co-cultured with 1 ml methanol. Negative results were also observed for cyclo-(ΔVal-l-Val) and cyclo-(l-Pro-l-Leu). However, remarkable antibacterial activity was indeed observed when strain NJ6-3-1 was co-cultured with cyclo-(l-Phe-l-Val) during the course of cultivation, and a new antibacterial substance appeared in the TLC plate with a R _f value of 0.1. The results demonstrated that cyclo-(l-Phe-l-Val) could induce NJ6-3-1 to produce antibacterial products under low cell density conditions, and cyclo-(l-Phe-l-Val) may actually be the AI signal molecule produced by bacterium NJ6-3-1.

The possible inhibitory effect on S. aureus of the three DKPs [cyclo-(ΔVal-L-Val), cyclo-(l-Phe-l-Val) and cyclo-(l-Pro-l-Leu)] was investigated; no such effect was observed (Fig. 5).

Fig. 5

TLC bioautographic overlay assay of antibacterial activity against S. aureus of different DKPs. 1 1 ml methanol (control), 2 cyclo-(ΔVal-l-Val), 3 cyclo-(l-Phe-l-Val), 4 cyclo-(l-Pro-l-Leu)

DKPs have been suspected of being the signal molecules of several Gram negative bacteria (Holden et al. 1999; Degrassi et al. 2002). Holden et al. (1999) found that Pseudomonas aeruginosa and other Gram-negative bacteria produce DKPs that could activate the LuxR-based AHL biosensor Escherichia coli (pSB401) and could compete with 3-oxo-C6-HSL for the same LuxR-binding site. DKPs were found to be capable of activating or antagonizing other quorum-sensing systems, such as the N-butanoylhomoserine lactone-dependent swarming motility of Serratia liquefaciens (Holden et al. 1999), pigment production in the C. violaceum AHL reporter strain CV026 (McClean et al. 1997) and the A. tumefaciens AHL biosensor strain NT1(pDCI41E33) (Cha et al. 1998; Shaw et al. 1997). We believe DKPs are indeed a kind of AI based on their observed capacity to induce production of antibacterial metabolites. Actually, DKPs have been isolated from many marine bacteria (Fdhila et al. 2003; Trischman et al. 2004), especially from sponge-associated bacteria (Jayatalike et al. 1996; De Rosa et al. 2003; Li et al. 2008), and many researchers have proposed them as potential quorum sensing sensors used by Gram-negative bacteria for cell–cell communication and for regulating gene expression in response to population density (De Kievit and Iglewski 2000; Degrassi et al. 2002; Li et al. 2008). Some results have demonstrated that DKPs have similar physicochemical properties to short-chain AHLs and play a biologically significant role in modulating quorum sensing and regulating different bacterial functions. (Swift et al. 1993; Holden et al. 1999; Brelles-Mariño and Bedmar 2001). Our finding that DKPs are a type of AI and capable of regulating the production of antibacterial metabolites thus provides new evidence in support of this hypothesis.

Screening for AHL in bacterium NJ6-3-1

Using the multiple monitor system overcomes the limitations of bioassays since bacteria often produce more than one AHL molecule. Bacterium NJ6-3-1 elicited diverse responses in the two AHL-monitor systems used here, thus demonstrating production of different AHL molecules.

Positive strains C. violaceum 31532 and A. tumefaciens KYC6 could induce the monitor strains C. violaceum CV026 and A. tumefaciens A136 to produce purple and blue pigment (Fig. 6b,e). C. violaceum CV026 did not produce the purple pigment, while A. tumefaciens A136 produced a blue pigment when they were streaked in parallel on strain NJ6-3-1 (Fig. 6c,f). This result indicates that NJ6-3-1 produces molecules that exhibit only a weak response to A. tumefaciens A136, i.e., it is likely that only low amounts of AHLs are produced by strain NJ6-3-1.

Fig. 6

a–f Agar plate assays for screening of acylated homoserine lactones (AHL production) in bacterium NJ6-3-1. Monitor strains were streaked parallel to themselves, positive strains, and the tested bacteria NJ6-3-1, respectively, as described in Materials and methods. a C. violaceum CV026 parallel to itself as negative control; b C. violaceum CV026 parallel to C. violaceum 31532 as positive control; c C. violaceum CV026 parallel to bacteria NJ6-3-1; d A. tumefaciens A136 parallel to itself; e A. tumefaciens A136 parallel to A. tumefaciens KYC6; f A. tumefaciens A136 parallel to bacteria NJ6-3-1

Many reports have now shown that AHLs as quorum-sensing AIs could regulate different biological functions in a range of Gram-negative bacteria (Swift et al. 1994; Dong et al. 2002). Furthermore, both interspecies and intraspecies communication systems have been reported to exist in marine bacteria (Surette et al. 1999; Bassler 2002; Federle and Bassler 2003). For example, Vibrio harveyi possesses two quorum sensing systems: AHLs mediate intraspecies quorum sensing and AI-2 mediates interspecies quorum sensing to regulate bioluminescence (Federle and Bassler 2003). In this case study, we observed that the production of antibacterial metabolites from the marine bacterium NJ6-3-1 was induced not only by its DKP metabolites but also by some signals secreted by the exogenous bacterium S. aureus. We postulate that strain NJ6-3-1 probably possesses both interspecies and intraspecies quorum sensing systems. The DKPs mediate the intraspecies quorum sensing while some unknown exogenous metabolite(s) from other bacteria (in this case S. aureus) mediate interspecies quorum sensing to regulate the production of antibacterial metabolites. At the beginning of the twenty-first century, several articles reported that marine bacteria co-cultured with exogenous microorganisms could induce them to metabolize antibiotics. These findings suggested that these techniques could be further explored to improve the efficiency of discovering new antibiotics (Yan et al. 2003; Trischman et al. 2004). We believe our results indicate that bacterium NJ6-3-1 is regulated by an interspecies quorum sensing system, and that this mechanism is essential as a chemical defense mechanism. However, the identities and biochemical properties of the antibacterial substances regulating quorum sensing require further investigation. The exogenous inducing signals also need to be further verified, and continuing studies along these directions are ongoing in our laboratory.