www.biomolther.org
Bioactive Cyclic Dipeptides from a Marine Sponge-Associated Bacterium, Psychrobacter sp.
Huayue LI1, Byung Cheol LEE2, Tae Sung KIM2, Kyung Sook BAE3,Jongki HONG4, Sang Ho CHOI5, Baoquan BAO1, and Jee Hyung JUNG1*
1 College of Pharmacy, Pusan National University, Busan, Republic of Korea
2 School of Life Sciences and Biotechnology, Korea University, Seoul, Republic of Korea
3 Korea Research Institute of Bioscience and Biotechnology, Daejon, Republic of Korea
4 College of Pharmacy, Kyung Hee University, Seoul, Republic of Korea
5 Department of Agricultural Biotechnology and Center for Agricultural Biomaterials, Seoul National University, Seoul, Republic of Korea
(Received October 12, 2008; Revised November 5, 2008; Accepted November 6, 2008)
Abstract−A bacterial strain with good antibacterial activities against Staphylococus aureus and Escherichia coli was isolated from a marine sponge Stelleta sp., and it was identified as a Psychrobacter sp. by compara- tive 16S rDNA sequence analysis. In our search for bioactive secondary metabolites from this psychrophillic and halotolerent bacterium, sixteen cyclic dipeptides (1-16) were isolated and their structures were identified on the basis of NMR analysis. In the test of the compounds for the protective effect against Vibrio vulnificus- induced cytotoxicity in human intestinal epithelial cells, cyclo-(L-Pro-L-Phe) (5) exhibited significant protective effect. Compounds 2, 6, and 11, which contain D-amino acid, were first isolated from bacteria.
Keywords: Sponge-associated, Psychrobacter sp., Cyclic dipeptide, D-proline, Vibrio vulnificus-induced cytotox- icity
INTRODUCTION
Recently, much attention has been given to marine sponge-associated microorganisms, which have emerged to be a rich source of novel and biologically active secondary metabolites (Piel, 2004). As a filter feeder, sponge is able to filter thousands of liters of water per day, which makes sponge harbor a large number of diverse microorganisms in its tissue. Generally, for microbe-abundant sponge, microorganisms can contrib- ute up to 40-70% of the sponge body volume, exceeding microorganisms in seawater by 2-4 orders of magnitude (Hentschel et al., 2006; Li et al., 2007). In almost all cases, development of sponge-derived drug leads is seri- ously hampered by the environmental and technical prob- lems associated with collecting or cultivating large amounts of animals. The possible existence of a bacte- rial symbiont, which can produce the relevant bioactive molecules, is therefore especially intriguing, because a
sustainable source of sponge-derived drug candidates could be generated by establishing a symbiont culture or by transferring symbiont biosynthetic genes into cultura- ble bacteria (Piel, 2004).
Psychrobacter sp., belonging to Gram-negative gamma-proteobacteria, has been isolated from several kinds of marine sponges including Stelletta tenui, Hali- chondria rugosa, and Dysidea avara (Li et al., 2007).
Members of the genus Psychrobacter, discovered at both poles, deep sea waters, and other diverse marine envi- ronments, are psychrotolerant or psychrophilic and halo- tolerant (Bowman et al., 1996). In spite of their diversity and abundance, the analysis of the physiological and bio- chemical adaptations of this kind of bacteria is still frag- mented. To the best of our knowledge, this is the first report on the investigation of biologically active metabo- lites from a bacterium Psychrobacter sp.
MATERIALS AND METHODS
Bacterial isolation and culture conditions
The sponge Stelletta sp. was collected by hand using SCUBA (20 m depth) in 2005 off the coast of Geoje
*Corresponding author
Tel: +82-51-510-2803, Fax: +82-51-513-6754 E-mail: jhjung@pusan.ac.kr
DOI: 10.4062/biomolther.2008.16.4.356
Island, Korea. Following a rinse with sterile sea water, small pieces (1×1 cm3) of the surface and inner tissue of the sponge were homogenized and then inoculated on Zobell 2216 marine agar (5 g/L of peptone, 1 g/L of yeast extract, 1 mg/L of FePO4·4H2O, and 15 g/L of agar in 75% aged sea water; pH was adjusted to 7.4 with 1 M NaOH solution) at 25oC for 3 days. By this process, 12 pure bacterial strains (J05B1-1~J05B1-12) were isolated from the sponge Stelletta sp. Among these 12 bacterial strains, the EtOAc extract of J05B1-11 (20 mg/mL, 10 µL) showed the highest inhibitory effect against Staphylococ- cus aureus 503 (ATCC) (inhibition zone: 16 mm) and Escherichia coli TEM (ATCC) (inhibition zone: 8 mm) in radial diffusion assay (Lehrer et al., 1991).
Single colonies of the bacterium J05B1-11 were subcul- tured in 500 mL Erlenmeyer flasks each containing 200 mL of Zobell 2216 broth medium (5 g/L of peptone, 1 g/L of yeast extract, 1 mg/L of FePO4·4H2O in 75% aged sea water; pH 7.4) and incubated at 25oC for 4 days on a rotary shaker (160 rpm). To provide cultures for metabo- lite isolation, 8 mL of a subculture was transferred to a 2 L Erlenmeyer flask containing 800 mL of the same medium, and incubated for 7 days under similar condition.
Identification of the bacterium
Genomic DNA extraction and PCR amplification of the 16S rRNA gene were conducted using the method previ- ously described by Kim et al. (1998). The PCR products were purified using a HiYieldTM Gel/PCR DNA extraction kit (Real Biotech Corp.) and sequence analysis was per- formed according to procedures demonstrated by Kim et al. (1998). The 16S rDNA sequence was aligned with those of representative members of the genus Psychro- bacter on the basis of similarities in the primary and sec- ondary RNA structures using PHYDIT version 3.1 (http://
plaza.snu.ac.kr/~jchun/phydit/).
Bioactivity-guided isolation of metabolites
The combined culture broth (80 L) was extracted with EtOAc at room temperature, and the EtOAc extract (4.5 g) was partitioned between 90% MeOH and n-hexane to yield 3.3 g and 1.0 g of residues, respectively. The aque- ous MeOH layer, which exhibited good antibacterial prop- erties against Staphylococcus aureus 503 (ATCC) (inhibition zone: 15 mm) and Escherichia coli TEM (ATCC) (inhibition zone: 5 mm), was subjected to a stepped-gradient MPLC (ODS-A, 120 Å, S-30/50 mesh) eluting with 20-100% MeOH to afford 18 fractions. HPLC was performed on a Gilson 370 pump with a YMC ODS- H80 column (250×10 mm i.d., 4 µm, 80 Å) using RI-101
detector. Compounds 1, 2, 4, 7, 8, 9, and 12-14 were obtained by separation of fraction 2 (inhibition zone: 11 mm against S. aureus; 8 mm against E. coli) on a reversed-phase HPLC (YMC ODS-H80 column) eluting with 23% aqueous MeOH at the flow rate of 1.3 mL/min.
Compounds 3, 5, 6, 10, and 11 were obtained by separa- tion of fraction 4 (inhibition zone: 15 mm against S.
aureus; 6 mm against E. coli) eluting with 40% MeOH (1.3 mL/min). Compounds 15 and 16 were obtained by separation of fraction 5 (inhibition zone: 23 mm against S.
aureus; 12 mm against E. coli) eluting with 50% MeOH (1 mL/min).
Structure elucidation
The 1H and 2D NMR (HSQC, HMBC, COSY, NOESY, and TOCSY) spectra were recorded at 500 MHz using Varian INOVA 500 spectrometer, and 13C NMR spectra were recorded on Varian UNITY 400 spectrometer.
Chemical shifts were reported with reference to the respective solvent peaks and residual solvent peaks (δH
3.30 and δC 49.0 for CD3OD; δH 7.28 and δC 77.2 for CDCl3). Optical rotations were measured using a JASCO P-1020 polarimeter.
Stereochemistry of compound 2 was defined by chemi- cal reaction method reported by Barrow and Sun (1994).
Cyclo-(L-Pro-L-Val) (1) (2 mg) was dissolved in NaOH (0.1 N, 1 mL) and the solution was left at room tempera- ture for 4 h. The solution was extracted with EtOAc more than three times and the organic layer was washed with dilute HCl followed by H2O. The solvent was removed under nitrogen gas and then cyclo-(D-Pro-L-Val) was obtained as a white amorphous solid (1.3 mg). The 1H NMR of the reaction product was compared with that of compound 2. Stereochemistry of other compounds was confirmed by comparing with the previously reported data.
V. vulnificus and human intestinal epithelial INT-407 cells
V. vulnificus strain MO6:24/O used in this study was isolated from patients. For the infection experiments, the bacterial strain was grown to log phase at 30oC in Luria- Bertani medium supplemented with 2.0% (w/v) NaCl (LBS). Bacterial concentrations were then confirmed via viable cell counts on LBS agar. The human intestinal epi- thelial cell-line (INT-407) was purchased from the Ameri- can Type Culture Collection (ATCC, Manassas, VA, USA), and maintained at 37oC in 5% CO2 in Minimum Essential Medium (MEM) supplemented with 10% fetal bovine serum (Gibco BRL, Grand Island, NY, USA) and antibiot- ics (10 unit/mL penicillin G and 10 g/mL streptomycin).
Infection of the INT-407 cells
Human intestinal epithelial cells (INT-407) were grown in culture T-75 flasks at 37oC in a 5% CO2 incubator.
Then, the INT-407 cells were seeded onto 96-well plates and cultured for 24 h in antibiotic-free growth medium.
Prior to infection, the bacteria were centrifuged for 3 min at 2,500×g, resuspended, and then adjusted to 6×108 CFU/mL in antibiotic and phenol red-free MEM medium.
INT-407 cells were incubated with V. vulnificus or cyclic dipeptide (3-5, 8, 9)-mixed V. vulnificus at a multiplicity of infection (MOI) of 10 (number of bacteria/number of epi- thelial cells = 10:1) in a 5% CO2 incubator at 37oC.
Cytotoxicity assay (LDH release assay)
The cytotoxicity was determined by measuring the activity of lactate dehydrogenase (LDH) in the superna- tant using a cytotoxicity detection kit (Roche, Mannheim, Germany) and expressed using the total LDH activity of the cells that were completely lysed by 1% Triton X-100 as 100% (Lee et al., 2008).
Morphological study
INT-407 (3×105 cells/well) cells were incubated with V.
vulnificus or cyclo-(L-Pro-L-Phe) (5) (50 µg/mL)-mixed V.
vulnificus in a 12-well plate for 3 h at an MOI of 10. The culture plates were then centrifuged at 3,000×g and washed twice with pre-warmed PBS (pH 7.4), and fixed with 4% para-formaldehyde (Sigma) for 10 min at room temperature. Then, the cells were washed twice with PBS, after which they were stained with Giemsa solution (Molecular Probe) for 1 h at room temperature. To deter- mine the cytotoxic effect of cyclo-(L-Pro-L-Phe) (5) on INT-407 cells, the cells treated with this compound (50 µg/mL) was compared with the blank (INT-407 cells in medium). Images of specimens were then acquired using a microscope (Olympus IX 71, Japan).
Bacterial growth test
Values of the bacterial growth were determined by a broth microdilution assay. Briefly, various dilutions of the cyclo-(L-Pro-L-Phe) (5) were prepared in an appropriate 2% NaCl LB culture medium (100 µL) in sterile 96-well round-bottom polystyrene microtiter plates. The V. vulnifi- cus (6×108 CFU/mL) was diluted in the same medium to 2×105 CFU/100 µL. Each well of the microtiter plate then received 2×105 CFU /100 µL of the inoculated medium, and the plates were incubated at 37oC for up 1 to 13 h.
The bacterial growth was determined by microplate reader at 540 nm.
RESULTS
Identification of the bacterium
Comparative analyses of the 16S rDNA sequence (1,390 nt) of strain J05B1-11 with those of representative members of the major lines of descent within the domain Bacteria revealed that the isolate was phylogenetically affiliated to the genus Psychrobacter. The 16S rRNA gene sequence of strain J05B1-11 showed the highest similarity to that of P. nivimaris 88/2-7T (98.7%) (Heuchert et al., 2006); the next highest similarity was observed in P. proteolyticus 116T (98.6%) (Denner et al., 2001) and P.
namhaensis SW-242 T (98.5%) (Yoon et al., 2005).
Structure elucidation
The MPLC fractions 2, 4 and 5 were separated repeat- edly by reversed-phase HPLC to give sixteen pure com- pounds. The structures of compounds 1-16 were elucidated by means of spectroscopic data (1D-NMR, 2D-NMR) analysis, and by comparing with previously reported data. The following cyclic dipeptides were iso- lated: cyclo-(L-Pro-L-Val) (1) (5.4 mg) (Schmitz et al., 1983; Jayatilake et al., 1996; Li et al., 2006), cyclo-(D- Pro-L-Val) (2) (1.2 mg) (Barrow and Sun, 1994), cyclo-(L- Pro-L-Leu) (3) (11.0 mg) (Jayatilake et al., 1996; Yan et al., 2004; Li et al., 2006), cyclo-(trans-4-hydroxy-L-Pro-L- Leu) (4) (66.0 mg) (Mitova et al., 2004), cyclo-(L-Pro-L- Phe) (5) (17.3 mg) (Jayatilake et al., 1996; Ström et al., 2002; Li et al., 2006), cyclo-(D-Pro-L-Phe) (6) (5.4 mg) (Barrow and Sun, 1994), cyclo-(trans-4-hydroxy-L-Pro-L- Phe) (7) (9.3 mg) (Ström et al., 2002), cyclo-(cis-hydroxy- D-Pro-L-Phe) (8) (8.9 mg) (Shigemori et al., 1998), cyclo- (L-Pro-L-Tyr) (9) (6.2 mg) (Rudi and Kashman, 1994 ; Jayatilake et al., 1996), cyclo-(L-Pro-L-Trp) (10) (4.2 mg) (Li et al., 2006), cyclo-(D-Pro-L-Trp) (11) (1.7 mg) (Sanz- Cervera et al., 2000), cyclo-(L-Ala-L-Leu) (12) (1.0 mg) (Stark and Hofmann, 2005), cyclo-(L-Ala-L-Ile) (13) (2.7 mg) (Stark and Hofmann, 2005), cyclo-(L-Phe-Gly) (14) (1.1 mg) (Kanoh et al., 1999), cyclo-(L-Phe-L-Val) (15) (1.1 mg) (Stark and Hofmann, 2005), and cyclo-(L-Phe-L- Leu) (16) (0.9 mg) (Tullberg et al., 2006). Among these, compounds 2, 6, 8, and 11 contained D-proline. Accord- ing to 1D and 2D NMR data, the planar structure of com- pound 2 was the same with compound 1, however, 1H chemical shift of compound 2 was different from that of compound 1, which indicated that they were stereoiso- mers. The stereochemistry of compound 2 was defined by conversion of the common cyclo-(L-Pro-L-Val) to cyclo-(D-Pro-L-Val) by chemical reaction. The 1H NMR
data of compound 2 and the synthetic cyclo-(D-Pro-L-Val) were identical. Therefore, compound 2 was defined as cyclo-(D-Pro-L-Val). As far as we know, only one literature (Kwon et al., 2001) has reported the 1H and 13C NMR data of cyclo-(D-Pro-L-Val), while it was not match with the data in this study, which seem to correspond more possibly with the L,L- isomer by comparison with previously reported data of cyclo-(L-Pro-L-Val) (Schmitz et al., 1983; Jayatilake et al., 1996; Li et al., 2006).
In the 1H NMR spectra of the cyclic dipeptides except 2, 6, 8, and 11, we observed that H-3 and H-6 were cou- pled with each other, with coupling constant value smaller than 2 Hz. This transannular long-range proton coupling was clearly observed in CD3OD and could be attributed to homoallylic coupling 5JHH in the diketopiperazine ring.
Lone-pair electrons of nitrogens and oxygens of the pep- tide bonds will participate in resonance equilibrium, and the pseudo-double bond character of the peptide bond is capable of transmitting the spin information from H-3 to H-6. Davies et al. (1976) mentioned that the observation of 5JHH could be used to determine the cis and/or trans- conformations of the peptide bond. From this observa- tion, we are able to further confirm the stereochemistry of D-proline containing compounds 2, 6, 8, and 11, which showed no long-range proton coupling because of their trans-conformations. This is the first report for D-amino acid containing compounds 2, 6, and 11 from bacteria.
Cyclo-(L-Pro-L-Val) (1) – colorless, crystalline solid;
[α]28D = -110 (c 0.48, MeOH).
Cyclo-(D-Pro-L-Val) (2) – white, amorphous solid;
[α]28D = +25 (c 0.12, MeOH). 1H NMR (CDCl3, 500 MHz) δ 6.21 (NH, bs), 4.10 (1H, dd, J = 9.5, 6.5 Hz, H-6), 3.74 (1H, bs, H-3), 3.68-3.75 (1H, m, H-9a), 3.50-3.55 (1H, m, H-9b), 2.39-2.44 (1H, m, H-7a), 2.21-2.27 (1H, m, H-10), 2.01-2.07 (1H, m, H-8a), 1.85-1.97 (2H, m, H-7b and H- 8b), 1.05 (3H, d, J = 7.0 Hz, H-12), 1.00 (3H, d, J = 7.0 Hz, H-11); 13C NMR (CDCl3, 100 MHz) δ 169.5 (C-1), 165.4 (C-4), 63.8 (C-3), 58.5 (C-6), 45.8 (C-9), 33.3 (C- 10), 29.6 (C-7), 22.1 (C-8), 19.2 (C-12), 17.7 (C-11).
Effects on Vibrio vulnificus-induced cytotoxicity in human intestinal epithelial cells
To determine the protective effect of several cyclic dipeptides (3-5, 8, 9) against V. vulnificus-induced cyto- toxicity in human intestinal epithelial cells (INT-407), the levels of lactate dehydrogenase (LDH) release were measured. As shown in Fig. 2A, V. vulnificus highly induced the release of LDH from the INT-407 cells (76.10
± 2.46 %) when incubated at MOI 10 for 3 h, while in cyclic dipeptide (3-5, 8, 9)-treated cells, LDH release was reduced. Especially, cyclo-(L-Pro-L-Phe) (5) significantly abolished the release of LDH (15.34 ± 2.46 %). The level of LDH release from the cells treated with V. vulnificus and cyclo-(L-Pro-L-Phe) (5) (100 µg/mL) was almost 7- fold less than that of the control (Fig. 2B). As shown in Fig. 2C, the release of LDH in V. vulnificus-infected INT-
Fig. 1. Structures of the cyclic dipeptides (1-16) isolated from a sponge-associated bacterium Psychrobacter sp.
407 cells approached 100% at the incubation time of 4 h, while the cells infected with cyclo-(L-Pro-L-Phe) (5) treated-V. vulnificus kept significantly reduced LDH val- ues for as long as 4 h at 50 or 100 µg/mL, which sug- gested that the protective effect of cyclo-(L-Pro-L-Phe) (5) was not decreased with time dependency. Morphological studies were also carried out using INT-407 cells. As shown in Fig. 2D, many INT-407 cells exhibited marked cellular damage, such as cytoplasmic loss and typical phenotypes of cell death after infection with V. vulnificus (D-1). In contrast, fewer damages were observed when incubated with cyclo-(L-Pro-L-Phe) (5) (50 µg/mL) and V.
vulnificus (D-2). The cells exhibited less-damaged sur-
face and less cytoplasmic loss. And also, cyclo-(L-Pro-L- Phe) (5) (50 µg/mL) itself had no cytotoxic effect in human intestinal epithelial cells (D-4) as compared to blank control (D-3). In the test for the effect of cyclo-(L- Pro-L-Phe) (5) on the growth of V. vulnificus, the com- pound inhibited the growth of V. vulnificus for as long as 13 h (Fig. 3A), and the inhibition effect was increased dose-dependently (20-100 µg/mL) (Fig. 3B).
DISCUSSION
Despite wide distribution and extreme environmental tolerance, few reports described bioactive secondary Fig. 2. Effects of cyclo-(L-Pro-L-Phe) (5) against V. vulnificus MO6:24/O-induced cytotoxicity to human intestinal epithelial cells (INT- 407). INT-407 cells in medium were used as a blank. (A) INT-407 cells were infected with the V. vulnificus or cyclic dipeptide (3-5, 8, 9) mixed-V. vulnificus at a multiplicity of infection (MOI) of 10 (number of bacteria/number of epithelial cells = 10:1) for 3 h, and the cytotoxicity was determined by LDH release. (B) INT-407 cells were infected with V. vulnificus or cyclo-(L-Pro-L-Phe) (5) dose dependency mixed-V. vulnificus at MOI 10 for 3 h. (C) INT-407 cells were infected with V. vulnificus or cyclo-(L-Pro-L-Phe) (5) mixed- V. vulnificus at MOI 10 for various times. All of the data represent the means ± standard errors (n = 3). *, P < 0.01, **, P < 0.001, rel- ative to the groups infected with V. vulnificus at MOI 10 for 3 h (A, B) or for various times (C). (D) Microscopic observation of INT-407 cells. [D-1, INT cells infected with V. vulnificus at MOI 10 for 3 h; D-2, INT cells infected with cyclo-(L-Pro-L-Phe) (5) (50 µg/mL)- mixed V. vulnificus; D-3, blank; D-4, INT cells treated with cyclo-(L-Pro-L-Phe) (5) (50 µg/mL)]
metabolites of the psychrophilic bacterium Psychrobacter sp. In this study, sixteen (1-16) cyclic dipeptides were iso- lated from a Psychrobacter sp., derived from a marine sponge Stelletta sp. Cyclic dipeptides possess diverse biological activities such as antitumor (Nicholson et al., 2006), antifungal (Houston et al., 2004), antibacterial (Fdhila et al., 2003), and antihyperglycaemic (Song et al., 2003) activities. Due to their chiral, rigid, and functional- ized structures, they bind to a large variety of receptors with high affinity, giving a broad range of biological activi- ties (Martins and Carvalho, 2007). Therefore, cyclic dipeptides are attractive structures for the discovery of new lead compounds by combinatorial chemistry, and are considered ideal for the rational development of new ther- apeutic agents (Martins and Carvalho, 2007). More than 90% of Gram-negative bacteria produce cyclic dipep- tides (Mitova et al., 2005), however, their origins, argu- ably from culture media, have been suspected for dozens of years (Lautru et al., 2002). Recently, it has been proved that sterile media do not contain cyclic dipeptides (Mitova et al., 2004; Li et al., 2005). In addition, gene clusters for the biosynthesis of cyclic dipeptides were identified in some bacterial strains (Lautru et al., 2002;
Sioud et al., 2007; Schultz et al., 2008). These evi- dences proved that cyclic dipeptides are biosynthesized by bacteria under the control of some specific genes, not obtained artificially from the culture media or separation process.
Among the isolated cyclic dipeptides, cyclo-(D-Pro-L- Val) (2), cyclo-(D-Pro-L-Phe) (6), cyclo-(cis-hydroxy-D-Pro- L-Phe) (8), and cyclo-(D-Pro-L-Trp) (11) were composed of D-amino acids. In most cases, proteins or peptides in nature are comprised solely of L-amino acids. D-amino acids, the enantiomer of L-amino acids, are not common in eukaryotic cells. However, in bacteria, there are abun-
dant D-amino acids, originating from cell wall peptidogly- can. D-amino acids have been found to constitute 5 and 15% of peptidoglycan amino acids in a Gram-negative and a Gram-positive bacterium, respectively (Jørgensen et al., 2006). It seems that D-amino acids widely distrib- uted among eubacteria have been eliminated from eukaryotes during the evolution (Nagata et al., 1998).
Recently, D-amino acids have also been identified in eukaryotic organisms such as fungi (Lee et al., 1977), algae (Brückner et al., 1994), and mammals (Hamase et al., 2006). And it was assumed that microbial enzymes, in particular racemases and epimerases, are responsible for the formation of D-amino acids (Brückner et al., 1993;
Gruenewald et al. 2004). As D-proline was recognized as a sole D-amino acid of the cyclic dipeptides from this bac- terium, we presume that an epimerase just responsible for proline epimerization might exist in this Psychrobacter sp.
In our search for the compounds exhibiting protective effect against V. vulnificus-induced cytotoxicity in human intestinal epithelial cells (INT-407), compounds 3-5, 8, and 9 were used in the test. The pathogenic marine bac- terium V. vulnificus is a causative agent of food-borne dis- eases, such as life-threatening septicemia and possibly gastroenteritis, in individuals with underlying predispos- ing conditions (Strom, et al., 2000; Park, et al., 2006). As a result, we found that cyclo-(L-Pro-L-Phe) (5) (50 µg/mL) significantly reduce V. vulnificus-induced cytotoxicity as gauged by the release of LDH from INT-407 cells (Fig.
2AC). In addition, we also clearly observed the protective effect of cyclo-(L-Pro-L-Phe) (5) against V. vulnificus- induced cell damage by morphological study (Fig. 2D). In a subsequent test for the effect of cyclo-(L-Pro-L-Phe) (5) on V. vulnificus growth, we found that this compound inhibit the growth of V. vulnificus in a dose-dependent manner (20-100 µg/mL) (Fig. 3B). Therefore, the protec- Fig. 3. Effect of cyclo-(L-Pro-L-Phe) (5) on the growth of V. vulnificus MO6:24/O.
tion of cyclo-(L-Pro-L-Phe) (5) against V. vulnificus- induced cytotoxicity in human intestinal epithelial cells might be due to its antibacterial effect.
As quorum-sensing in bacterial cells is becoming a ris- ing focus of the research for cyclic dipeptides produced by bacteria, we examined the inhibition effect of cyclic dipep- tides (3-5, 8, 9) against SmcR, which is an important glo- bal regulator for genes associated with the survival and pathogenesis of V. vulnificus (Lee et al., 2007), but none of them exhibited inhibitory activity (data not shown). Con- sidering that cyclic dipeptides exhibit diverse biological and pharmacological activities in higher plants and ani- mals (Martins and Carvalho, 2007), it is quite plausible that cyclic dipeptides do not function as bacteria-to-bacte- ria signaling molecules, but they might have a role in mod- ulating prokaryotic-eukaryotic interactions (Holden et al., 2000). Understanding the functional roles of cyclic dipep- tides in prokaryotic-eukaryotic interactions will greatly assist the research on symbiotic interactions between marine sponges and their bacterial communities.
In summary, we isolated sixteen cyclic peptides (1-16) from a sponge-associated psychrophilic bacterium Psy- chrobacter sp, among which, compounds 2, 6, and 11 were first obtained from bacteria. And cyclo-(L-Pro-L-Phe) (5) exhibited significant protective effect against V. vulnifi- cus-induced cytotoxicity in human intestinal epithelial cells (INT-407). These cyclic dipeptides, having a variety of pharmaceutical potential, may play important roles in sponge-bacteria interaction to make their co-beneficial symbiotic relationship. Additionally, a better understanding of Psychrobacter's metabolic products may aid in research on potential extraterrestrial cryo-dwelling organisms.
ACKNOWLEDGMENTS
This research was supported by a grant from Marine Biotechnology Program funded by Ministry of Land, Transport, and Maritime Affairs, Korea.
REFERENCES
Barrow, C. J. and Sun, H. H. (1994). Spiroquinazoline, a novel substance P inhibitor with a new carbon skeleton, isolated from Aspergillus flavipes. J. Nat. Prod. 57, 471-476.
Bowman, J. P., Cavanagh, J., Austin, J. J. and Sanderson, K.
(1996). Novel Psychrobacter species from Antarctic ornitho- genic soils. Int. J. Syst. Bacteriol. 46, 841-848.
Brückner, H., Becker, D. and Lüpke, M. (1993). Chirality of amino acids of microorganisms used in food biotechnology.
Chirality 5, 385-392.
Brückner, H., Haasmann, S., Langer, M., Westhauser, T., Witt- ner, R. and Godel, H. (1994). Liquid-chromatographic
determination of D-amino and L-amino acids by derivatization with o-phthaldialdehyde and chiral thiols. J. Chromatogr. A 666, 259-273.
Davies, D. B. and Khaled, M. A. (1976). Conformations of peptides in solution by nuclear magnetic resonance spectroscopy. Part II. Homoallylic coupling in cyclic dipeptides. J. Chem. Soc. Perkin Trans. 2, 187-196.
Denner, E. B. M., Mark, B., Busse, H. J., Turkiewicz, M. and Lubitz, W. (2001). Psychrobacter proteolyticus sp. nov., a psy- chrotrophic, halotolerant bacterium isolated from the Antarc- tic krill Euphausia superba Dana, excreting a cold-adapted metalloprotease. Syst. Appl. Microbiol. 24, 44-53.
Fdhila, F., Vázquez, V., Sánchez, J. L. and Riguera, R. (2003).
DD-Diketopiperazines: antibiotics active against Vibrio anguil- larum isolated from marine bacteria associated with cultures of Pecten maximus. J. Nat. Prod. 66, 1299-1301.
Gruenewald, S., Mootz, H. D., Stehmeier, P. and Stachelhaus, T.
(2004). In vivo production of artificial nonribosomal peptide products in the heterologous host Escherichia coli. Appl. Envi- ron. Microb. 70, 3282-3291.
Hamase, K., Takagi, S., Morikawa, A., Konno, R., Niwa, A. and K. Zaitsu. (2006). Presence and origin of large amounts of D- proline in the urine of mutant mice lacking D-amino acid oxi- dase activity. Anal. Bioanal. Chem. 386, 705-711.
Hentschel, U., Usher, K. M. and Taylor, M. W. (2006). Marine sponges as microbial fermenters. FEMS. Microbial. Ecol. 55, 167-177.
Heuchert, A., Glöckner, F. O., Amann, R. and Fischer, U. (2004).
Psychrobacter nivimaris sp. nov., a heterotrophic bacterium attached to organic particles isolated from the south Atlantic (Antarctica). Syst. Appl. Microbiol. 27, 399-406.
Holden, M., Swift, S. and Williams, P. (2000). New signal mole- cules on the quorum-sensing block. Trends Microbiol. 8, 101- 103.
Houston, D. R., Synstad, B., Eijsink, V. G. H., Stark, M. J. R., Eggleston, I. M. and Aalten, D. M. F. (2004). Structure-based exploration of cyclic dipeptide chitinase inhibitors. J. Med.
Chem. 47, 5713-5720.
Jayatilake, G. S., Thornton, M. P., Leonard, A. C., Grimwade, J.
E. and Baker, B. J. (1996). Metabolites from an Antarctic sponge-associated bacterium, Pseudomonas aeruginosa. J.
Nat. Prod. 59, 293-296.
Jørgensen, N. O. G. and Middelboe, M. (2006). Occurrence and bacterial cycling of D amino acid isomers in an estuarine environment. Biogeochemistry 81, 77-94.
Kanoh, K., Kohno, S., Katada, J., Takahashi, J., Uno, I. and Hayashi, Y. (1999). Synthesis and biological activities of phe- nylahistin derivatives. Bioorgan. Med. Chem. 7, 1451-1457.
Kim, S. B., Falconer, C., Williams, E. and Goodfellow, M. (1998).
Streptomyces thermocarboxydovorans sp. nov. and Strepto- myces thermocarboxydus sp. nov., two moderately thermo- philic carboxydotrophic species from soil. Int. J. Syst.
Bacteriol. 48, 59-68.
Kwon, O. S., Park, S. H., Yun, B. S., Pyun, Y. R. and Kim, C. J.
(2001). Cyclo(D-Pro-L-Val), a specific â-glucosidase inhibitor produced by Aspergillus sp. F70609. J. Antibiot. 54, 179-181.
Lautru, S., Gondry, M., Genet, R. and Pernodet, J. L. (2002).
The albonoursin gene cluster of S. noursei: biosynthesis of
diketopiperazine metabolites independent of nonribosomal peptide synthetases. Chem. Biol. 9, 1355-1364.
Lee, B. C., Lee, J. H., Kim, M. W., Kim, B. S., Oh, M. H., Kim, K.
S., Kim, T. S. and Choi, S. H. (2008). Vibrio vulnificus rtxE is important for virulence, and its expression is induced by expo- sure to host cells. Infect. Immun. 76, 1509-1517.
Lee, C. and Bada, J. L. (1977). Dissolved amino acids in the equatorial Pacific, the Sargasso Sea, and Biscayne Bay. Lim- nol. Oceanogr. 22, 502-510.
Lee, J. H., Rhee, J. E., Park, U., Ju, H. M., Lee, B. C., Kim, T.
S., Jeong, H. S. and Choi, S. H. (2007). Identification and functional analysis of Vibrio vulnificus SmcR, a novel global regulator. J. Microbiol. Biotechnol. 17, 325-334.
Lehrer, R. I., Rosenman, M., Harwig, S. S. S. L., Jackson, R.
and Eisenhauer, P. (1991). Ultrasensitive assays for endoge- nous antimicrobial activity. J. Immunol. Methods 137, 167- 173.
Li, X., Bobretsov, S., Xu, Y., Xiao, X., Hung, O. S. and Qian, P.
Y. (2006). Antifouling diketopiperazines produced by a deep- sea bacterium, Streptomyces fungicidicus. Biofouling 22, 201- 208.
Li, Z., Hu, Y., Liu, Y., Huang, Y., He, L. and Miao, X. (2007). 16S rRNA clone library-based bacterial phylogenetic diversity associated with three South China Sea sponges. World J.
Microbiol. Biotechnol. 23, 1265-1272.
Li, Z., Yan, X., Xu, J., Chen, H. and Lin, W. (2005). Hymeniaci- don perleve associated bioactive bacterium Pseudomonas sp. NJ6-3-1. Appl. Biochem. Micro. 41, 29-33.
Martins, M. B. and Carvalho, I. (2007). Diketopiperazines: bio- logical activity and synthesis. Tetrahedron 63, 9923-9932.
Mitova, M., Tommonaro, G., Hentschel, U., Müller, W. E. G. and Rosa, S. D. (2004). Exocellular cyclic dipeptides from a Rue- geria strain associated with cell cultures of Suberites domun- cula. Mar. Biotechnol. 6, 95-103.
Mitova, M., Tutino, M. L., Infusini, G., Marino, G. and Rosa, S. D.
(2005). Exocellular peptides from Antarctic psychrophile Pseudoalteromonas haloplanktis. Mar. Biotechnol. 7, 523- 531.
Nagata, Y., Fujiwara, T., Kawaguchi-Nagata, K., Fukumori, Y.
and Yamanaka, T. (1998). Occurrence of peptidyl D-amino acids in soluble fractions of several eubacteria, archaea and eukaryotes. Biochim. Biophys. Acta. 1379, 76-82.
Nicholson, B., Lloyd, G. K., Miller, B. R., Palladino, M. A., Kiso, Y., Hayashi, Y. and Neuteboom, S. T. C. (2006). NPI-2358 is a tubulin-depolymerizing agent: in-vitro evidence for activity as a tumor vascular-disrupting agent. Anti-Cancer Drug 17, 25-31.
Park, N. Y., Lee, J. H., Kim, M. W., Jeong, H. G., Lee, B. C., Kim, T. S. and Choi, S. H. (2006). Identification of the Vibrio vulnificus wbpP gene and evaluation of its role in virulence.
Infect. Immun. 74, 721-728.
Piel, J. (2004). Metabolites from symbiotic bacteria. Nat. Prod.
Rep. 21, 519-538.
Rudi, A. and Kashman, Y. (1994). Amino acid derivatives from the marine sponge Jaspis digonoxea. J. Nat. Prod. 57, 829- 832.
Sanz-Cervera, J. F., Stocking, E. M., Usui, T., Osada, H. and Williams, R. M. (2000). Synthesis and evaluation of microtu- bule assembly inhibition and cytotoxicity of prenylated deriva- tives of cyclo-L-Trp-L-Pro. Bioorgan. Med. Chem. 8, 2407- 2415.
Schmitz, F. J., Vanderah, D. J., Hollenbeak, K. H., Enwall, C. E.
L., Gopichand, Y., SenGupta, P. K., Hossain, M. B. and Helm, D. (1983). Metabolites from the marine sponge Tedania ignis.
A new atisanediol and several known diketopiperazines. J.
Org. Chem. 48, 3941-3945.
Schultz, A. W., Oh, D. C., Carney, J. R., Williamson, R. T., Udwary, D. W., Jensen, P. R., Gould, S. J., Fenical, W. and Moore, B. S. (2008). Biosynthesis and structures of cyclomar- ins and cyclomarazines, prenylated cyclic peptides of marine actinobacterial origin. J. Am. Chem. Soc. 130, 4507-4516.
Shigemori, H., Tenma, M., Shimazaki, K. and Kobayashi, J.
(1998). Three new metabolites from the marine yeast Aure- obasidium pullulans. J. Nat. Prod. 61, 696-698.
Sioud, S., Karray-Rebai, I., Aouissaoui, H., Aigle, B., Bejar, S.
and Mellouli, L. (2007). Targeted gene disruption of the cyclo (L-Phe, L-Pro) biosynthetic pathway in Streptomyces sp.
US24 strain. J. Biomed. Biotechnol. 2007, 1-9.
Song, M. K., Hwang, I. K., Rosenthal, M. J., Harris, D. M., Yamaguchi, D. T., Yip, I. and Go, V. L. W. (2003). Anti-hyper- glycemic activity of zinc plus cyclo(His-Pro) in genetically dia- betic goto-kakizaki and aged rats. Exp. Biol. Med. 228, 1338- 1345.
Stark, T. and Hofmann, T. (2005). Structures, sensory activity, and dose/response functions of 2,5-diketopiperazines in roasted cocoa nibs (Theobroma cacao). J. Agric. Food Chem. 53, 7222-7231.
Ström, K., Sjögren, J., Broberg, A. and Schnürer, J. (2002). Lac- tobacillus plantarum MiLAB 393 produces the antifungal cyclic dipeptides cyclo(L-Phe-L-Pro) and cyclo(L-Phe-trans-4- OH-L-Pro) and 3-phenyllactic acid. Appl. Environ. Microb. 68, 4322-4327.
Strom, M. and Paranjpye, R. N. (2000). Epidemiology and pathogenesis of Vibrio vulnificus. Microbes Infect. 2, 177-188.
Tullberg, M., Grøtli, M. and Luthman, K. (2006). Efficient synthe- sis of 2,5-diketopiperazines using microwave assisted heat- ing. Tetrahedron 62, 7484-7491.
Yan, P. S., Song, Y., Sakuno, E., Nakajima, H., Nakagawa, H.
and Kimiko, Y. (2004). Cyclo-(L-leucyl-L-prolyl) produced by Achromobacter xylosoxidans inhibits aflatoxin production by Aspergillus parasiticus. Appl. Environ. Microb. 70, 7466-7473.
Yoon, J. H., Lee, C. H., Yeo, S. H. and Oh, T. K. (2005). Psy- chrobacter aquimaris sp. nov. and Psychrobacter namhaen- sis sp. nov., isolated from sea water of the South Sea in Korea. Int. J. Syst. Evol. Microbiol. 55, 1007-1013.
