|Home | About | Journals | Submit | Contact Us | Français|
Feeding of bacterially encapsulated heat shock proteins (Hsps) to invertebrates is a novel way to limit Vibrio infection. As an example, ingestion of Escherichia coli overproducing prokaryotic Hsps significantly improves survival of gnotobiotically cultured Artemia larvae upon challenge with pathogenic Vibrio campbellii. The relationship between Hsp accumulation and enhanced resistance to infection may involve DnaK, the prokaryotic equivalent to Hsp70, a major molecular chaperone in eukaryotic cells. In support of this proposal, heat-stressed bacterial strains LVS 2 (Bacillus sp.), LVS 3 (Aeromonas hydrophila), LVS 8 (Vibrio sp.), GR 8 (Cytophaga sp.), and GR 10 (Roseobacter sp.) were shown in this work to be more effective than nonheated bacteria in protecting gnotobiotic Artemia larvae against V. campbellii challenge. Immunoprobing of Western blots and quantification by enzyme-linked immunosorbent assay revealed that the amount of DnaK in bacteria and their ability to enhance larval resistance to infection by V. campbellii are correlated. Although the function of DnaK is uncertain, it may improve tolerance to V. campbellii via immune stimulation, a possibility of significance from a fundamental perspective and also because it could be applied in aquaculture, a major method of food production.
The brine shrimp Artemia, found worldwide in extreme environments, is used as a model organism for the study of stress response (Clegg et al. 2000a, b; Frankenberg et al. 2000, MacRae 2003), aquatic diseases (Verschuere et al. 2000; Soto-Rodriguez et al. 2003; Marques et al. 2005; Sung et al. 2007), feed quality (Marques et al. 2004a, b), and probiont characteristics (Marques et al. 2005, 2006a). As one example, a recently established gnotobiotic system, where larvae are hatched, grown in axenic conditions, and exposed to known bacteria, has facilitated research on host–microbe interactions, particularly as they relate to Artemia larvae (Marques et al. 2006b). Among the topics investigated were the contribution of bacteria toward immune stimulation in Artemia, the role of bacteria as a food source for this crustacean, and the influence of bacteria on the nutritional properties of Artemia used as feed in aquaculture (Yasuda and Taga 1980; Hessen and Andersen 1990; Gorospe et al. 1996; Marques et al. 2005).
Physiological stress promotes synthesis of intracellular heat shock proteins (Hsps), often referred to as stress proteins or molecular chaperones, and they influence the synthesis, structure, localization, and function of other cell proteins (Parsell and Lindquist 1993). Members of the Hsp70 family, a prominent group of molecular chaperones, are robustly induced when cells experience temperature variation, oxygen deprivation, nutritional deficiency, and infection (Morimoto 1998; Feder and Hofmann 1999). Hsp70 occurs in all types of eukaryotic organisms and it has approximately 60% similarity to DnaK, the prokaryotic Hsp70 equivalent (Pockley 2003). As an unexpected finding, extracellular Hsp70, in combination with other Hsps, is thought to modulate peptide positioning on cell surfaces, thus enhancing immune recognition of aberrant cells (Athman and Philpott 2004; Johnson and Fleshner 2006). Of interest in this regard is research on economically important organisms used in aquaculture such as Penaeus monodon (de la Vega et al. 2006) and Artemia franciscana (Sung et al. 2007, 2008, 2009), where Hsp70 is implicated in protection against pathogenic microbes, perhaps by stimulation of the immune response.
Aquaculture, an important source of human nutrition, is the fastest-growing food production sector in the world, increasing about 8.8% annually during the past decade and significantly supplementing capture fisheries. With this growth is the expanding use of live food such as Artemia larvae. High nutritional properties, including those contributed by lipids and unsaturated fatty acids, make newly hatched Artemia larvae particularly suitable as a major starter diet in the rearing of farmed finfish and crustacean larvae. However, this advance necessitates rearing Artemia larvae under conditions where infection by bacteria and other pathogens must be controlled. In relation to this requirement, endogenous Hsp70 (Sung et al. 2007, 2008) and exogenously supplied DnaK synthesized in transformed Escherichia coli (Sung et al. 2009) were shown to protect Artemia larvae from Vibrio infection, perhaps by immune stimulation. The objective of the present study was, therefore, to substantiate the observations that bacteria overproducing DnaK protect Artemia larvae from Vibrio infection.
Bacillus sp. (LVS 2), Aeromonas hydrophila (LVS 3), Vibrio sp. (LVS 8), Cytophaga sp. (GR 8), Roseobacter sp. (GR 10), and Vibrio campbellii strain LMG 21363 (Table 1) were stored in 40% glycerol at −80°C. Bacteria were grown at 28°C for 48 h on marine agar and then to log phase in marine broth 2216 (Difco Laboratories, Detroit, MI, USA) by overnight incubation at 28°C. Hsp synthesis was induced in bacteria, with the exception of V. campbellii, by a 30-min exposure to 38°C at a t of 4°C/min in a preheated water bath. The bacterial strains were then transferred individually to sterile tubes, centrifuged at 2,200×g for 15 min at 28°C, and suspended in filtered autoclaved sea water, used throughout the study. The turbidity of bacterial cultures was determined spectrophotometrically at 550 nm and these values were used to determine bacterial numbers according to the McFarland standard (BioMerieux, Marcy L'Etoile, France), assuming that an optical density of 1.000 corresponds to 1.2×109 cells per milliliter.
Axenic Artemia larvae were obtained by cyst decapsulation (Sorgeloos et al. 1986) followed by incubation at 28°C for 24 h in sea water (Sung et al. 2008). Larvae developing within the next 4–6 h, and thus with the ability to ingest bacteria, were harvested and incubated for 6 h with either heat-shocked or nonstressed bacteria at approximately 1×107 cells per milliliter. Challenge tests were then performed by adding pathogenic V. campbellii at 1×107 cells per milliliter and incubating with constant agitation and light for 36 h prior to recovery of swimming larvae, fixation in Lugol's solution, and counting. Survival percentage was calculated as where Nt and No are the final and initial numbers of live larvae, respectively. The increase in survival conferred by each bacterial strain upon heating was calculated as where Sf and Si are the percentage survival of Artemia-fed heated and nonheated bacteria, respectively. Axenity was confirmed by plating 100 µL of culture medium containing Artemia, but receiving no bacteria, on marine agar and incubating for 5 days at 28°C (Marques et al. 2004a). Results were discarded if contamination was found.
The nutritional value of bacterial strains (Table 1) was determined by a single feeding of 1×107 bacteria per milliliter to axenically cultured Artemia. Viability, as survival percentage, was determined after 2 days as previously described. Fixed nauplii were measured by using a dissecting microscope equipped with a drawing mirror, a digital plan measure, and Artemia 1.0® software (courtesy of Marnix Van Domme).
Larval survival (in percent) in challenge and nutritional tests were arcsine-transformed to satisfy normality and homoscedasticity requirements while length measurements were either logarithmic- or square root-transformed as necessary. Differences in larval survival and lengths were investigated by performing analysis of variances and Tukey's multiple comparisons range using the statistical analysis software SPSS® version 11.5 for Windows®.
Protein extraction was as described (Sung et al. 2009). Bacteria were homogenized in the presence of 0.1 mm diameter glass beads in ice-cold buffer K (150 mM sorbitol, 70 mM potassium gluconate, 5 mM MgCl2, 5 mM NaH2PO4, 40 mM HEPES, pH 7.4) (Clegg et al. 2000a) containing protease inhibitors (catalog no. P8465, Sigma-Aldrich, USA) at the highest recommended level. Protein concentrations were determined by the Bradford method and equal amounts of protein were loaded in each gel lane.
Bacterial proteins resolved in 10% sodium dodecyl sulfate polyacrylamide gels (Laemmli 1970) were either stained with Coomassie Biosafe (BioRad™ Laboratories, USA) or transferred to polyvinylidene fluoride membranes (BioRad Immun-Blot™ PVDF, USA) before incubating for 60 min with blocking buffer (phosphate-buffered saline containing 0.2% [v/v] Tween-20 and 5% [w/v] bovine serum albumin). Membranes were incubated for 1 h at room temperature with rabbit polyclonal antibody raised against the ATPase domain of E. coli DnaK (Bucca et al. 2000), a generous gift from Dr. Bernd Bukau, ZMBH, Germany. Subsequent to washing, the membrane was incubated with horseradish peroxidase-conjugated goat antirabbit IgG (Gentaur BVBA, Belgium) for 1 h, washed, and exposed to 0.7 mM diaminobenzidine tetrahydrochloride dihydrate and 0.01% (v/v) H2O2 in 0.1 M Tris–HCl (pH 7.6) to detect antibody-reactive proteins. The 70-kDa protein detected with this antibody in various bacterial strains is subsequently referred to as DnaK.
DnaK was quantified with the Sensiflex™ ELISA Development Kit (Molecular Probes, Invitrogen, USA) following the manufacturer's instructions. Approximately 50 µg of bacterial protein in 100 µL of sodium bicarbonate buffer, pH 9.3, was added to each well of a 96-well round-bottom polystyrene plate (Nunc-Immunoplate Maxisorp, Denmark) and incubated at 4°C overnight. Plates were washed three times with phosphate-buffered saline containing 0.1% Tween-20 and blocked 4 h at 28°C with phosphate-buffered saline containing 5% bovine serum albumin. One hundred microliters of rabbit polyclonal antibody raised against E. coli DnaK was added to each well, and plates were incubated for 30 min at 37°C prior to washing. Goat antirabbit IgG (H+L) with β-lactamase TEM-1 conjugate was then added, and after 30 min, at 28°C, the plates were washed. Detection of antibody reactivity was by incubation for 30 min with 100 µL of 0.9 mM Fluorocilin™ Green reagent. Fluorescence intensity was determined at 495 nm excitation and 525 nm emission in a Tecan® Infinite M200 ELISA microplate reader. A standard curve, constructed by use of recombinant DnaK, was used to convert sample absorbance to DnaK content with values expressed as E. coli equivalent. All experiments were done in duplicate.
Feeding Artemia larvae heat-shocked as opposed to nonheated bacteria significantly (p<0.05) increased survival upon exposure to V. campbellii (Fig. 1). Strain LVS 8, although conferring the lowest protection on Artemia larvae whether or not bacteria were heated, nonetheless, greatly improved protection when heat-shocked bacteria were employed. In contrast, GR 10, which gave the best protection of any strain prior to heat shock, provided the least enhancement of larval protection upon heating. Increases in protection granted by heating strain LVS 3 were between that bestowed by heat shocking LVS 8 and GR 10, whereas the effects of LVS 2 and GR 8 were variable, although both provided significantly increased protection against Vibrio infection upon heating (Table 2).
Coomassie stained gels containing protein extracts from heated and nonheated bacteria of single strains were similar to one another, although LVS 2 exhibited a slight increase in a 70-kDa polypeptide upon heating (Fig. 2a). Immunoprobing of Western blots with an antibody to DnaK revealed a single reactive polypeptide of approximately 70 kDa in all strains, except LVS2 where two 70 kDa polypeptides were detected. Staining intensity of the antibody-reactive 70-kDa polypeptides was greater in protein extracts from heat-shocked versus nonheated bacteria (Fig. 2b).
Additionally, quantification by enzyme-linked immunosorbent assay (ELISA) demonstrated that DnaK increased from 2.0- to 2.3-fold in heated bacteria (Fig. 3). The higher amounts of DnaK in heated as opposed to nonstressed bacteria correlated with enhanced ability to promote survival of Artemia larvae.
When Artemia were starved, mortality was high, even in the absence of Vibrio exposure, with only 9±5% and 13±5% of larvae alive in replicate experiments after 48 h. Conversely, incubating larvae with any of the bacterial strains employed in this study, except V. campbellii, significantly increased the number of live larvae (p<0.05) (Table 3). Bacterial strains varied in nutritional quality. In comparison to LVS 8, feeding with both LVS 2 and LVS 3 substantially increased viable larvae, while moderate but insignificant increases were obtained with GR 8 and GR 10 (p<0.05). Differences in the nutritional values of heated versus nonheated bacteria were insignificant (p>0.05), except for LVS 8 and GR 8 in experiment A which gave slight but significant increases in viable larvae. In no case was the difference in nutritional value great enough to account for the enhanced survival conferred on larvae by feeding with heated versus nonheated bacteria. Regardless of heat exposure, the average length of Artemia larvae, which also acted as an index of bacterial nutritional value, was highest when larvae were fed strain LVS 2 and lowest when LVS 8 was employed. Generally, length variation was insignificant when comparing results obtained by feeding larvae with LVS 2 and other bacterial strains (p>0.05), although some differences were apparent (Table 3). Incubation with only V. campbellii killed all larvae.
Feeding with bacteria, apart from virulent strains such as V. campbellii (Soltanian et al. 2007; Sung et al. 2008), V. proteolyticus (Marques et al. 2006a; Sung et al. 2007), and V. harveyi (Soto-Rodriguez et al. 2003; Defoirdt et al. 2006), enhances Artemia resistance against disease and infection. As one example, colonization of cultures with nine different live bacterial strains protected Artemia juveniles against V. proteolyticus CW8T2 (Verschuere et al. 1999). Moreover, dead bacteria contribute similar beneficial effects, suggesting as one possibility that enhanced tolerance of Artemia to virulent bacteria occurs through stimulation of the nonspecific or innate immune response (Verschuere et al. 2000; Marques et al. 2006a).
In this study, feeding gnotobiotic Artemia with heat-shocked bacteria significantly increased larval resistance against the detrimental effects of V. campbellii. Additionally, immunoprobing of Western blots and protein quantification by ELISA demonstrated that enhanced protection is associated with increased DnaK accumulation in bacteria used as feed. Previously, feeding with E. coli overexpressing DnaK, produced under the control of either its own promoter or a heterologous arabinose-inducible promoter, boosted Artemia survival approximately threefold (Sung et al. 2009). The present study extends these observations, demonstrating clearly that DnaK-producing bacteria other than E. coli confer protection upon Artemia larvae. Moreover, the protective effects obtained by feeding many different strains of heat-shocked bacteria possessing increased DnaK suggest direct involvement of this protein.
The effects of beneficial bacteria in an aquaculture system can be explained by various mechanisms such as improvement of water quality, antagonism towards pathogens including competition for adhesion sites, enzymatic contribution to digestion in the host, and stimulation of the host immune response (Verschuere et al. 2000; Farzanfar 2006; Tinh et al. 2008). How Hsp-producing bacteria protect brine shrimp has yet to be determined, but there is compelling evidence that these proteins induce strong immunological responses in other organisms (Pockley 2003). Bacterial Hsps robustly stimulate the production of proinflammatory cytokine in human monocytes (Galdiero et al. 1997) and induce interleukin-1 secretion from macrophages (Retzlaff et al. 1994). Additionally, extracellular Hsp70 increases the production of inducible nitric oxide synthase and nitric oxide, as well as tumor necrosis factor-α, interleukin-1β, and interleukin-6 in macrophages and neutrophils (Asea et al. 2000; Panjwani et al. 2002; Campisi and Fleshner 2003; Johnson and Fleshner 2006). Furthermore, phagocytes and granulocytes release lysozyme, reactive oxygen species, and cationic peptides upon exposure to Hsps, all of which suppress infection (Jacquier-Sarlin et al. 1994; Basu et al. 2002).
There is accumulating evidence that stimulation of the innate immune response by Hsps protects invertebrates from disease and infection. For example, buildup of Hsp70 after short-term hyperthermic stress correlates with attenuation of gill-associated virus replication in the black tiger prawn, P. monodon (de la Vega et al. 2006). Acting through a conserved pathway involving heat shock transcription factor-1 and the associated DAF-2/DAF-16 pathway, stress-regulated small Hsps and Hsp90 trigger immunity in the nematode C. elegans against Pseudomonas aeruginosa (Singh and Aballay 2006). Additionally, endogenous Hsp70 accretion enhances resistance of gnotobiotic Artemia larvae to V. campbellii and V. proteolyticus (Sung et al. 2007, 2008). In both studies, Hsps were proposed to activate the Artemia innate immune system, thus promoting recognition and destruction of pathogens by defensive mechanisms.
In this study, administration of heat-shocked bacteria enriched in DnaK, and perhaps other Hsps, represented an efficient strategy for biocontrol of vibriosis. This observation has applied significance in aquaculture where disease associated with luminescent vibrios causes mass mortalities of cultured organisms and economic losses worldwide (Diggles et al. 2000; Austin and Zhang 2006). Antibiotic therapy is used but this results in microbial resistance, tissue accumulation of antibiotic residues (Vadstein 1997), and immunosuppression (Hameed and Balasubramanian 2000; Smith et al. 2003), consequently posing serious threats to human health. Application of the results described in this paper has the potential to alleviate these types of problems. The current findings also comment upon the relationship between Hsps, the innate immune response, and bacterial resistance in invertebrates, all of fundamental and applied significance.
This work was supported by the University Malaysia Terengganu (UMT, formerly known as University College of Science and Technology Malaysia, KUSTEM) through a doctoral grant to YYS. Research funding was from the Belgian Foundation for Scientific Research (FWO) through the project “Probiont-induced functional responses in aquatic organisms” (no. G.0491.08) and from the Natural Sciences and Engineering Research Council of Canada to THM. We thank Prof. Dr. Bernd Bukau and Dr. Axel Mogk from the Centrum for Molecular Biology, University Heidelberg, Germany for the generous gift of polyclonal antibody to DnaK.