Interaction between innate immune cells and a bacterial type III secretion system in mutualistic and pathogenic associations

Abstract

Animals house a community of bacterial symbionts in their digestive tracts that contribute to their well being. The medicinal leech, Hirudo verbana, has a remarkably simple gut population carrying two extracellular microbes in the crop where the ingested blood is stored. This simplicity renders it attractive for studying colonization factors. Aeromonas veronii, one of the leech symbionts, can be genetically manipulated and is a pathogen of mammals. Screening transposon mutants of A. veronii for colonization defects in the leech, we found one mutant, JG752, with a transposon insertion in an ascU homolog, encoding an essential component of type III secretion systems (T3SS). Competing JG752 against the wild type revealed that JG752 was increasingly attenuated over time (10-fold at 18 h and >10,000-fold at 96 h). This colonization defect was linked to ascU by complementing JG752 with the operon containing ascU. Fluorescence in situ hybridization analysis revealed that at 42 h 38% of JG752 cells were phagocytosed by leech macrophage-like cells compared with <0.1% of the parental strain. Using mammalian macrophages, a lactate dehydrogenase release assay revealed that cytotoxicity was significantly reduced in macrophages exposed to JG752. In a mouse septicemia model, JG752 killed only 30% of mice, whereas the parent strain killed 100%, showing the importance of T3SS for both pathogenesis and mutualism. Phagocytic immune cells are important not only in defending against pathogens but also in maintaining the mutualistic symbiont community inside the leech, demonstrating that animals use similar, conserved mechanisms to control bacterial populations, even when the outcomes differ dramatically.

The digestive-tract microbiota provide animals with critical benefits, including the synthesis of essential nutrients, digestion of food, and resistance to colonization by pathogens (1). Most digestive-tracts are colonized by a complex microbial community composed of hundreds of different species or, in the case of some insects, tens of species (2–4). This complexity and the inability to cultivate many of these bacteria outside the host have hampered our understanding of the molecular requirements for colonizing digestive tracts, especially when compared with the detailed knowledge available from pathogens (5). Investigations of model systems with simple and culturable digestive-tract microbiota, such as those found in entomophagous nematodes or medicinal leeches, can provide insight into the underlying molecular mechanisms of gut symbioses (6–8).

Hirudo verbana, the medicinal leech, houses a two-member microbial community in the crop of its gut: Aeromonas veronii and an uncultured member of the Bacteroidetes, a Rikenella-like bacterium (9–11). These symbionts are thought to provide essential nutrients such as vitamins, aid in digesting ingested blood, and prevent other bacteria from colonizing the gut (12). The application of molecular techniques to A. veronii and the development of a colonization assay have allowed us to gain insight into colonization requirements and may also reveal factors that contribute to this unusual simplicity (8, 9, 13). H. verbana, usually sold as Hirudo medicinalis from commercial suppliers (43), feeds exclusively on vertebrate blood. During feeding, the leech secretes anticoagulants and vasodilators that stimulate blood flow and produce a beneficial effect on wound healing (14). Medicinal leeches have recently been approved as a medical device by the Food and Drug Administration (15). A leech can consume five times its body weight in a single blood meal (14) with some of the blood's heat-sensitive properties, i.e., the complement system, remaining active and killing susceptible bacteria (16, 17). This large amount of blood is stored in the crop of the digestive tract, where salts and water are absorbed, and the two bacterial symbionts are found (14). The symbionts form polysaccharide-embedded, mixed-species microcolonies that might form floating, granular biofilms or a biofilm associated with erythrocytes (11). In addition to being a symbiont, A. veronii is also a human pathogen, implicated in causing a variety of diseases ranging from diarrhea to septicemia (18). The importance of Aeromonas species as emerging pathogens is highlighted by a recent report of Aeromonas spp. as the leading cause of wound infections in victims of the tsunami in Thailand (19). In immunocompromised patients, one of the most severe complications of an Aeromonas infection is life-threatening septicemia (20). The ability of A. veronii strains to cause septicemia can be assessed in mice by i.p. injections (21). The two lifestyles of A. veronii allow us not only to discover colonization factors for beneficial associations but also to directly compare the similarities and differences of the colonization factors in beneficial and pathogenic associations.

Many opportunistic pathogens (e.g., Yersinia pestis and enterohemorrhagic Escherichia coli) are associated with an alternative host animal where they can be a pathogen, commensal, or mutualist. Studies of Pseudomonas aeruginosa have revealed that some virulence factors are critical for causing disease in plants, worms, and mice (22). Other investigations have revealed that some colonization factors of mutualists are similar to virulence factors (23, 24), yet it remains unclear how similar the colonization factors are for one microorganism that causes disease in mammals yet coexists peacefully with invertebrates. Understanding the molecular basis of this switch is fundamental for understanding the evolutionary dynamics of emerging pathogens.

Researchers searching for virulence factors in Aeromonas hydrophila and Aeromonas salmonicida demonstrated the importance of a type III secretion system (T3SS) in causing disease in mammals and fish (25–28). T3SS are critical virulence factors of many Gram-negative pathogens of plants and animals, including humans (29). The T3SS acts as a molecular syringe that penetrates the membrane of a eukaryotic cell and injects bacterial proteins (effectors) into the cytosol. A wide range of effectors can inhibit or stimulate phagocytosis, confer cytotoxicity, and induce apoptosis (29). A T3SS is also required for establishing a few intracellular beneficial associations [i.e., Sodalis glossinidius–tsetse fly (30) and Rhizobium–leguminous plant (31)], yet to our knowledge, a T3SS has not been demonstrated to be important for an extracellular, beneficial symbiont to colonize a host animal.

Results and Discussion

During a signature-tagged mutagenesis screen of mutants derived from the leech-symbiotic A. veronii strain HM21R, which will be reported elsewhere [A.C.S., N. M. Rabinowitz, S. Küffer, and J.G., unpublished data; (32)], a colonization mutant, JG752, was identified. In the signature-tagged mutagenesis screen, each mutant competes against other strains during the colonization of the leech crop; thus, we initially assessed the colonization ability of JG752 in a competition assay. The kanamycin-resistant mutant was competed against a streptomycin-resistant derivative of the parent strain, HM21RS, by feeding them to the leech in blood that was heat-inactivated to eliminate heat-sensitive antimicrobial properties of sheep blood (17). From 18 h after feeding onward, JG752 had a significantly lower colonization level inside the leech (Fig. 1A). No growth difference between the two strains was detected in aliquots of the inoculated blood that was not fed to the leech (data not shown), indicating that the colonization defect of JG752 was due to a component intrinsic to the leech crop and not because of an inability to grow in blood. In addition, when JG752 was introduced into the leech by itself, 42 h after feeding the mutant had reached a significantly lower concentration (median = 7.3 × 10⁶ cfu/ml, P ≤ 0.035, t test) than the parent strain (median = 4.6 × 10⁸ cfu/ml).

Fig. 1.

The T3SS mutant has a reduced ability to colonize the medicinal leech. (A) Decreased ability of the T3SS mutant JG752 to colonize the leech. JG752 was coinoculated with the competitor strain in a 1:1 ratio. The competitive index [CI = (mutant_output/competitor_output)/(mutant_input/competitor_input)] was calculated for each animal and plotted over time. A CI of 1 (dashed line) indicates that the mutant and competitor strains colonize to equal levels. A CI <1 indicates that the mutant is outcompeted and has a colonization defect. The decrease in CI is statistically significant from 18 h onward (P < 0.05, single-sided, two-tailed t test). (B) The entire T3SS operon 1 was used for complementation. The region used for complementation (black line) encompasses the entire first operon as well as 25 bp downstream of ascU and 233 bp upstream of ascN. The transposon insertion site is indicated by the lollipop. The dashed arrow indicates the remaining portion of the T3SS locus (SI Fig. 4). (Scale bar = 1 kb.) (C) Complementation of the ascU mutant with operon 1. The observed phenotype was linked to the disruption of ascU by complementing the JG752 colonization defect. The parent strain, HM21R, and JG752, carrying either the broad-host-range vector pMMB207 or pAS7 (pMMB207 containing the complementing region) were tested against the competitor strain HM21RS and assessed at 96 h. The CI for each animal is shown; the horizontal line indicates the median CI value. A statistically significant difference between data sets is indicated by an asterisk (single-sided, two-tailed t test, P < 0.05, Mann–Whitney test); ns, no significant difference.

The Tn lesion of JG752 was characterized by Southern analysis, inverse PCR, and DNA sequence analysis. The Southern analysis demonstrated the presence of a single Tn insertion in JG752 (data not shown). Inverse PCR was used to amplify the DNA flanking the Tn insertion. DNA sequence analysis of the amplified fragment revealed that the Tn inserted into a gene with the highest similarity to ascU from A. salmonicida (99% deduced amino acid identity), a homolog of yscU that encodes an inner membrane protein of the T3SS (29, 33). The presence of a complete T3SS was verified by cloning a 40-kb fragment flanking the Tn insertion of JG752 and sequencing a 26 kb region corresponding to the four operons of the T3SS reported in A. hydrophila and A. salmonicida [supporting information (SI) Table 2 and SI Fig. 4]. Sequence analysis revealed a complete T3SS system that was most similar to the T3SS from A. salmonicida even though the gyrB, rpoD, and 16S rRNA gene sequences suggest a closer relationship of A. veronii with A. hydrophila (34). ascU is the terminal gene of the first operon (Fig. 1B) and previous studies in Yersinia enterocolitica demonstrated that inactivation of the ascU homolog resulted in the lack of secretion through the T3SS (33). These results indicate that JG752 has a defect in its T3SS and suggests that a functional T3SS is critical for the colonization of the leech by its native symbiont.

The observed colonization defect was linked to the disruption of ascU by complementing JG752 with the entire operon 1 on pAS7, a derivative of the broad-host-range vector pMMB207 [Fig. 1B (35)]. We chose this approach because ascU is the terminal gene of operon 1 and the native promoter could be used to drive its transcription. The complementation was assessed 96 h after feeding by competing HM21R, the parent strain, or JG752 carrying pAS7 or pMMB207 against HM21RS (Fig. 1C). The complemented mutant colonized at significantly higher levels than JG752 carrying the empty control plasmid. The inability of JG752 (pAS7) to reach wild-type levels could be because of poor maintenance of the plasmid, which is also reflected in the CI values being less than one for HM21R carrying either plasmid. These results demonstrate that ascU is critical for colonizing the leech and the presence of a complete operon 1 is sufficient to restore symbiotic competence to JG752.

In a recent study, we determined the location of A. veronii inside the leech crop and discovered that most A. veronii cells were not associated with the host epithelium but localized within the lumen of the crop (11); however, T3SS require contact with host cells to inject the effectors. We wanted to determine whether the in vivo localization of the mutant was different from that of the parental strain. Using FISH with an Aeromonas-specific probe, sections prepared from leeches 18 and 42 h after feeding confirmed a dramatic reduction of A. veronii cells in JG752-colonized animals (Table 1). Interestingly, cells of the T3SS mutant were detected often in clusters that were intimately associated with nucleated cells, appearing to be intracellular and near its nucleus (Fig. 2 and Table 1). The blood used in these experiments had been reconstituted after centrifugation to remove sheep leukocytes, indicating that these nucleated cells were macrophage-like cells called hemocytes of leech origin (36). By 42 h after feeding, 38.6% of JG752 cell clusters appeared to be intracellular, suggesting that the leech hemocytes phagocytosed the T3SS mutant. In contrast, less than 2.7% of the parent strain clusters were associated with leech hemocytes 42 h after feeding and only 0.1% appeared to be intracellular. The remaining clusters we termed surface-associated (Fig. 2 and Table 1).

Table 1.

The number (percentage) of A. veronii clusters associated with hemocytes in the crop

Strain	Time, h	Surface-associated*	Intracellular*
HM21R	18	3.7 ± 0.6 (0.9 ± 0.4%)	0
	42	42 ± 26 (2.6 ± 1.5%)	1.3 ± 2.3 (0.1 ± 0.1%)
JG752	18	2.7 ± 2.3 (1.8 ± 1.5%)†	4.3 ± 3.2 (2.7 ± 2.2%)‡
	42	0‡	7.7 ± 3.8 (38.6 ± 6.7%)‡

*Number of clusters ± SD. The percentage of associated clusters is shown in parentheses. Statistical differences between number of clusters of HM21R and JG752 were tested using the χ² test. Significant differences are indicated (†, P < 0.05; ‡, P < 0.0001).

Fig. 2.

Fluorescence micrographs of A. veronii clusters in the crop of the leech. (A) Crop sections of leeches inoculated with HM21R or T3SS mutant JG752 at 42 h after feeding. Aeromonas (red) and nuclei (blue) were visualized by staining with an Aeromonas-specific probe AER66 and DAPI, respectively. Bacteria were either not associated with nucleated cells (arrows) or were associated with nucleated cells either surface-associated (SA, green arrowhead) or apparently intracellular (IC, white arrowheads). Erythrocytes have a dark center surrounded by a red autofluorescence (blue arrowheads). (Scale bars: 10 μm.) (B) Representative SA and IC A. veronii in the crop. Merged images of fluorescence and differential interference contrast micrographs are shown. The blue arrowheads indicate erythrocytes of the ingested blood. (Scale bars: 5 μm.)

Our data indicate that a functional T3SS protects A. veronii against phagocytosis by leech hemocytes and that leech hemocytes circulate through the intraluminal fluid inside the crop, enforcing the unusual simplicity of this symbiosis. Avoiding this innate immune response is essential for A. veronii to maintain the symbiosis and, importantly for the symbiotic interaction, this escape is achieved without interfering with the hemocyte's ability to remove other strains as demonstrated by the competition assay. The protection provided by the T3SS appears to act locally. The polysaccharide-embedded microcolonies reported by Kikuchi and Graf (11) may also protect the Rikenella-like symbionts from phagocytosis. Leech hemocytes represent a second layer of selection that enforces the specificity of this symbiosis; we previously demonstrated that the complement system of the ingested blood is an important selective barrier (16, 17). Hemocytes have also been shown to be active in the Vibrio fischeri–Euprymna scolopes light organ symbiosis, but V. fischeri does not possess a T3SS (37, 38). It is intriguing that the hosts in at least two mutualistic symbioses use phagocytes, a common defense mechanism against pathogens, to monitor their microbial community and that A. veronii uses machinery analogous to a well established virulence factor to overcome this powerful layer of the innate immune defense.

A. veronii is also a human pathogen, and this raises the question as to whether the same T3SS is similarly active against mammalian macrophages. Using a lactate dehydrogenase (LDH) release-based cytotoxicity assay, we compared the cytotoxicity of the T3SS mutant with that of the parent strain using mouse RAW 264.7 macrophages (Fig. 3A). This method measures the release of a cytosolic LDH from the host cells as a result of cell lysis. As shown in Fig. 3A, LDH release was significantly reduced in macrophages exposed to JG752 compared with HM21R, indicating that this T3SS was also active against vertebrate macrophages. We further investigated the importance of this T3SS in mammals by using a mouse septicemia model where the bacteria were introduced by i.p. injection. When injected at an equal dose (3 LD₅₀), JG752 killed only 30%, whereas the parent strain killed 100% of the animals in 2 days (Fig. 3B). These results clearly demonstrate the importance of a functional T3SS system for virulence in mammals.

Fig. 3.

Effect of T3SS mutation on virulence in mice. (A) Mouse macrophages were exposed to Aeromonas strains at a multiplicity of infection of 10, and the cytotoxicity was monitored by the release of LDH after 2–3.5 h. The T3SS mutant JG752 had a significantly reduced cytotoxicity compared with the parent strain HM21R at all observed time points (∗, P < 0.05). As a positive control, A. hydrophila strain SSU was included. (B) An equal number of JG752 and HM21R cells were inoculated by i.p. injections into mice, and the mortality of the mice was monitored over a 14-day period. The T3SS mutant had a dramatic and significantly reduced virulence (P < 0.02; Fisher's exact test).

In this study, we have demonstrated the importance of a functional T3SS for the colonization of a digestive tract by an extracellular, beneficial symbiont. Our experiments have revealed that A. veronii strain HM21R requires a functional T3SS for successful beneficial and pathogenic colonization of different host animals. This finding provides evidence that animals use the cellular innate immunity not just to control pathogenic but also beneficial associations. The similarity of the control mechanisms of the host and the bacterial counter measures blurs the distinction between virulence factors and colonization factors of beneficial symbioses (23). Our data suggest that a wide range of beneficial microbes carry the molecular machinery to become deleterious to different host animals thus possessing the means to become a pathogen. Investigating strains like this leech symbiont, which can be both beneficial and pathogenic, will provide further insight into the similarities and differences in the underlying mechanisms of all bacteria–animal associations.

Methods

Bacterial Strains, Plasmids, and Growth Conditions.

The A. veronii biovar sobria T3SS mutant JG752 was derived from the rifampin-resistant parent strain JG84 (HM21R), which is a spontaneous rifampin-resistant mutant derived from HM21, a symbiotic isolate from the leech digestive tract (9). The competitor strain HM21RS is a spontaneous streptomycin-resistant mutant derived from HM21R (13). The E. coli strains TOP10 (TA Cloning Kit; Invitrogen, Carlsbad, CA) and BW20767 (39) were used for cloning and conjugation, respectively. The vector pCR2.1 (TA Cloning Kit; Invitrogen) was used for cloning PCR products. Operon 1 of the T3SS was cloned into the broad-host-range vector pMMB207 (35) yielding pAS7, which was then used for complementation. Its construction is described below. The bacteria were cultured at 200 rpm in LB broth or on LB agar plates at 37°C and 30°C for E. coli and Aeromonas species, respectively. The growth medium was supplemented with the appropriate antibiotics at the following concentrations: ampicillin, 100 μg/ml; chloramphenicol, 1 μg/ml; kanamycin, 100 μg/ml; rifampin, 100 μg/ml for selection and 10 μg/ml for maintenance; and streptomycin, 100 μg/ml.

Inverse PCR.

Inverse PCR was used to amplify DNA flanking the Tn insertion (40). Genomic DNA was isolated as described previously (13). Two micrograms of genomic DNA were digested with the restriction enzymes PstI in a 20-μl reaction. The digest was precipitated after 6 h using 6.5 μl of 10 M NH₄ acetate and 65 μl of 95% ethanol and centrifugation at 13,200 × g for 15 min at 4°C. The pelleted DNA was washed twice with 100 μl of 70% ethanol and dried at 70°C. The DNA was resuspended in 90 μl of dH₂O by heating it to 50°C for 10 min. After cooling, 10 μl of 10× T4 ligation buffer was added. Two microliters of T4 DNA ligase (NEB) were added, and the reaction was incubated overnight at 16°C. Inverse PCR amplification of the interrupted gene was performed using an outward-facing primer pair located on mTn5, P2 (5′-TACCTACAACCTCAAGCT) or P4 (5′-TACCCATTCTAACCAAGC) and P8 (5′-GAGACACAACGTGGCTTT) (32). Each reaction contained 2 μl from the ligation, 1× PCR amplification buffer, 1 mM MgCl₂, 200 μM of each deoxynucleotide triphosphate, 0.5 μM of each primer, and 0.5 units of Platinum TaqDNA polymerase (Invitrogen) in a final volume of 20 μl. The amplification conditions were as follows: (i) 5 min at 95°C and (ii) 20 cycles of 30 s at 95°C, 30 s at 50°C, and 1.5 min at 72°C. Nested PCR amplification of the inverse PCR product was done using a primer pair located on pUTminiTn5 outside of P2/P4 and P8, P6 (5′-CCTAGGCGGCCAGATCTGAT) and P9 (5′-CGCAGGGCTTTATTGATTC) (32). Each reaction contained 2 μl of inverse PCR product, 1× PCR amplification buffer, 1 mM MgCl₂, 200 μM of each deoxynucleotide triphosphate, 0.5 μM of each primer, and 0.5 units of Platinum TaqDNA polymerase (Invitrogen) in a final volume of 20 μl. The amplification conditions were as follows: (i) 5 min at 95°C and (ii) 35 cycles of 30 s at 95°C, 30 s at 50°C, and 1.5 min at 72°C. PCR products were purified using the QIAquick PCR purification kit (Qiagen, Valencia, CA). Purified PCR products were sequenced (described below) using primers P6 and P9.

Fosmid Constuction.

The fosmid for JG752 was constructed using the CopyControl Fosmid Library Production Kit according to the manufacturer's instructions (Epicentre Technologies, Madison, WI). Genomic DNA isolated from JG752 was randomly sheared by passing it through a 200-μl small bore pipette tip, blunt-end repaired, and separated in 1% SeaPlaque GTG agarose gel at 17 V overnight at 4°C. The fragments corresponding to 40 kb were excised, gel-purified, and ligated into pCC1FO vector. The ligation was packaged using MaxPlax Lambda packaging extract and used to transduce EPI300-T1 E. coli. Transduced cells were plated on LB agar containing chloramphenicol and kanamycin to select for CopyControl Fosmid clones containing DNA flanking the transposon. Clones were streaked for isolation, and fosmid DNA was purified using the FosmidMAX DNA Purification Kit (Epicentre Technologies).

DNA Sequencing.

The fosmid was sequenced by primer walking using the BigDye sequencing kit (Applied Biosystems, Foster City, CA). For sequencing fosmids, each sequencing reaction contained 4 μl of BigDye Terminator version 1.1, 5.5 μl of purified fosmid DNA, and 2.5 μM of primer. The sequencing conditions were as follows: 65 cycles of 30 s at 95°C, 20 s at 50°C, and 4 min at 65°C. PCR products were sequenced using primers P6 and P9. Each sequencing reaction contained 0.75 μl of BigDye version 1.1, 3.45 μl of the amplicon, and 0.16 μM of primer. The sequencing conditions were as follows: (i) 5 min at 95°C and (ii) 25 cycles of 30 s at 95°C, 20 s at 55°C, 4 min at 60°C, and 10 min at 72°C. The reactions were then run on an ABI PRISM 3100 (Applied Biosystems) capillary DNA sequencer. The DNA sequences obtained in this study were deposited in GenBank (accession no. EF215451).

Sequence Analysis.

Contiguous DNA sequences were aligned using ContigExpress and analyzed using VectorNTI 7. The sequences were compared with the NCBI database using BLASTX and BLASTN (41).

Competition Assay.

The competition assay used in this study compares the colonization ability of a test strain against a competitor strain, HM21RS, by inoculating a blood meal with 250 cfu/ml of each strain. Otherwise the conditions were identical to the assay we described previously except that heat-inactivated blood (Quad Five, Ryegate, MT) was used (9, 16). For each time point, at least three animals were used. At 6, 18, 42, 72, 96, and 120 h after feeding, animals were dissected and the intraluminal fluid was collected and serially diluted as described previously (9). The limit of detection was 10 cfu/ml. The competitive index was calculated as follows: CI = (mutant_output/competitor_output)/(mutant_input/competitor_input). A CI of 1 indicates that the mutant colonized to the same level as the competitor strain and a CI <1 indicates that the mutant had a colonization defect.

Growth in Heat-Inactivated Blood.

Heat-inactivated blood was inoculated with 250 cfu/ml of mutant and competitor strains. The inoculated blood was incubated at room temperature (23°C). Aliquots were removed at 0, 18, 24, and 42 h after inoculation and plated as described for the competition assay.

Statistical Analysis.

Data were analyzed using GraphPad Prism 4.0a (GraphPad, San Diego, CA) using the methods indicated in the figure legends. A two-tailed, one-sided t test was used to determine whether the CI differed from 1, with P ≤ 0.05.

Complementation of the T3SS Mutant.

A 7-kb fragment containing the first operon of the T3SS was PCR-amplified. The primer pair, Op1-F (5′-GATAAGCTTCGTGTTGCGCTGTTTGTTCG) and Op1-R (5′-TCGAAGCTTTTGACGGGATGCGACCTCTG), each contained HindIII restriction sites. Primer Op1-F anneals 25 bp downstream of ascU, and primer Op1-R anneals 233 bp upstream of ascN. The reaction contained 100 ng of DNA, 1× PCR buffer, 500 μM of each deoxynucleotide triphosphate, 0.4 μM of each primer, 0.05 units/μl AccuTaq LA DNA polymerase mix (Sigma, St. Louis, MO), and 2% DMSO in a final volume of 50 μl. The amplification conditions were as follows: (i) 30 s at 98°C; (ii) 30 cycles of 15 s at 94°C, 20 s at 57°C, and 7 min at 68°C; and (iii) 10 min at 68°C. The TA Cloning Kit was used to clone the PCR product into pCR 2.1 according to manufacturer's instructions (Invitrogen). The resulting plasmid, pAS6, was isolated using the Qiagen Plasmid purification kit. The cloned 7-kb fragment was ligated into the broad host range vector pMMB207 from pAS6 by using the HindIII restriction sites introduced during PCR and present in the multiple cloning site of pMMB207. The resulting plasmid, pAS7, was electroporated into the E. coli strain BW20767. The presence of the insert was verified by PCR and DNA sequencing. The primer pair pAS7F (5′-TTTGCGCCGACATCATAACG) and pAS7R (5′-CAGACCGCTTCTGCGTTCTG) flanks the multiple cloning site of pAS7. The PCR amplification was performed as above. pAS7 and pMMB207 were each introduced into HM21R and JG752 by conjugation (13, 42). The resulting strains were competed against HM21RS.

FISH.

Leeches were fed 3 ml of blood inoculated with 250 cfu/ml and incubated at room temperature (23°C). The blood meal was reconstituted after three centrifugations to remove sheep leukocytes and over 95% of the leukocytes were removed and heat-inactivated. Three replicates were collected 18 and 42 h after feeding. The serial tissue sections of the leech midbody were prepared and subjected to FISH as described previously (11). The 16S rRNA targeting Aeromonas-specific probe Cy3-AER66 (5′-CTACTTTCCCGCTGCCGC-3′) was used for detecting A. veronii. Hybridization buffer (150 μl; 20 mM Tris·HCl pH 8.0, 0.9 M NaCl, 0.01% SDS, and 30% formamide) containing 50 pmol/ml of each probe and 4 nmol/ml DAPI was applied to the tissue samples and incubated at room temperature overnight. The samples were washed in hybridization buffer for 10 min at 37°C. The fluorescence signals were observed with an epifluorescence microscope (TE2000; Nikon, Florham Park, NJ), and images were captured with a digital camera (Spot RT-KE; Diagnostic, Sterling Heights, MI). For each animal, the total number of the clusters and individual cells was counted in one randomly selected section at ×400 magnification. For ease of discussion, we report the data as clusters even though A. veronii cells were usually present as single cells except when associated with hemocytes.

Cytotoxicity Assay.

The wild-type A. hydrophila SSU as a positive control (21), HM21R, and JG752 were grown in 3 ml of LB medium in 50-ml disposable tubes and incubated at 37°C overnight with shaking (180 rpm). The bacterial cells were centrifuged and washed three times with PBS. RAW 264.7 murine macrophages (American Type Culture Collection, Manassas, VA) were seeded into 96-well plates (1 × 10⁵ cells per well) and infected with the live bacterial cultures (as prepared above) at a multiplicity of infection of 10. After incubation at 37°C for 2.5 to 3.5 h, the tissue culture medium was examined for the release of LDH enzyme using a CytoTox96 kit (Promega, Madison, WI). The released LDH in tissue culture supernatants converts tetrazolium salt into a red formazan product. The intensity of the red color formed is proportional to the number of lysed cells and is measured at 490 nm. Student's t test was used to compare the data.

Animal Experiments.

Two groups of 10 Swiss–Webster mice (Taconic Farms, Germantown, NY) were infected by the i.p. route with 5 × 10⁷ bacteria (HM21R or JG752) in accordance with approved animal care protocols. Deaths were recorded for 14 days after infection. The statistical analyses were performed using Fisher's exact test.

Abbreviations

LDH

lactate dehydrogenase

T3SS

type III secretion system

CI

competitive index.