|Home | About | Journals | Submit | Contact Us | Français|
Hemolysin and the type II secretion system (T2SS) have been shown to be important for virulence in many pathogens, but very few studies have shown their importance in beneficial microbes. Here, we investigated the importance of the type II secretion pathway in the beneficial digestive-tract association of Aeromonas veronii and the medicinal leech Hirudo verbana and revealed a critical role for the hemolysis of erythrocytes. A mutant with a miniTn5 insertion in exeM, which is involved in forming the inner membrane platform in the T2SS, was isolated by screening mutants for loss of hemolysis on blood agar plates. A hemolysis assay was used to quantify the mutant's deficiency in lysing sheep erythrocytes and revealed a 99.9% decrease compared to the parent strain. The importance of the T2SS in the colonization of the symbiotic host was assessed. Colonization assays revealed that the T2SS is critical for initial colonization of the leech gut. The defect was tied to the loss of hemolysin production by performing a colonization assay with blood containing lysed erythrocytes. This restored the colonization defect in the mutant. Complementation of the mutant using the promoter region and exeMN revealed that the T2SS is responsible for secreting hemolysin into the extracellular space and that both the T2SS and hemolysin export by the T2SS are critical for initial establishment of A. veronii in the leech gut.
Animal digestive tracts house an array of microorganisms that play an important role in aiding digestion and providing essential nutrients to the host (33, 50, 53). The vast number of species that make up the microbiome, the presence of microbes recalcitrant to cultivation in the laboratory, and the lack of genetic tools for many cultured microbes contribute to the challenges of studying digestive-tract microbiota (15, 52). The availability of versatile invertebrate symbioses with simpler, naturally occurring microbial communities overcomes these difficulties and provides important new insights that are applicable to more complex communities.
The digestive-tract symbiosis of the medicinal leech Hirudo verbana is such a simple model system (15, 52). The largest region of the digestive tract, the crop, contains a simple microbial community dominated by two species, Aeromonas veronii and a Rikenella-like bacterium (15, 52). To date, two factors that contribute to the simplicity of this symbiosis have been identified: (i) the complement system of the ingested blood meal (4) and (ii) leech hemocytes (46). A. veronii is a Gram-negative, facultative anaerobe and a human pathogen causing diseases such as wound infections and gastroenteritis (21, 22). The Rikenella-like bacterium is a member of the phylum Bacteroidetes. 16S rRNA gene sequence analysis revealed that this constitutes a novel genus related to Rikenella microfusus (52), which has been isolated from cecal and fecal samples of Japanese calves, chickens, and quails (27). In a single feeding, the medicinal leech can consume more than five times its body weight, after which the blood is stored in the crop (14, 43). It has been reported that the ingested erythrocytes remain physically intact inside the crop and are degraded when they enter the intestinum (14, 32). Recent direct microscopic observations also confirmed the presence of visually intact erythrocytes (30). We recently discovered that A. veronii utilizes heme as an iron source inside the crop and that this ability is critical during the initial colonization phase (M. Maltz, B. LeVarge, and J. Graf, unpublished data). This access to heme implies that at least a portion of the erythrocytes are permeabilized or lysed to an extent that allows hemoglobin to be released.
One way that bacteria can lyse erythrocytes is by secreting hemolysin into the extracellular space (20, 44, 49, 51). The best-studied hemolysin produced by aeromonads, aerolysin, is a pore-forming protein that leads to the release of glucose and potassium from erythrocytes, which leads to colloid-osmotic lysis and the release of hemoglobin, resulting in beta-hemolysis (8, 16, 49). Many Aeromonas species have been shown to excrete hemolysins, e.g., aerolysin, into the environment (8). Studies have also revealed that hemolysin expression is regulated by iron-limiting conditions (49), which could mimic iron restriction inside a host and serve as an entry signal.
The establishment of beneficial microbe-host associations follows the traditional infection cycle of pathogens, i.e., entry, multiplication, and avoidance of host defenses (13, 35), except that no overt damage is done to the host. One of the most widespread systems that can contribute to this process is the type II secretion system (T2SS), also known as the general secretory pathway (GSP). In most Gram-negative bacteria, the T2SS is used to transport proteins into the extracellular space (9, 10). This system requires two main steps; the sec-dependent GSP translocates proteins from the cytoplasm into the periplasmic space, and with the signal sequenced removed, the fully folded protein is translocated out of the cell through the T2SS. The exact mechanism, configuration, and order of this apparatus are still unknown. Studies have shown that there are 12 conserved core components of the T2SS, and several other proteins are not conserved, including ExeN, which is believed to be involved in the placement of the exporting proteins in the outer membrane (25, 41, 42). The genomes of A. hydrophila and A. salmonicida encode a T2SS, and both have been studied in detail (17, 18, 28, 37, 45).
There have been many links between the T2SS and pathogenesis (42). While the T2SS exports many hydrolytic enzymes, in many cases, the T2SS is also used to export toxins or extracellular virulence factors, including aerolysin, from A. hydrophila (2, 10, 41). Although genomic data revealed the presence of the T2SS in many symbionts, there are very few functional studies linking T2SS and beneficial relationships (10, 47a). Here, we investigated the role of the T2SS in the symbiosis of A. veronii strain Hm21, an isolate from the digestive tract of the medicinal leech (14).
A. veronii strains were cultured at 30°C and Escherichia coli was cultured at 37°C in LB medium or on LB or blood agar (BA) plates (40) (Table (Table1).1). LB contained 10 g Bacto Tryptone, 5 g yeast extract, and 10 g NaCl per liter. For agar plates, 15 g/liter Bacto Agar was added. BA contained 44 g Columbia BA base per liter. A 50-ml volume of sheep blood (Quad Five, Ryegate, MT) was added to BA after the agar had cooled to 50°C. Growth medium was supplemented with the appropriate antibiotics at the following concentrations: ampicillin (Ap), 100 μg/ml; streptomycin (Sm), 100 μg/ml; kanamycin (Km), 100 μg/ml; trimethoprim (Tp), 100 μg/ml; rifampin (Rf), 100 μg/ml for selection and 10 μg/ml for maintenance. All chemicals and reagents were purchased from Fisher Scientific (Pittsburgh, PA) unless otherwise stated.
The A. veronii T2SS mutant (HE-1095) was derived from HM21S by conjugation with E. coli strain BW20767 harboring pRL27 (31) (Table (Table1),1), which carries a miniTn5 transposon, as described previously (31; Maltz et al., unpublished). More than 18,000 mutants were screened on BA plates for the inability to lyse erythrocytes, which is represented by a clear halo around a colony (49). BA plates were prepared as described above and supplemented with Sm and Km.
The ability of the mutant to produce hemolysin was assessed using a modified hemolysis assay previously described by Brillard et al. (5, 6). Sheep erythrocytes were rinsed and resuspended in phosphate-buffered saline (PBS) in triplicate. The supernatant from overnight cultures was filter sterilized, and 250 μl was added to 250 μl of washed erythrocytes to a final concentration of 5%. The samples were then incubated for 1 h at 37°C. After incubation, the samples were centrifuged and the optical density at 540 nm of the supernatant was measured. The percentage of hemolysis was calculated with the following formula: [(A540 of samples with hemolysis − A540 of negative control)/A540 of positive control] × 100 (5, 6).
Inverse PCR was used to amplify and verify the DNA flanking the Tn insertion (36). Genomic DNA was extracted from HE-1095 using the MasterPure DNA purification kit (Epicentre Biotechnologies, Madison, WI). Genomic DNA was digested with NcoI (New England Biolabs, Ipswich, MA). A self-ligation was preformed on the digest using T4 DNA ligase (New England Biolabs). The PCR mixture contained 2 μl ligation reaction mixture, 1× GOTaq Green Master Mix (Promega, Madison, WI), and 4 μM tpnRL17-1 and tpnRL13-2 (31) in a final volume of 25 μl. The amplification conditions were as follows: (i) 5 min at 95°C and (ii) 32 cycles of 30 s at 95°C, 30 s at 60°C, and 180 s at 72°C.
DNA sequencing and analysis were done as previously described by Silver et al. (47), except that it was supplemented with a draft of the A. veronii genome.
The ability of the mutant to colonize the crop of H. verbana was done as described previously (14, 38, 46, 47), with the following modifications. The blood was reconstituted to remove nutrients released from lysed erythrocytes present in the blood prior to inoculating it with A. veronii (46). Sheep blood was centrifuged, and plasma and the buffy coat containing leukocytes were removed. The erythrocytes were washed with PBS to remove lysed erythrocytes and resuspended in heat-inactivated sheep serum. In the experiments with lysed blood, after erythrocytes were washed, 1 ml of nanopure H2O was added to 2.5 ml of erythrocytes and then heat-inactivated sheep serum was added. The leech and the inoculated blood as an in vitro control were incubated at room temperature for 3, 6, 18, 24, 42, and 72 h after inoculation. The leech was then sacrificed, and intraluminal fluid and blood samples were plated on LB-Sm. Differences in colonization levels were analyzed using Prizm software. The Mann-Whitney test was performed to calculate P values.
A competition assay was performed with Hm21R and HE-1095 as described by Silver et al. (46), except that the blood meal was modified as described above. Also, the ratio of Hm21S to HE-1095 in the inoculum was varied (50:50, 25:75, and 15:85). The competitive index (CI) was calculated as follows: CI = (mutantoutput/competitoroutput)/(mutantinput/competitorinput). A CI of 1 indicates that the mutant colonized to the same level as the parent strain, and a CI of <1 indicates that the mutant had a colonization defect.
The HE-1095 mutant was complemented using Tn7 as previously described (7; Maltz et al., unpublished), with the following modifications. The 192-bp promoter region of the T2SS was PCR amplified from Hm21 genomic DNA using primers PromoterR(NHE1) (5′ACCGCTAGCGTCATCCCTGAATCGTAG 3′) and PromoterF (5′ GATTGGCGAGGGTGAGTA 3′), with added NheI restriction sites on the PromoterR(NHE1) primer. The reaction mixture and amplification conditions were as described above, except that the annealing temperature was 62°C and the elongation time was 60 s. The T2SS promoter region PCR product was inserted into pCR2.1 using a TA cloning kit (Invitrogen, Carlsbad, CA), yielding pMMPR1 (Table (Table1).1). The 1,300-bp fragment containing the exeMN genes was PCR amplified from Hm21 genomic DNA using primers exeMNF(Nhel) (5′ ACCGCTAGCGCTCAAGGGTAAATCATCA 3′) and exeMNR(Nhel) (5′ ACCGCTAGCGCCGCGAGATAAACAAATA 3′) with added NheI restriction sites. The reaction mixture and amplification conditions were as described above, except that the annealing temperature was 65°C and the elongation time was 90 s. The PCR product was then cloned into pCR2.1 using a TA cloning kit (Invitrogen, Carlsbad, CA), yielding pMMN2. The exeMN genes were then excised from pMMN2 using NheI and inserted into the NheI site at the 3′ end of the promoter region in pMMPR1, yielding pMMPM3 (Table (Table1).1). The direction of exeMN was verified by colony PCR on transformants using primers PromoterF and exeMNtestR (5′ AGTAGAAGAGACCGACCA 3′). The reaction mixture and amplification conditions were as described above, except that the annealing temperature was 55°C, the elongation time was 60 s, and 1 μl of template DNA from a colony was diluted in 20 μl of water. The promoter region and exeMN were then removed from pMMPM3 using EcoRI restriction sites and cloned into pTnTp using EcoRI restriction sites located on Tn7, yielding pTn7PM (Table (Table1).1). Quadriparental matings were performed to conjugally transfer the transposon containing the promoter region and exeMN genes into Hm21R and HE-1095, respectively. The insertion of Tn7 downstream of glmS was verified as previously described (7; Maltz et al., unpublished).
The DNA sequences obtained in this study were deposited in GenBank under accession number HM623429.
The reports of finding intact erythrocytes inside the crop (26) that is inhabited by beta-hemolytic A. veronii (14, 23, 29) seems surprising but consistent with the notion of the crop being a storage area for the ingested blood meal and suggested that somehow the beta-hemolytic activity of A. veronii was suppressed. Our recent finding of A. veronii acquiring heme inside the crop suggests that at least a portion of the erythrocytes are lysed or permeabilized to such an extent that hemoglobin can exit (Maltz et al. unpublished). Many studies have shown the role of hemolysin in the virulence of bacteria such as E. coli, Xenorhabdus nematophilus, Pseudomonas aeruginosa, Burkholderia cepacia, and vibrios (11, 19, 20, 44). There have been fewer studies showing the importance of hemolysin to beneficial interactions, some examples are Chlorochromatium aggregatum, Bacteroides fragilis, and Rhizobium sp. (34, 39, 51).
Our approach to determining the role that hemolysis plays in the ability of leech isolate Hm21 to colonize the leech was to screen 18,000 miniTn5 mutants on BA plates for the inability to lyse erythrocytes, which is represented by a clear halo around a colony (49). This screen revealed one mutant with dramatically reduced hemolysin activity, HE-1095, which had no clear halo around the colony (Fig. (Fig.1).1). Many other mutants that initially appeared to be hemolysis negative were slow growing and displayed beta-hemolysis when subcultured.
The inability of HE-1095 to lyse erythrocytes was further verified with a hemolysis assay (Fig. (Fig.2)2) (6, 49). For Hm21S, the average was 96.45% hemolysis and for HE-1095, the average was less than 1% hemolysis, showing a decrease in the mutant's hemolysin activity. These data suggest that due to the Tn insertion, HE-1095 is unable to secrete hemolysin proteins under these conditions.
The insertion site was identified by inverse PCR (31, 36). A BLASTX search of the NCBI database revealed that the deduced amino acid sequence of the inactivated gene was 55% identical to ExeM in A. hydrophila (1). ExeM is predicted to be part of the T2SS involved in forming the inner membrane platform (17, 25, 41, 42). Searching of an Hm21 draft genome revealed that exeM is the second-to-last gene in the 11.8-kb T2SS locus. Downstream of exeM is exeN, which is predicted to be involved in placement of the exporting proteins in the outer membrane (Fig. (Fig.3).3). The inability of HE-1095 to form the inner ring platform and place exporting proteins renders the mutant unable to complete the formation of a functional T2SS. Thus, in HE-1095, proteins cannot be secreted through the T2SS, which would include aerolysin homologs. The loss of hemolysis activity in HE-1095 demonstrates that unlike E. coli, which uses the type I secretion system to secrete hemolysin, A. veronii uses the T2SS. Further searching of a draft genome of A. veronii revealed the presence of multiple genes encoding hemolysins, offering an explaining why the only mutant found in this screen was a T2SS mutant. Given the inactivation of the T2SS, we predict that many other hydrolytic enzymes are also not secreted by this mutant, which preliminary phenotypic tests support.
A. veronii is capable of colonizing the leech crop and also causing many human diseases, including septicemia (21, 23, 24). This suggests that it can thrive in a blood-derived environment, which allows us to use blood as a relevant in vitro control. The leech modifies a blood meal by removing water and osmolytes (14), as well as infiltrating the intraluminal fluid with hemocytes (46), macrophage-like cells of invertebrates. This control allows us to differentiate between a general growth defect in blood and a more specific colonization defect inside the leech gut. The ability of HE-1095 to grow in blood was assessed with a time course assay (3, 6, 18, 24, and 42 h). At all time points, the mutant was able to grow to the same levels as Hm21S (Fig. (Fig.44 A.). These data suggest that the T2SS is not necessary for A. veronii to grow in blood.
The ability of HE-1095 to colonize the crop of H. verbana was monitored over time (6, 18, 24, and 48 h) (Fig. (Fig.4B).4B). In these experiments, the introduced strain has to compete with the native microbiota. However, the long starvation of the leeches by the suppliers reduces the levels of the native microbiome so that initially there is less competition than in our standard competition assay, where two strains are introduced at the same time (46, 47). At 6 h, the mutant colonized to significantly lower levels then Hm21S (Fig. (Fig.4B),4B), suggesting that the T2SS is important for initial establishment of A. veronii in the leech gut. At 18 and 24 h, the mutant still has a significant growth defect, but by 24 h there is a slight recovery in some animals. In contrast to the in vitro blood control, competition with the native Rikenella-like symbiont or native A. veronii seems a likely explanation, but it could be that the hemolysin is needed to provide protection against leech hemocytes that patrol the leech crop (46). At 42 h, HE-1095 grows to the same levels as the parent strain. These data suggest that after 42 h, HE-1095 could be cheating off the natural microflora in the crop for nutrients, allowing the mutant to recover once the native population has grown, or that proteins secreted through other pathways are expressed 42 h after feeding.
A competition assay was performed with the mutant and a competitor strain introduced at different ratios to determine if the early growth delay of HE-1095 was due to competition with native A. veronii. The assay was done as previously described by Silver et al. (46), except that the blood meal was modified as described above. In the competition assay, the blood contained the mutant at 250 CFU/ml and decreasing numbers of bacteria of strain Hm21R, a spontaneously Rfr mutant. The CI was calculated as follows: CI=(mutantoutput/competitoroutput)/(mutantinput/competitorinput). A CI of 1 indicates that the mutant colonized to the same level as the parent strain, and a CI of <1 indicates that the mutant had a colonization defect. When inoculated at a 1:1 ratio, the average CI value for HE-1095 versus Hm21S was 0.59 ± 0.56 (standard deviation). A single-sided, two-tailed t test revealed that there was no significant difference from a value of 1 (Fig. (Fig.5).5). This competition assay revealed that the competitor complemented the T2SS mutant; which could be cheating off secreted enzymes or released nutrients. To further investigate the idea of cheating, animals were fed blood containing fewer HM21R cells, i.e., Hm21R/HE-1095 ratios of 25:75 and 15:85. The CI values for both ratios were significantly different from a value of 1 (Fig. (Fig.5),5), indicating that the T2SS mutant required sufficient numbers of Hm21R bacteria to proliferate inside the leech gut at 6 h. The simplest explanation is that the T2SS mutant is cheating off Hm21R secreted enzymes or nutrients released by their activity. This complementation by the wild-type strain also explains why this mutant was not recovered in our previous signature-tagged mutagenesis screen, in which many strains were fed to individual animals (47).
The T2SS is the general secretion pathway, causing HE-1095 to be pleiotropic, but we wanted to examine if there was a link between the lack of hemolysin production and the colonization phenotype. Leeches were fed HE-1095 in a blood meal containing either 100, 1, or 0.1% lysed erythrocytes or intact, washed erythrocytes. When inoculated into blood meals containing 100 or 1% lysed erythrocytes, HE-1095 grew to the same levels as the parent strain, while in the animals fed washed erythrocytes the mutants failed to proliferate (Fig. (Fig.6).6). In animals fed 0.1% lysed erythrocytes, HE-1095 grew but reached a significantly lower density than the parent strain. These results suggest that the lack of hemolysin activity and the inability to lyse erythrocytes each play a role in the T2SS mutant's leech crop colonization defect. These data imply that the ability to lyse erythrocytes is essential for the initial establishment of A. veronii in the leech gut. However, it remains possible that the release of nutrients by adding lysed erythrocytes complements a different nutritional defect that is unrelated to the lysis of erythrocytes. This will be further investigated by inactivating the multiple hemolysins that A. veronii possesses. In a recent study, A. veronii was shown to require iron in the form of heme to colonize the leech gut (Maltz et al., unpublished). A. veronii could be lysing erythrocytes to obtain nutrients such as iron or a carbon source that are stored within the erythrocytes, which contradicts previous studies that show that erythrocytes remain physically intact for many months inside the crop of the leech and do not lyse until they enter the intestinum (32). A. veronii has to compete for the released nutrients with the Rikenella-like bacterium, the other dominant symbiont (30, 52). The high concentrations of HE-1095 in the leech crop at 42 h (Fig. (Fig.4B)4B) could be due to the mutant cheating off the hemolysin production by the native A. veronii or the Rikenella-like bacteria that increase in abundance after feeding or other hemolysins that are expressed later on in a T2SS-independent manner.
The colonization defect was also linked to the inactivation of exeMN genetically by complementing HE-1095 with exeMN and the promoter region of the T2SS locus carried on a Tn7 transposon, yielding Hm21C-1343 and HEC-1344 (Table (Table1).1). Complementation was assessed by streaking bacteria on BA plates and performing a hemolysis assay (Fig. (Fig.11 and and2).2). HEC-1344 and Hm21S had similar clear halos on BA plates and percentages of hemolysis, suggesting that the inactivation of exeMN was responsible for the significantly lower percentage of hemolysis in HE-1095. Complementation of the leech colonization phenotype was assessed 6 h after feeding (Fig. (Fig.7).7). HEC-1344 grew to the same levels as Hm21S in the leech crop, showing that the interrupted exeMN genes in the T2SS locus are responsible for the colonization defect in HE-1095. These data link the inactivation of the T2SS to hemolysin export and its importance for A. veronii colonization of the leech gut.
The T2SS is a widespread mechanism used for protein transport in many Gram-negative bacteria. Although it is well studied in pathogenic relationships, less is known about the role the T2SS plays in beneficial relationships. Our studies indicate that in A. veronii, the T2SS exports hemolysin into the extracellular space and suggest that the export of hemolysin into the environment is important for the initial colonization of the leech crop by A. veronii. These studies ultimately link the T2SS and hemolysin as critical colonization factors in beneficial relationships. We are currently working on demonstrating the link between hemolysis and the colonization defect more directly, but the presence of multiple hemolysins complicates this approach.
We thank Matt Hung, Katherine Miller, and Janet Yi from the UConn Mentor Connection for their help in making Tn mutants and screening them. We also thank Lindsey Bomar and Corey Cates for helpful comments on the manuscript.
This research was funded by NFS Career Award MCB 0448052 to J.G.
Published ahead of print on 19 November 2010.