Gel shift analysis of the empA promoter region in Vibrio anguillarum

Background The induction of metalloprotease encoded by empA in Vibrio anguillarum occurs at high cell density in salmon intestinal mucus. Previously we have shown that there are significant differences in empA expression in two strains of V. anguillarum, M93Sm and NB10. It is hypothesized that differences in empA regulation are due to differences in binding of regulatory elements. Results Two strains of V. anguillarum, M93Sm and NB10, were examined and compared for the presence of DNA regulatory proteins that bind to and control the empA promoter region. Gel mobility shift assays, using a digoxigenin (DIG)-labeled oligomer containing a lux box-like element and the promoter for empA, were done to demonstrate the presence of a DNA-binding protein. Protein extracts from NB10 cells incubated in Luria Bertani broth + 2% NaCl (LB20), nine salts solution + 200 μg/ml mucus (NSSM), 3M (marine minimal medium), or NSS resulted in a gel mobility shift. No gel mobility shift was seen when protein extracts from either LB20- or NSSM-grown M93Sm cells were mixed with the DIG-labeled empA oligomer. The azocasein assay detected protease activity in all incubation conditions for NB10 culture supernatants. In contrast, protease activity was detected in M93Sm culture supernatants only when incubated in NSSM. Since the luxR homologue in V. anguillarum, vanT, has been cloned, sequenced, and shown to be required for protease activity, we wanted to determine if vanT mutants of NB10 exhibit the same gel shift observed in the wild-type. Site-directed mutagenesis was used to create vanT mutants in V. anguillarum M93Sm and NB10 to test whether VanT is involved with the gel mobility shift. Both vanT mutants, M02 and NB02, did not produce protease activity in any conditions. However, protein extracts from NB02 incubated in each condition still exhibited a gel shift when mixed with the DIG-labeled empA oligomer. Conclusions The data demonstrate that protein extracts of V. anguillarum NB10 cells contain a protein that binds to a 50 bp oligomer containing the empA promoter-lux box-like region. NB10 cells express empA during stationary phase in all growth conditions. The DNA binding protein is not present in M93Sm extracts. M93Sm cells express protease activity only when incubated at high cell density in fish gastrointestinal mucus. The gel shift observed with NB10 cells is not due to VanT binding. The data also suggest that the DNA binding protein is responsible for the less restrictive expression of empA in NB10 compared to M93Sm.


Background
Vibrio anguillarum is the causative agent of vibriosis, one of the major bacterial diseases affecting fish, bivalves, and crustaceans [1][2][3]. Vibriosis has been a major problem for the aquaculture industry around the world. Large economic losses due to this fish pathogen are sustained by the fish farming industry. Annual losses of cultured fish species in Japan alone exceed $30 million [1]. Vibriosis in fish is observed as a hemorrhagic septicemia. Infected fish display skin discoloration and erythema around the base of the fins, vent, and mouth. Necrotic lesions form in the abdominal muscle. Mortalities within affected fish farm stocks range from 30-100% [1,4,5].
Extracellular metalloproteases are important virulence factors for many pathogenic bacteria including Vibrio cholerae [6,7] and Vibrio vulnificus [8][9][10]. The EmpA metalloprotease genes of V. anguillarum strains NB10 and M93Sm have been cloned and sequenced by Milton et al [11] and Denkin and Nelson [12], respectively. The EmpA metalloprotease of V. anguillarum shares significant sequence homology with other known proteases. Although the role of the EmpA metalloprotease is not completely understood, we showed previously that V. anguillarum empA mutants are either avirulent or attenuated for virulence in Atlantic salmon depending on the route of inoculation [12]. It can be hypothesized that EmpA causes tissue damage during pathogenesis.
Recently, the Vibrio harveyi luxR (LuxR Vh ) homologue in V. anguillarum, vanT, was cloned, sequenced, and shown to be required for protease activity [13]. LuxR Vh dependent regulation of luminescence in V. harveyi occurs at the promoter of the lux operon [14,15]. While the amino acid sequence of LuxR Vh in V. harveyi is not similar to the Lux-R Vf protein in V. fischeri which also regulates luminescence [16], it is thought that LuxR Vh binds to the lux box and activates the expression of the luciferase genes. Since Lux-R Vh functions as a transcriptional activator in V. harveyi and VanT is required for protease expression in V. anguillarum, we wanted to determine whether any proteins bind to the regulatory regions (e.g. the lux box-like element and promoter) immediately upstream of empA.
We have previously shown that V. anguillarum cells grow rapidly in Atlantic salmon intestinal mucus [17,18] and that empA is strongly induced when cells are incubated at high density in mucus [17]. Further, we have demonstrated that empA transcription is RpoS-dependent [12]. Our findings also suggest that quorum sensing and undescribed autoinducer(s) help to regulate the expression of empA in V. anguillarum [12]. Additionally, Milton et al [19] previously identified a lux box-like element about 70 bp upstream of the translational start of empA, which we subsequently showed to span the transcriptional start site [12]. In this study, we examine, compare, and contrast the expression of metalloprotease encoded by empA in two strains of V. anguillarum (M93Sm and NB10). Differences in empA expression were investigated with regards to various growth conditions, the presence or absence of DNAbinding proteins, and putative regulatory proteins.

Gel shift analysis of the empA promoter
The possibility of DNA-binding proteins interacting with regulatory sequences of empA, were examined using a digoxigenin (DIG) labeled oligomer containing the lux boxlike element and empA promoter region. The location of double stranded oligomer relative to the promoter and coding sequences is shown in Fig. 1. Protein extracts from V. anguillarum M93Sm and NB10 cells incubated in LB20 and NSSM were mixed with the DIG-labeled oligomer (Fig. 2). NB10 and M93Sm cells were incubated in LB20 and NSSM for 3 h and protein extracts from cells at 0 and 3 h were prepared. DNA binding was observed in protein extracts from NB10 incubated in LB20 and NSSM ( Fig. 2A and 2B) at both 0 and 3 h. A 3.7-fold increase (from T = 0 h) in the amount of DNA binding was observed by 3 h in extracts from NSSM-grown NB10 cells (Fig. 2B, NB10 lane 2). An additional 1.4-fold increase in DNA binding was observed when the amount of protein was increased from 5 µg to 7.5 µg (Fig. 2B, NB10 lane 3). Although only a minor change in the amount of DNA binding (1.4-fold increase) was detectable by 3 h in NB10 extract from LB20-grown cells, a 4.7-fold decrease (from T = 0 h) in free DIG-labeled oligomer was observed ( Fig. 2A, NB10 lane 2). In addition, a 8.9-fold decrease in free DIGlabeled oligomer was detected when 7.5 µg protein was added ( Fig. 2A, NB10 lane 3). Specificity of protein binding to the DIG-labeled oligomer was tested by competitive inhibition of binding by the addition of excess amounts of an identical unlabeled oligomer. When excess unlabeled lux box-empA promoter oligomer was added to protein extracts from LB20-and NSSM-grown NB10 cells binding to the DIG-labeled oligomer was abolished. Further, the addition of unlabeled oct2A (a non-competitive oligomer) to protein extracts did not affect the gel mobility shift of the DIG-labeled oligomer caused by protein extracts from NB10 cells incubated either in LB20 or NSSM. In contrast to protein extracts from NB10 cells grown in LB20 or NSSM causing a gel shift, protein extracts from M93Sm cells incubated in either condition did not show any binding to the DIG-labeled oligomer ( Fig. 2A and 2B). Even when 7.5 µg of protein was used, a band shift was not observed. In a separate experiment, protein extracts from NB10 cells incubated in 3M or starved in NSS also caused a similar gel mobility shift for the DIG-labeled oligomer (Fig. 3). Additionally, even when 5 µg of protein extract from exponentially growing NB10 cells in LB20 was added to the DIG-labeled oli-gomer, a band shift was observed. Further, in this gel shift experiment, M93Sm protein extract was added as a negative control to the labeled DIG-labeled oligomer and no shift was observed.

Protease activity in LB20, NSSM, 3M, and NSS
Expression of empA in NB10 occurs in both LB20 and NSSM, while M93Sm exhibits empA expression only in NSSM [12]. Since a gel mobility shift was observed in NB10 protein extracts in all conditions of growth and during starvation, these cells were tested for protease activity in 3M and NSS in addition to LB20 and NSSM (Fig. 4). Protease activity was detected in NB10 cells after 1 and 3 h of incubation in NSSM or LB20, respectively (Fig. 4A). Maximum protease activity was observed at 4 h for both conditions of incubation. In addition, protease activity was observed in NB10 cells incubated in 3M or starved in NSS after 1 h and continued to increase during the next hour of the 4 h experiment. Throughout the incubation, NB10 cells in 3M produced 3-4 fold greater amounts of protease activity than cells in NSS. The amount of protease measured in NSS at 4 h was only 14 and 17% of the amount produced in LB20 and NSSM, respectively. In contrast, protease activity was observed in M93Sm cells only when incubated in NSSM (Fig. 4B). M93Sm cells did not produce protease activity in any of the other conditions examined.

Effects of vanT mutations on protease activity and gel mobility shift
Since vanT of V. anguillarum NB10 was cloned, sequenced and shown to be required for protease activity [13], vanT mutants of both M93Sm and NB10 were created by site directed mutagenesis and tested for protease activity in LB20 and NSSM. When the mutants (NB02 and M02) were incubated in NSSM for 3 h, a small amount of protease activity was detected (5 and 13% of the wild-type levels, respectively) ( Fig. 5). No protease activity was observed in M93Sm, M02, or NB02 incubated in LB20.
The vanT mutant strains (M02 and NB02) and their parental wild-types (M93Sm and NB10, respectively) were also examined for vanT transcription by RT-PCR (Fig. 6). Neither vanT mutant produced vanT transcripts. Interestingly, both wild-type strains produced vanT mRNA in both LB20 and NSSM.
To determine if VanT was binding to the lux box-empA promoter, protein extracts from NB02 were tested in the gel shift assay (Fig. 7). Protein extracts from NB02 cells incubated in NSSM and LB20 exhibited binding to the DIG-labeled oligomer similar to that observed with protein extracts from wild-type NB10 cells (compare to Fig.  2). These data show that VanT was not responsible for the gel mobility shift of the lux box -empA promoter oligomer.

Discussion
It has been demonstrated that the transcriptional activator, LuxR Vf , in Vibrio fischeri, activates expression of the lux operon via interaction with the lux box [15,16,20,21]. Although there is incomplete understanding of LuxR Vf interaction with the lux box, good evidence is available that shows a specific lux box sequence required for efficient LuxR Vf binding [21]. LuxR Vh from V. harveyi and Lux-R Vf from V. fischeri are not the same transcriptional activator proteins [22]. These proteins differ in their activation of the lux operon. In V. fischeri, LuxR Vf interacts with an acylhomoserine lactone autoinducer and then binds to the lux box to activate transcription of the lux Nucleotide sequence of the promoter region of empA Figure 1 Nucleotide sequence of the promoter region of empA. The 5'-coding sequence and immediate upstream/promoter sequence of empA is presented. The base-paired sequence indicates the annealed primers empA3W and empA3C ( Table 2). The transcriptional start site is indicated by +1, the rpoS-dependent promoter regions are indicated as -10 and -35 with the bases in italics [12], and the lux box-like element [19] is underlined. Additionally, the start translation site (ATG) is in bold with the translated product below and the ribosome binding site (RBS) is marked above (*).
genes [23]. In V. harveyi, the regulation of luminescence occurs via a phosphorelay signal transduction system, which involves LuxR Vh binding to the lux box independent of an autoinducer [24]. Additionally, a central response regulator, LuxO, has been shown to regulate the expression of luxR [24] in V. harveyi. Recently, the LuxR Vh homologue of V. fischeri was identified as LitR [25]. However, it is not known if V. fischeri uses a phosphorelay cascade to control LitR interaction with the lux operon.
In this study, two strains of V. anguillarum (NB10 and M93Sm) were examined for a empA promoter region binding protein. In all conditions, incubation with NB10 protein extracts resulted in a gel mobility shift to the DIGlabeled oligomer. In contrast, M93Sm extracts did not show any protein binding to the labeled DNA in any conditions tested. This was interesting because NB10 protease production occurs in all conditions, while M93Sm protease activity is detected only when the cells are incubated in mucus. Recently, VanT, the LuxR Vh homolog in V. anguillarum, was identified by Milton et al [13] and shown to positively regulate EmpA metalloprotease and biofilm production. We created vanT mutants in both M93Sm and NB10 and tested them for vanT expression, protease activity, and DNA binding. These two mutants, M02 and NB02, did not produce detectable vanT mRNA. However, both exhibited low levels of protease activity in mucus -  about 5% and 13% of wild type activity in mucus, respectively. For NB02, no protease activity in LB20 was observed. These results demonstrate that vanT is required for protease activity in a non-mucus containing medium (V. anguillarum NB10 in LB20), confirming the result reported by Milton et al [13]. Our results also show that in NSSM, vanT mutants still exhibit some protease activity, but VanT is required for full protease activity in both M02 and NB02 cells.
We tested a vanT mutant of V. anguillarum for the possibility of VanT binding to and activation of empA expression. Since the V. anguillarum NB10 strain expresses empA during stationary phase in all conditions tested and protein extracts from these conditions show DNA binding to the empA promoter region, it was hypothesized that VanT binds to this region. The vanT mutant, NB02, was analyzed for gel shift using protein extracts from cells incubated in LB20 and NSSM. Protein extracts from NB02 cells incubated in each condition still caused a band shift. This result suggests that VanT, the LuxR Vh homolog in V.
anguillarum, either does not bind to the empA promoter region containing the lux box-like element or perhaps VanT binds to a second protein that binds to this DNA sequence. The former possibility is more likely since the gel shift observed in NB02 is identical to that observed in NB10 extracts. Further, since the M93Sm vanT mutant, M02 still exhibits about 13% of the wild type protease activity in NSSM and does not exhibit protein binding in the gel shift analysis, it is probable that VanT does not bind to the empA promoter-lux box region. However, Croxatto et al. [26] demonstrated recently that purified VanT protein binds to the V. anguillarum NB10 promoter regions for empA, vanOU, and vanT. Examination of their data shows that VanT binds more strongly to the promoters for vanOU and vanT (binding observed with 0.3 µg purified VanT per 2 ng DIG-labeled DNA) than to the empA promoter (binding observed with 2.5 µg VanT per 2 ng DIG-labeled DNA). It should be noted that in our study only 5 µg of total protein extract was used in the gel shift analysis. That binding to the empA promoter is still observed in vanT mutants strongly suggests that an unidentified protein binds to the promoter-lux box region.

NSSM LB20
Previously, we have shown that empA in V. anguillarum M93Sm is positively regulated by at least two components of salmon gastrointestinal mucus [17]. In this study we show that NB10 cells exhibit protease activity under all nutritional conditions including starvation in NSS when at high density, whereas M93Sm only induces protease activity in mucus. RpoS is required for expression of empA in both M93Sm and NB10 [12]. This was demonstrated by the observation that rpoS mutants of M93Sm (M03) and NB10 (NB03) exhibit no protease activity when cells are incubated in either LB20 or mucus during stationary phase. Extracellular factors present in conditioned media showed both positive and negative effects on empA expression [12]. LB20 conditioned supernatants from M93Sm cause a more rapid and increased production of protease activity (in NB10 and M93Sm cells). However, conditioned LB20 from the luxS mutant, M01added to wildtype cells results in a more rapid induction of protease activity, which suggests that AI-2 may act as a negative regulator of empA.
In this study we confirmed that VanT is required for empA expression in NB10 when incubated in LB20; however, when NB10 or M93Sm cells are incubated in NSSM, a small percentage of protease activity remains in the vanT mutants. Our data support the observation by Croxatto et al [13] that VanT positively regulates empA expression. Further, the role of VanT as a transcriptional activator of empA via the lux box-like element has not been established. Our data show that VanT does not bind to the empA promoter lux box-like region, but is required for maximum protease activity. In addition, we show that in NB10 an unknown protein binds to the empA promoterlux box region that may increase protease activity in all conditions. In contrast, M93Sm protein binding to this region is not observed and protease activity is detected only when cells are incubated in mucus. In conclusion, the expression of the empA virulence gene in V. anguillarum is regulated transcriptionally by the alternative σfactor RpoS (σ S ) and the transcriptional activator VanT (a LuxR Vh homologue). Additionally, an unknown DNA binding protein binds specifically to the empA promoter lux box-like region in NB10 cells, but not in M93Sm cells. We hypothesize that this protein permits empA expression in the absence of the inducing factors in fish gastrointestinal mucus. A better understanding of how and why these two wild-type strains of V. anguillarum exhibit different levels of empA expression will further reveal how virulence genes are regulated in this fish pathogen.

Conclusions
Differences in empA expression between the two strains of V. anguillarum, M93Sm and NB10, correspond with the production of an unknown regulatory protein that binds to a 50 bp oligomer identical to the promoter-lux box region of empA. For NB10, protein binding to the oligomer correlated with the expression of empA in all incubation conditions. However, M93Sm, which expresses empA only in mucus, did not exhibit protein binding in this condition. Additionally, a null mutation in vanT (luxR homologue) results in severely decreased or no empA expression, but does not abolish protein binding (in NB10) to the oligomer. Our data suggest that while empA expression in both V. anguillarum M93Sm and NB10 is dependent upon RpoS and VanT, the two strains differ in production of an unknown protein that binds to the promoter-lux box-like region of empA. We also hypothesize that the presence of this DNA binding protein permits NB10 cells to express empA in the absence of fish gastrointestinal mucus.

Bacterial strains, plasmids, and growth conditions
All bacterial strains and plasmids used in this report are listed in Table 1. V. anguillarum strains were routinely grown in Luria-Bertani broth + 2% NaCl (LB20) [18,27] supplemented with the appropriate antibiotic, on a rotary shaker at 27°C. Escherichia coli SM10 was grown in LB10 (LB + 1% NaCl) on a rotary shaker at 37°C. Experimental media included: LB20, 3M (marine minimal media) [18,28], nine salt solution (NSS, a carbon-, nitrogen-, and phosphorus-free salt solution) and NSS plus 200 µg mucus protein/ml (NSSM) [18]. Gastrointestinal mucus was harvested from Atlantic salmon as previously described by Garcia et al. [18]. Overnight cultures of V. anguillarum were grown in LB20, centrifuged (9,000 × g, 10 min), and pelleted cells washed twice with NSS [18]. Washed cells were resuspended to appropriate cell densities in experimental media. Specific conditions are described in the text for each experiment. Cell densities were determined by serial dilution and plating on LB20 agar plates or by measuring optical density at 600 nm (OD 600 ). Antibiotics were used at the following concentrations: streptomycin (Sm), 200 µg/ml and chloramphenicol (Cm), 5 µg/ml.

Bacterial matings
Plasmids were introduced into V. anguillarum M93Sm and NB10 from E. coli SM10 by conjugation using the procedure described by Milton et al. [11,29,30]. Briefly, overnight cultures of V. anguillarum M93Sm or NB10 and E. coli SM10 containing the pNQ705-1 vanT clone, pNQVanT, were prepared and mixed using ratios of 1:1 or 3:1 (recipient: donor) in NSS plus 10 mM MgSO 4 . The cell suspension was vacuum filtered onto a 0.22 µm nylon membrane, which was placed on an LB15 (LB + 1.5% NaCl) agar plate and allowed to incubate overnight at 27°C. Following incubation, the cells were removed from the filter by vigorous mixing in NSS plus 10 mM MgSO 4 . The cell suspension (100 µl) was plated on LB20 Sm 200 Cm 5 (for M93Sm mutants) or on TCBS Cm 5 (for NB10 mutants) and allowed to incubate at 27°C until V. anguillarum colonies were observed (usually 16-24 h).

Site-directed mutagenesis of vanT
Site-directed mutagenesis was used to create gene interruptions within the structural gene of vanT. Primers (Table  2) were generated based on the vanT sequence for V. anguillarum NB10 (accession # AF457643). A 320 bp region from vanT was PCR amplified using Qiagen Taq DNA polymerase and cloned into the suicide vector, pNQ705, using SacI and XbaI restriction endonucleases to yield pNQVanT. The presence of the 320 bp empA-derived insert was confirmed by both PCR amplification and restriction analysis using SacI and XbaI. The mobilizable suicide vector, pNQVanT, was transferred into V. anguillarum by conjugation with E. coli SM10. E. coli SM10 contains the λ pir protein that is required for replication of pNQ705. Chloramphenicol resistant colonies were selected and screened for insertion within vanT. PCR and Southern blot analysis were used to confirm the incorporation of pNQVanT. For PCR analysis, a primer described previously by Milton et al. [11] complementary to the pNQ705 vector was utilized ( Table 2). The forward primer, SD vanT-Forward (Table 2) is complementary to a region upstream of the insertion. PCR products were analyzed by electrophoresis through a 1.0 % agarose gel in Tris-acetate EDTA (TAE) [31] buffer containing 0.2 µg/ml ethidium bromide. The gene was interrupted within the 320 bp region of vanT rendering the mutants resistant to chloramphenicol at 5 µg/ml. The resulting V. anguillarum vanT mutants were designated M02 (derived from M93Sm) and NB02 (derived from NB10) ( Table 1).

Gel mobility shift assay
Protein extracts were prepared from 2 ml of cells using the Sigma CelLytic™ B bacterial cell lysis/extraction reagent containing 0.2 mg/ml lysozyme, 1 mM DTT, 2 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride (PMSF). Briefly, the cells were pelleted from the experimental conditions by centrifugation at 9,000 × g for 10 min, 4°C. In addition, cells were lysed by multiple freeze-thaw cycles in crushed dry ice or in liquid nitrogen and then heated at 37°C. Insoluble cell debris was centrifuged at 12,000 × g for 10 min, 4°C. Protein concentration was determined using the Bradford assay (Bio-Rad).