Brucella suis urease encoded by ure1 but not ure2 is necessary for intestinal infection of BALB/c mice

Background In prokaryotes, the ureases are multi-subunit, nickel-containing enzymes that catalyze the hydrolysis of urea to carbon dioxide and ammonia. The Brucella genomes contain two urease operons designated as ure1 and ure2. We investigated the role of the two Brucella suis urease operons on the infection, intracellular persistence, growth, and resistance to low-pH killing. Results The deduced amino acid sequence of urease-α subunits of operons-1 and -2 exhibited substantial identity with the structural ureases of alpha- and beta-proteobacteria, Gram-positive and Gram-negative bacteria, fungi, and higher plants. Four ure deficient strains were generated by deleting one or more of the genes encoding urease subunits of B. suis strain 1330 by allelic exchange: strain 1330Δure1K (generated by deleting ureD and ureA in ure1 operon), strain 1330Δure2K (ureB and ureC in ure2 operon), strain 1330Δure2C (ureA, ureB, and ureC in ure2 operon), and strain 1330Δure1KΔure2C (ureD and ureA in ure1 operon and ureA, ureB, and ureC in ure2 operon). When grown in urease test broth, strains 1330, 1330Δure2K and 1330Δure2C displayed maximal urease enzyme activity within 24 hours, whereas, strains 1330Δure1K and 1330Δure1KΔure2C exhibited zero urease activity even 96 h after inoculation. Strains 1330Δure1K and 1330Δure1KΔure2C exhibited slower growth rates in tryptic soy broth relative to the wild type strain 1330. When the BALB/c mice were infected intraperitoneally with the strains, six weeks after inoculation, the splenic recovery of the ure deficient strains did not differ from the wild type. In contrast, when the mice were inoculated by gavage, one week after inoculation, strain 1330Δure1KΔure2C was cleared from livers and spleens while the wild type strain 1330 was still present. All B. suis strains were killed when they were incubated in-vitro at pH 2.0. When the strains were incubated at pH 2.0 supplemented with 10 mM urea, strain 1330Δure1K was completely killed, strain 1330Δure2C was partially killed, but strains 1330 and 1330Δure2K were not killed. Conclusion These findings suggest that the ure1 operon is necessary for optimal growth in culture, urease activity, resistance against low-pH killing, and in vivo persistence of B. suis when inoculated by gavage. The ure2 operon apparently enhances the resistance to low-pH killing in-vitro.


Background
A number of environmentally and medically important bacteria produce the enzyme urease (urea amidohydrolase) [1], which catalyzes the hydrolysis of urea, leading to the production of carbamate and ammonia. In an aqueous environment, the carbamate rapidly and spontaneously decomposes to yield a second molecule of ammonia and one of carbonic acid. The carbonic acid equilibrates in water, as do the two molecules of ammonia, which become protonated to yield ammonium hydroxide ions. The reaction results in an increased pH of the environment [reviewed in [2][3][4][5]]. In sites where microorganisms colonize epithelial surfaces, such as the normal flora of the oral cavity or intestines, or when certain pathogenic bacteria infect tissues, the metabolism of urea by microbial ureases can have a profound impact on tissue integrity, microbial ecology, and the overall health of the host.
The ureases of most microbes are composed of three subunits α, β, and γ that are encoded by ureA, ureB and ureC genes respectively. The plant jack bean produces a singlesubunit urease [12], whereas, in gastroduodenal pathogen H. pylori, the ureA and ureB genes are sufficient to encode urease. Nevertheless the UreAB subunits of H. pylori can be aligned with the UreABC subunits of other ureolytic bacteria and with the single polypeptide of the jack bean urease. The crystal structure of the Klebsiella aerogenes urease reveals a trimeric configuration [13]. Biochemical analyses of ureases by gel filtration have shown that other bacterial ureases are multimeric and probably have similar stoichiometry [4].
Ureases are structurally complex enzymes, and additional urease subunits are required for the production of a catalytically active holoenzyme in-vivo. Ureases are among the few enzymes that require nickel for activity. Biogenesis of a functional urease in prokaryotes requires the presence and expression of four urease accessory genes, ureDEFG. In vitro experiments using purified accessory proteins support the idea that UreE likely acts as a carrier of nickel [14] and that UreDFG form a chaperone-like complex that keeps the apoenzyme in a configuration competent to accept nickel [15].
Urease activity can be a critical factor in the colonization, persistence and pathogenesis of bacteria. Considering the products produced by urease, it would be logical to assume that one of the enzyme's functions is to allow nitrogen assimilation. In fact, urea represents an assimilable nitrogen source for bacteria that can colonize the human body and there is evidence suggesting that ammonia assimilation from urea occurs in-vivo. A significant proportion of the urea produced in the liver ends up in the intestines, where it can be hydrolyzed and assimilated by several different species of anaerobic, ureolytic bacteria [3]. Similarly, the oral bacterium, Actinomyces naeslundii can use urea as a primary nitrogen source for growth [6]. So there is little doubt that nitrogen acquisition as the result of urease activity can be important in the ecology of complex populations colonizing the human body. However, it is an open question as to whether the capacity to assimilate ammonia produced by urease contributes to the pathogenic potential of bacteria. Instead, it appears that the release of the strongly alkaline ammonia released by urease is a major cause of the damage to the host tissue, and in some cases, a key factor in persistence of pathogens [reviewed in [2]]. Jubier-Maurin et al., (30) identified the nikABCDE operon encoding the specific transport system for nickel in B. suis. Insertional inactivation of nikA strongly reduced the activity of the nickel metalloenzyme urease, which was restored by addition of nickel excess. Intracellular growth rates of the B. suis wild-type and nikA mutant strains in human monocytes were similar, indicating that nikA was not essential for this type of infection.
The Brucellae are gram-negative, facultative intracellular bacterial pathogens of a wide range of vertebrates [7]. This pathogen is the etiologic agent of the disease brucellosis and the pathological manifestations of brucellosis include abortion and sterility in animals [7], and meningitis, endocarditis, spondylitis and arthritis in humans [8]. Paulsen et al., [9] annotated the genome of B. suis strain 1330 (biovar 1), and discovered that unlike many other organisms, Brucella have two urease gene operons located on chromosome I (GenBank accession no. NC_004310). Urease activity is important for the nitrogen assimilation and persistence of other bacterial species like Helicobacter pylori [10,11]. We investigated the role of the two B. suis urease operons on the infection, intracellular persistence, growth, and resistance to low-pH killing. We report that the B. suis ure1 operon, in contrast to ure2, appears to be principally responsible for determining urease activity, optimum growth and resistance to low-pH killing in-vitro and persistence in-vivo.

Organization, and nucleotide and amino acid sequences of urease genes
The ure1 and ure2 operons are located on the chromosome I of B. suis strain 1330 (GenBank accession number NC_004310). The ure1 operon is 5284-bp long and composed of seven coding sequences (CDS). The ure2 operon is 6571-bp long and comprised eight CDS ( Figure 1). The ureA gene was the same in size in both operons (302-bp). All the other genes of ure2 operon were slightly longer than their counterparts in ure1 operon. The ureC gene was the longest in each operon (1712-bp in operon-1 and 1721-bp in operon-2). The G+C content of each ure gene was compared with that of its counterpart of the other operon and found not differ substantially between ure genes of operon-1 and operon-2 ( Table 1). The identity of each ure gene was compared with that of its counterpart of the other operon. The ureA, ureB, ureC, and ureG genes of the two operons exhibited 52 to 60% identity, whereas the ureD, ureE, and ureF genes did not share significant identity ( Table 1).
The deduced amino acid sequences encoded by the ureA, ureB, and ureC genes in both operons displayed great identity with the structural urease subunits of a vast range of organisms including Gram-positive bacteria, Gramnegative bacteria, photosynthetic bacteria, fungi, and higher plants (see Table 5). For instance, urease subunits of other organisms exhibited up to 81% identity with the UreC of ure1 and up to 69% identity with the UreC of ure2. The ureases of alpha and beta-proteobacteria exhibited the greatest identity with UreC of operon-1, whereas, the ureases of all species of Yersinia exhibited the greatest identity with the UreC of operon-2.
Real-time PCR assays produced amplicons in sizes exactly similar to the expected sizes for each ure gene (data not shown). The UreB subunit in the ure2 operon contains a predicted hydrophobic signal sequence and suggests that the subunit may localize in the periplasmic space. All other Ure subunits lack any signal sequences and were predicted to localize in the cytoplasm (data not shown).

Genomic characterization of generated mutant B. suis strains
Four mutant B. suis strains were generated by allelic exchange, i.e., 1330Δure1K, 1330Δure2K, 1330Δure2C, and 1330Δure1KΔure2C. The PCR assays produced a predicted 2.2-kb amplicon from the wild type B. suis strain 1330 and an approximately 3.2-kb amplicon from the mutant strain 1330Δure1Kwith the ureONE-Forward and ureONE-Reverse primers (see Table 6); a predicted 2.2-kbsize amplicon from the strain 1330 and an approximately 2.8-kb product from strain1330Δure2K with ureTWO-Forward and ureTWo-Reverse primers; and a predicted 2.9kb-sizeamplicon from the strain 1330 and an approximately 3.4-kb product from strain1330Δure2C with primers Ure-2-AB-Forward and Ure-2-AB-Reverse. The PCR assays with the primer pairs ureONE-Forward/ureONE-Reverse and Ure-2-AB-Forward/Ure-2-AB-Reverse confirmed that the double-mutant strain 1330Δure1KΔure2C carried a 575-bp deletion from the ure1DA region and a 1.2-kb deletion from the ure2ABC region ( Figures 1A and  1B).

Expression of urease, urease enzyme activity and growth rates of B. suis strains
Native polyacrylamide gel electrophoresis revealed urease activity at approximately 95-kDa from strains 1330, 1330Δure2K and 1330Δure2C, but not from strains 1330Δure1K or 1330Δure1KΔure2C (Figure 2).
In a quantitative urease assay, mutants 1330Δure1K and 1330Δure1KΔure2C exhibited 0 activity, the wild type and the mutant 1330Δure2K displayed maximal activity, and mutant 1330Δure2C showed slightly reduced activity ( Table 2). In qualitative urease assay, urease test broth started turning positive within 4 h after either strain 1330, 1330Δure2K, or 1330Δure2C were introduced, and acquired a bright pink color after approximately 24 h (Figure 3 and Table 2). In contrast, strains 1330Δure1K and 1330Δure1KΔure2C failed to cause a pink color in the urease test broth even after 96 h of incubation ( Figure 3 and Table 2).
Strains 1330Δure1K and 1330Δure1KΔure2C, both urease negative, grew approximately 25% slower than wild type strain 1330. In contrast, strains 1330Δure2K and 1330Δure2C, both urease positive, did not display any measurable differences in growth rate compared to strain 1330 ( Table 2).

Survival of B. suis strains in macrophage cell lines
When used to infect J774A.1 or H36.12a [Pixie 12a] mouse macrophage cell lines, the recovery of all the B. suis strains declined 2-3 log 10 cfu between 0 and 4 h postinoculation. During the next 20 h, all the B. suis strains increased 1-2 log 10 cfu. There were no significant differences between the wild type and the urease mutant strains in terms of their ability to replicate in macrophages (data not shown).

Survival of B. suis strains in BALB/c mice
Following an intraperitoneal inoculation, the recovery of ure mutants from spleens did not differ from the wild type strain at 6 wks post-infection (Table 3). When the mice were inoculated by gavage, one week after inoculation, strain 1330 was recovered from spleens ( Figure 4) as well as from livers ( Figure 5). When the mice were inoculated with strain 1330 supplemented with 10 mM urea, nearly 2.2 log 10 greater cfu was recovered from spleens and nearly 3.5 log 10 greater cfu was recovered from livers. However, when the mice were inoculated with strain 1330Δure1KΔure2C, with or without urea supplementation, no cfu were recovered from spleens ( Figure 4) but nearly 2.5 log 10 cfu was recovered from livers only when the inoculum was supplemented with 10 mM urea ( Figure  5).

Resistance of B. suis strains against low-pH killing
The wild type and the ure mutants did not differ with respect to the survival after 90 min incubation at pH 4.0 or 7.0 (data not shown). All the strains including the wild type were killed when incubated at pH 2.0 for 90 min ( Figure 6). When the strains were supplemented with 5 mM urea during incubation at pH 2.0, more than 6.0 log 10 cfu of strains 1330 and 1330Δure2K were recovered. In comparison to strain 1330, the recovery of the strain 1330Δure2C was nearly 1.5 log 10 lower at 5 mM urea concentration and nearly 1.0 log 10 lower at 10 mM urea. In contrast to strains 1330, 1330Δure2K and 1330Δure2C, strain 1330Δure1K was not recovered after incubation at pH 2.0 supplemented at any urea concentration ( Figure   6). Addition of urea did not change the pH of the incubation media.

Discussion
The ureA, ureB, and ureC genes of B. suis ( Figure 1) encode the γ, β, and α subunits respectively, and the urease holoenzyme of B. suis is likely to be assembled in a trimeric configuration. The total predicted mass of the B. suis urease holoenzyme (UreA+B+C) is 91-kDa. The native The schematic representations of the ure operons with corresponding Ure subunits, and deletion sites of mutant strains 1330Δure1K, 1330Δure2K and 1330Δure2C The DNA region deleted when making strain 1330 ure2C 1B polyacrylamide gel reveals urease activity at approximately 95-kDa ( Figure 2) and supports a trimeric configuration of this enzyme. The Brucella genome also contains ureDEFG genes ( Figure 1) in each of ure1 and ure2 operons and are predicted to produce the UreD, UreE, UreF and UreG proteins. Unlike many other microorganisms, Brucella contains two operons encoding urease subunits ( Figure 1) located on the chromosome I. Based on the similarities of G+C contents among genes in ure1 and ure2 operons, it is unlikely that any of the operons were acquired by horizontal gene transfer. The genes of the ure1 operon shared less than 60% identity with their counterparts of the ure2 operon (Table 1). In particular, the ureE and ureF genes of the ure1 operon did not share considerable similarity with those genes in the ure2 operon. Based on the relatively low identity among genes between ure1 and ure2 operons, it seems unlikely that they were the result of a recent duplication event. However, further analyses are required to confirm these predictions.
We generated a series of mutants by disrupting the first few genes encoding structural subunits of each urease. All seven genes of ure1 operon appear to be transcribed in a single direction. The gaps between individual ure genes are extremely small (Figure 1), so that all or most of the genes are possibly expressed under a single, common promoter -leading to a polycistronic mRNA. The ureG is the last gene of the ure1 operon. All three genes downstream of the ureG are transcribed in the opposite direction, from the complementary strand ( Figure 1). The closest non-ure Native 8% polyacrylamide gel assay with B. suis extracts    The deletion of genes within the ure1 operon caused the disappearance of urease activity on a native polyacrylamide gel ( Figure 2). Even though the ure2 operon was present in the ure1 mutant, it failed to produce urease activity as measured by either of two assays. In addition, a disruption of the ure2 operon did not have any impact on production of a detectable urease activity.
The deletion of genes within the ure1 operon caused the slower growth and loss of urease activity ( Table 2) suggest-

Recovery of
Brucella cfu from spleens one week after BALB/c mice were inoculated by gavage with wild type strain 1330 or strain 1330Δure1KΔure2C with or without urea supplementation Figure 4 Recovery of Brucella cfu from spleens one week after BALB/c mice were inoculated by gavage with wild type strain 1330 or strain 1330Δure1KΔure2C with or without urea supplementation. P value for the difference among mean values was <0.01. The mean values that share the same symbol do not differ from one another significantly; and the mean values designated by different symbols differ from one another significantly.  ing that the genes are necessary for maximal growth and urease activity of B. suis. Strains 1330Δure2K and 1330Δure2C made by deleting genes of ure2 operon did not display any change in growth rate or urease activity, suggesting that the genes of ure2 operon are not required for these functions. Overall, B. suis is apparently capable of exhibiting urease activity even without involvement of ure2 operon. This observation raises the question as to whether the genes within ure2 are actually being expressed. Our measurements using RT-PCR of RNA extracted from B. suis 1330 cultured in TSB (data not shown) suggest that all the genes in ure2 are being expressed. It is apparent that much more work will have to be done to determine the role of the ure2 operon. Except for a detectable role in resistance to acidic pH invitro, there were no other detectable phenotypes associated with the ure2 mutants under the various conditions employed. One obvious question is whether there is any post-translational interaction going on between the subunits of both ure operons. Given that there is no urease activity in an ure1 mutant, it is possible to conclude that the corresponding Ure2 subunits are not acting to restore Ure1 activity.
In addition to the ureolytic bacteria [3] that can use urea as a primary nitrogen source for growth, the urinary tract pathogen Ureaplasma ureolyticum [16] and some alkalophilic bacteria [17] can use ureolysis to generate ATP. Urea, particularly at millimolar concentrations can readily enter the cell [4]. Thus, in bacteria that utilize urea for nitrogen assimilation or ATP generation, the urease activity is expected to occur intracellularly. In most organisms, the ureases are found in the cytoplasm, although there is a report of urease membrane association and cell surface localization as well [4]. The ureases for which primary sequence information is available do not have characteristics consistent with being integral membrane proteins or secreted through the general secretory pathways. Consistent with the ureases of other bacteria, the deduced B. suis ure encoded subunits (with exception of UreB encoded by the ure2 operon) are predicted to localize in the cell cytoplasm (data not shown). Thus it may be possible that urease functions within the cytoplasm in order to assimilate nitrogen. Further work is needed to characterize the exact location of urease subunits and whether they play a role in nitrogen uptake.
The ure mutants did not exhibit any decline in persistence in BALB/c mice when inoculated intraperitoneally ( Table  3), suggesting that the urease activity is not critical once the pathogen has entered the host. Using ure mutants of Bordetella bronchiseptica, Monack and Falkow [18] Recovery of Brucella cfu from livers one week after BALB/c mice were inoculated by gavage with the wild type strain 1330 or strain 1330Δure1KΔure2C with or without urea supplementation Figure 5 Recovery of Brucella cfu from livers one week after BALB/c mice were inoculated by gavage with the wild type strain 1330 or strain 1330Δure1KΔure2C with or without urea supplementation. P value for the difference among mean values was <0.025.
The mean values that share the same symbol do not differ one another significantly; and the mean values designated by different symbols differ one another significantly. reported that urease is not essential for the colonization of the guinea-pig respiratory and digestive tracts. One of the common ways that humans acquire brucellosis is through consumption of contaminated milk. Brucella needs to have a mechanism to resist the low-pH killing in the gastrointestinal tract. When the mice were inoculated by gavage, the wild type strain was recovered from livers and spleens, whereas, strain 1330Δure1KΔure2C was not. It is noteworthy that when the infecting doses of B. suis were supplemented with urea, the recovery of strain 1330 from livers and spleens was increased, but the mutant strain 1330Δure1KΔure2C was recovered only from livers. These findings suggest that the urease activity and sufficient substrate is needed for B. suis to cope with the low pH in the gastrointestinal tract i.e. either stomach or upper intestine. Similar observations have been recently reported by Sangari et al., [37] in ure-mutant strains of B. abortus. Thus, similar to the urease of H. pylori [10,11], that of B. suis appears to be a factor in coping with the pH of the gastrointestinal tract. Even though the ure1, ure2 mutant B. suis exhibited no urease activity, urea supplementation raised the recovery of this strain from livers. This is possible because other bacterial species in stomach and gastro-intestinal tract may have utilized the supplemented urea to reduce the acidity and facilitated enhanced survival of Brucella. It is worth noting that inoculation via an oral route versus direct inoculation in the stomach may be more representative of a natural infection. However, we chose to deliver a known dose of Brucella and measure uptake through the gastro-intestinal tract (i.e., into the spleen and liver).
For an intracellular bacterium like Brucella that replicates inside phagocytic cells associated with the various tissues including the reproductive tract [19], urease may not be important for regulating the pH in an intracellular milieu. Correspondingly, the lack of differences between wild type and ure mutants inoculated intraperitoneally during splenic clearance was not surprising (Table 3). Nevertheless, in the case of H. pylori and perhaps other organisms, urease is an important factor in survival in severely acidified environments, probably because neutralization of the extracellular milieu around the cells is needed to prevent irreversible membrane damage. Acidification and subsequent alkalinization in the phagosomal compartment is a prerequisite for a successful Brucella infection in-vitro [20].
Survival of B. suis strains 1330, 1330Δure1K, 1330Δure2K, and 1330Δure2C after incubation at pH 2.0 with or without urea Figure 6 Survival of B. suis strains 1330, 1330Δure1K, 1330Δure2K, and 1330Δure2C after incubation at pH 2.0 with or without urea. At each urea concentration, the P value for the difference among mean cfu was <0.005. At each urea concentration, the mean values designated by different symbols differ one from another significantly; and the mean values that share the same symbol do not differ one from another significantly.
However, the amount of urea present in macrophages maybe low enough as not to cause a significant pH change. As such, the impact of urease activity on macrophage pH can be considered insignificant, and therefore, similar survival rates in macrophages between wild-type and ure mutants was not unexpected.

Conclusion
The B. suis genome contains two operons encoding urease. The genes in the ure1 operon are important for optimal growth in culture, and indispensable for urease activity, resistance to low-pH killing and survival of the pathogen when inoculated by gavage. The genes in the ure2 operon seem to be dispensable for the above functions, but slightly enhance the resistance to low-pH killing measured in-vitro. The apparent lack of urease activity encoded by ure2 is consistent with the observation by Hooper and Berg who reviewed over 20 microbial genomes with respect to gene innovation following gene duplication (38). They concluded that most gene copies are deleted but of the ones retained, they function in weak or ancillary roles. Thus it is possible that ure2 in B. suis is diverged enough to encode a new function that has yet to be defined.

DNA and protein sequence analyses
The nucleotide sequence of the urease genes was analyzed with DNASTAR software (DNASTAR, Inc., Madison, Wis.). The destination of the deduced proteins upon translation and processing was predicted using the Subloc v1.0 server of the Institute of Bioinformatics of the Tsinghua University http://www.bioinfo.tsinghua.edu.cn/. Identity of the ureA, ureB, and ureC genes of B. suis ure1 and ure2 operons with sequences of the EMBL/GenBank/ DDBJ databases was analyzed using the BLAST software [22] at the National Center for Biotechnology Information (Bethesda, MD).

Bacterial strains, plasmids, and reagents
B. suis strain 1330 was obtained from our culture collection. Escherichia coli strain Top10 (Invitrogen Life Technologies, Carlsbad, Calif.) was used for producing plasmid constructs. E. coli were grown in Luria-Bertani (LB) broth or on LB agar (Difco Laboratories, Sparks, MD). Brucella were grown either in Trypticase soy broth(TSB) or on Trypticase soy agar (TSA) plates (Difco) at 37°C in the presence of 5% CO 2 as previously described [23]. Theplasmids used in this study are listed in Table 4. Bacteria containing plasmids were grown in the presence of ampicillinor kanamycin at a 100-μg/ml concentration ( Table 4).
All experiments with live Brucella were performed in a Biosafety Level 3 facility in the Infectious Disease Unit of the Virginia-Maryland Regional College of Veterinary Medicine per standard operating procedures approved by the Centers for Disease Control and Prevention.

Recombinant DNA methods
Genomic DNA was isolated from B. suis strain 1330 by use ofa QIAGEN blood and tissue DNA kit (QIAGEN Inc.,   [25].

Mutation of ure1 operon
A 2,241-bp region including the whole length of the ure1D, ure1A, and ure1B genes and a portion of ure1C gene ( Figure 1) was amplified via PCR using the genomic DNA of B. suis strain 1330 and the primers UreaseONE-Forward and UreaseONE-Reverse (RansomHill Bioscience, Inc., Ramona, CA) (see Table 6). The amplified gene fragment was cloned into the pCR2.1 vector of the TA cloning system (Invitrogen) to produce plasmid pCRure1. From this plasmid the ure1 region was isolated by EcoRI digestion and cloned into pGEM-3Z (Promega) and the resulting 5.0-kb plasmid was designated pGEMure1. The suicide vector pGEMure1K was constructed as follows: the plasmid pGEMure1 was digested with NcoI to delete a 575-bp region from the ure1 region. The NcoI sites on the 4.4-kb plasmid were filled in by reaction with Klenow enzyme and ligated to the1.6-kb PvuII fragment of pUC4K (also blunt ended) containing the Tn903 npt gene [26], which confers kanamycin resistance(Kan r ) to B. suis. The resulting suicide vector was designated pGEMure1K. The E. coli Top10 cells carrying the recombinant plasmid were picked from TSA plates containing kanamycin (100 μg/ ml).
One ug of pGEMure1K was used to electroporate B. suis strain 1330; several colonies of strain 1330 were obtained from a TSA plate containing kanamycin (100 μg/ml). These colonies were streaked on TSA plates containing ampicillin (100 μg/ml) to determine whether a single-or double-crossover event had occurred. Five of the colonies did not grow on ampicillin-containingplates, suggesting that a double-crossover event had occurred. PCR with the primers UreaseONE-Forward and UreaseONE-Reverse (see Table 6) confirmed that a double-crossover event had taken place in all five transformants. One of these strains was chosen for further analyses and designated 1330Δure1K.

Mutation of ure2B and ure2C
A 2,214-bp region including the whole length of the ure2A and ure2B genes, and a portion of the ure2C gene was amplified via PCR using the primers UreaseTWO-Forward and ureaseTWO-Reverse (see Table 6). The amplified gene fragment was cloned into the pCR2.1 vector to produce plasmid pCRure2ABC. From this plasmid, the ure2ABC region was isolated by BamHI and XbaI digestion and cloned into the same sites of plasmid pGEM-3Z (Promega). The resulting 5.0-kb plasmid was designated pGEMure2ABC. The suicide vector pGEMure2ABCK was constructed as follows: the plasmid pGEMure2ABC was digested with SacII to delete a 940-bp region from the ure2ABC region, sticky sites filled in with Klenow enzyme, and ligated to the 1.6-kb PvuII fragment of pUC4K. The resulting suicide vector was designated pGEMure2ABCK. One microgram of pGEMure2ABCK was used to electroporate B. suis strain 1330, and a transformed strain containing a double-crossover event was verified by PCR and designated 1330Δure2K.
Mutation of ure2A, ure2B and ure2C A 2,923-bp region including the whole length of ure2A and ure2B genes, and a portion of ure2C gene (Figure 1) was amplified via PCR using the primers Ure-2-AB-Forward and Ure-2-AB-reverse (see Table 6). The amplified gene fragment was purified using a Qiagen PCR purification kit (Qiagen), digested with BamHI and XbaI, cloned into the same sites of plasmid pGEM-3Z (Promega) to produce the 5.7-kb plasmid pGEMure2ABC-2. The suicide

Real-time PCR assays
RNA was isolated from broth cultures of B. suis strain 1330 by the procedure described previously [31]. After a 75% ethanol wash, the dried RNA pellet was resuspended in RNase-and DNase-free water (Sigma). The concentration of the RNA was be determined with the RiboGreen RNA Quantitation kit (Molecular Probes). Genomic DNA was digested with RNase-free DNase (Ambion), and precipitated with GlycoBlue (Ambion). RNA samples not treated with reverse transcriptase was also subjected to PCR to measure the level of contamination from genomic DNA. For each RNA sample, the control transcript (sigA) and the target mRNA were reverse-transcribed using the ReverTra Dash kit (Toyobo). The cDNA was amplified using the Light Cycler (Roche) in conjunction with the DNA Master SYBR Green I kit (Roche). Primers specific for the internal standard of ure genes were purchased from Sigma-Genosys. The target cDNA was normalized internally to the sigA cDNA level in the same sample [32,33].

Growth rates of B. suis
Single colonies of strains 1330, 1330Δure1K, 1330Δure2K, 1330Δure2C and 1330Δure1KΔure2C were grown at 37°C for 72 h to stationary phase in 5 ml of TSB. These cultures were used to inoculate 25 ml of TSB in a Klett side-arm flask to 12 to 16 Klett units. Cultures were grown at 37°C at 200 rpm; Klett readings were recorded every 2 h in a Klett-Summerson colorimeter (New York, NY).

Native polyacrylamide gel electrophoresis
Extracts were prepared from B. suis strains 1330,1330Δure1K, 1330Δure2K, 1330Δure2C and 1330Δure1KΔure2C using glass beads and vortex, in Tris-HCl 30 mM, pH 8.0). Buffer without SDS or mercaptoethanol were added to the extracts and the extracts were loaded into the gel that did not contain SDS. After running was complete, the gel was placed in 0.02% cresol red-0.1 EDTA, washed several times until it became yellow, and incubated with 1.5% urea at room temperature until pink bands appeared, i.e., the positive urease reaction.

Urease enzyme activity of strains-qualitative
Fifty μl of culture (grown for 72 h in TSB) volumes of strains 1330, 1330Δure1K, 1330Δure2K, 1330Δure2C and 1330Δure1KΔure2C were used to inoculate 5 ml volumes of urease test broth (Difco). The contents were incubated at 37°C with 200 rpm shaking. At 8, 24 and 48 hours after incubation, the cultures were centrifuged to remove the cells. The color change in the clarified urease test broth was measured using a Klett-Summerson colorimeter. The native color of the urease test broth was used as a blank.

Urease enzyme activity of strains-quantitative
The Coomassie Brilliant Blue G, TRIS, NADPH, 2-oxoglutarate and glutamate dehydrogenase (from bovine liver) were from Sigma-Aldrich. Urea was obtained from Qiagen. All other reagents were of analytical grade.
The concentration of protein in the extracts was determined using a Bradford-modified assay [34]. All assays (final volume of 2 mL) were performed in 31 mM Tris-HCl pH 8.0 buffer at 28°C, using a Beckman DU 800 UV/ Vis spectrophotometer, with a stirred, temperature-controlled multi-cell holder. Urease activity was determined using a coupled assay with glutamate dehydrogenase [35] Glutamate dehydrogenase 12 U/mL was incubated with 250 μM NADPH for 5 min. 2-Oxoglutarate (1 mM) and B. suis extracts were added and the reaction was followed at 340 nm. The observed decrease in absorbance monitored during this period is due to nonspecific oxidation of NADPH by several enzymes in the extract. When the absorbance was stable, urea (10 mM) was added and the decrease in absorbance at 340 due to urease activity was measured for 5 min. Initial rates were calculated from the linear portion of the curves, by linear regression using the least squares method. The absorption coefficient used for NADPH was 6.22 M -1 cm -1 [36]. The volume of extracts was varied and the specific activity of urease was calculated. One unit of urease activity was defined as the amount of enzyme that hydrolyzes 1 μmol of urea per min. Specific activities were calculated as units of urease per mg of protein in the extract.