Genetic Manipulation Of B Pseudomallei Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Generally, few stages are involved in the mutagenic process of allelic replacement system which includes the cloning of targeted allele or region to be deleted into a vector followed by the mobilization of the gene replacement vector into the host bacterium. Subsequently, the allele is integrated into the chromosome most likely within the complementary gene and finally the selection and verification of putative mutant strains underwent double homologous recombination.

To this end, the gene replacement vector, pBCSA01 has been successfully transferred into B. pseudomallei UKMS-01 through the selection of transconjugants by utilizing chloramphenicol cassette located on the suicide plasmid. Entailing the conjugal transfer of gene replacement vector pBCSA01 from E. coli, the suicide plasmid must have integrated into the recipient's chromosome via homologous recombination in order for it to be maintained. The integration, which is also termed as first crossing over or first allele exchange is promoted by inoculating a single colony of merodiploid clone in LB without NaCl (Blomfield) broth and incubated at 24°C for overnight. In this study, it has been found that the efficiency of latter counter-selection is negatively correlated with cell density in this isolate. Therefore, bacterial growth at optimum temperature (37°C) is avoided at this stage to allow for plating at low cell density that consequently facilitates the selection of mutants.

The resolution of the resulting strain, B. pseudomallei UKMS-01 bcsA::pBCSA01 merodiploid was achieved through sucrose counter-selection which is critical for second cross-over and vector loss. After 16 hours of growth in Blomfield broth without any selection pressure, the overnight culture was serially diluted and plated on Blomfield agar containing 10% [w/v] sucrose. Colonies observed on Blomfield containing sucrose agar plates after 72 hours of incubation at 24°C were patched in duplicate on LB agar supplemented with 150 μg/ml chloramphenicol and Blomfield agar containing 10% [w/v] sucrose. A double cross-over mutant is phenotypically catS and sacBR due to the lost of chloramphenicol cassette and sacB gene located on the plasmid backbone. Genomic DNAs were extracted from five randomly pick colonies which are catS and sacBR and subjected to PCR validation to differentiate the deletion mutant from wild type strain which is also phenotypically catS and sacBR and merodiploid strain which experienced spontaneous mutation in both markers.

All five putative deletion mutants were confirmed via PCR analysis. A total of two primers, namely bcsA UVe 5'F and bcsA DVe 3'R (Table 3.2) were used for this purpose. These primers were designed to anneal just outside the deleted region in the chromosomal DNA with the location of the each primer's annealing site shown in Figure 4.3. Since the primers used are complementary to the flanking region outside the deleted region, they are expected to anneal to chromosomal DNAs of both wild type and deletion mutant strains. Therefore, the resulting PCR products of both strains should be of variable lengths. The presence of deletion in bcsA gene should yield an expected PCR fragment of 1.8 kb while the wild type chromosomal DNAs should give a larger fragment of 4.2 kb when the same primers were used (Figure 4.3). Through PCR screening, one out of five analyzed colonies was found to contain the mutation. PCR of the wild type B. pseudomallei UKMS-01 and its deletion allele are shown in Plate 4.4. B. pseudomallei UKMS-01 strain carrying the desired deletion was selected and designated as B. pseudomallei UKMS-01 ∆bcsA01.

The PCR fragment derived from deletion strain was cloned and sequenced to confirm that the correct deletion was made. A total of three sequencing primers were used (Table 3.2), namely bcsA USVe 5'F, bcsA USVe 3'R and bcsA USeq5'F (Figure 4.3). The analysis of BLASTX results showed that the upstream fragment of mutant allele (sequencing primer: bcsA USVe 5'F) was from gene which encoded hypothetical protein BPSS 1578 while the downstream fragment of the mutant allele (sequencing primer: bcsA USVe 3'R) was from gene which encoded hypothetical protein BPSS 1576. Connectively, the sequencing results have validated the correct deletion of bcsA gene in B. pseudomallei UKMS-01. The illustration of sequencing result is shown in Figure 4.5 and 4.6.

4.1.5 The use of pDM4 to create an unmarked deletion mutant in B. pseudomallei

Taken together, results obtained from mutagenesis system employed in this study demonstrate that pDM4 is suitable for allelic replacement in B. pseudomallei and most importantly, this vector can be recycled to introduce multiple mutations into the same strain. The successful excision of both resistance determinant (catR) and counter selection marker (sacBS) from the mutant strains allows for the reutilization of both markers for subsequent rounds of genetic selection without encountering potential intricacies due to marker effects on bacterial physiology. Due to previous experience with replicative vector in generating deletion mutants, we noticed that non-replicative vector enhance the effectiveness of plasmid curing.

Gene replacement vector pDM4 and its derivatives carry a R6K origin of replication which is suicidal in B. pseudomallei strains. Therefore, the low copy number of plasmid (~ 1 per cell) has comparatively eased the curing process and thus smoothen the screening of double cross-over mutants. Nevertheless, we have routinely recovered false positive transformants on selection plates which were either antibiotic-tolerant or sucrose-tolerant cells which did not undergo double cross-over.

Counter-selectable markers have been a common yet efficacious tool in the studies of fundamental genetics and functional genomics. They play a direct role towards the successful construction of unmarked mutants via genetic manipulation, which are required for functional investigation of numerous unannotated ORFs in the genomic databases, for better understanding of the mechanism of pathogenicity and ultimately the development of live attenuated strains for vaccination purpose (Reyrat et al., 1998). Among the array of counter-selectable markers, the fusaric sensitivity system (tetAR), streptomycin sensitivity system (rpsL) and the sucrose sensitivity system (sacB) are the most popular ones that have been proven to be most useful (Reyrat et al., 1998).

In this study, the latter counter-selection of merodiploid was mediated by sacB gene presents on the backbone of pDM4 vector. Although sacB-mediated mutagenesis has been previously reported for some Burkholderia spp. (Brown et al., 2004; Chan and Chua, 2005; Essex-Lopresti et al., 2005; Ulrich et al., 2005), it is indeed a problematical and leaky counter-selectable marker in some instances (Barrett et al., 2008). Furthermore, the chromosome of B. pseudomallei was found to harbor an endogenous copy of sacB gene (BPSS0543; (Holden et al., 2004) that obviates the use of sucrose sensitivity system for counter-selection during allelic replacement. Consequently, many studies have used either alternative counter-selection system including tetAR, rpsL, pheS, thyA, and ccdB (Reyrat et al., 1998) or worked with potentially polar single cross-over mutants (Logue et al., 2009). However, a recent report by Logue et al. (2009) demonstrated that sacB counter-selection is possible in B. pseudomallei under optimized conditions (10% [w/v] sucrose, no NaCl, 24°C). For B. pseudomallei strains which display sucrose-resistance, it was found that the expression of pDM4-encoded sacB is relatively higher to the chromosomally sacB expression under the optimized conditions and after extended incubation. Nevertheless, most of the sequenced B. pseudomallei strains are naturally phenotypically sucrose-resistant despite of the presence of chromosomally encoded sacB gene.

As a whole, pDM4-mediated genetic system is reliable for generating unmarked deletions in B. pseudomallei. There are two drawbacks in this genetic system, one being the use of non select-agent compliant antibiotic resistance marker (catR), and secondly the lack of reporter gene. According to the select-agent guidelines, the choice of antibiotic resistance markers in the genetic manipulation of B. pseudomallei is greatly restricted. As mentioned before, the only approved antibiotic resistance markers for B. pseudomallei are zeocin, gentamicin and kanamycin. Therefore, the utilitarian of chloramphenicol resistance cassette can be possibly replaced by kanamycin resistance gene to fulfill compliant and responsible genetic manipulation of the select agent B. pseudomallei. The lack of reporter gene has hindered the study of gene expression and regulation of targeted biological system, for instances through the isolation of transcriptional or translational lacZ fusions, transcriptional luciferase (lux) fusions and tagging with green fluorescent proteins (Choi et al., 2008). Besides that, the addition of reporter gene also facilitated the screening of mutants by allowing visual detection of B. pseudomallei positive transformants.

4.2 Characterization of B. pseudomallei UKMS-01 ∆bcsA mutant

4.2.1 Morphotype screen via Congo red and Calcofluor binding

Over the years, Congo red and Calcofluor have been the most commonly used binding dyes in the detection of cellulose secreted from bacterial colonies on agar plates. Congo red has been shown to bind β(1→4)-D-glucopyranosyl units as well as other extracellular matrix components in several bacteria whilst Calcofluor white has an affinity towards β(1→4) and β(1→3) glycosidic bonds of polysaccharides. Congo red has an affinity to cellulose and curli via non-covalent bonding. Therefore, growth of bacteria which express cellulose and/or curli on agar medium supplemented with Congo red dye often display characteristic colony morphology described as Rdar morphotype system. This system has been vastly described in E. coli, S. typhimurium and other Enterobacteriaceae (Zogaj et al., 2001; Zogaj et al., 2003; Romling, 2005).

The rdar morphotype is the abbreviation for red, dry and rough and was used to elaborate bacterial colony which displays a red, dry and rough morphology on Congo red agar plates. Interestingly, differential expression patterns of extracellular matrix components lead to the development of various type of distinct colony morphology. The characteristic colony morphology display in relation the expression of extracellular matrix components are summarized in Table 4.1.

In this study, Congo red-mediated morphotype system was used to phenotypically characterize the ability to produce cellulose in both B. pseudomallei UKMS-01 (wild type) and B. pseudomallei UKMS-01 ∆bcsA strains. Theoretically, the knockout of bcsA gene which encoded for the catalytic subunit of cellulose synthase in mutant strain will suppress the cellulose biosynthesis giving rise to a change in the exopolysaccharide composition of the bacteria. The absence of cellulose causes different absorbance spectrums of Congo red and leads to the development of distinct colony morphology between wild type and mutant strains. B. pseudomallei UKMS-01 form colonies that display a morphology most closely resembling the pdar morphotype when grown on Congo red agar plates at 37°C (Figure 4.7), indicating production of cellulose only. Strangely, knock-out of bcsA in mutant strain did not abolish the pdar morphology; instead was indistinguishable from colony morphology of wild type strain (Figure 4.8).

Calcofluor (fluorescent brightener 28), a non-specific fluorochrome is another indicator of cellulose production. Fluorescence of colonies that were grown on agar plates containing Calcofluor were detected and visualized under a 366 nm UV light source. Calcofluor phenotype is analyzed by comparing fluorescence intensity of experimental strain relative to a control strain with known calcofluor binding capacity. In our case, bcsA knockout mutant served as the control strain. Under long wave UV light, both wild type and mutant strains fluoresced with low intensity, only at the edge of the colony on calcofluor agar plates (Figure 4.9).

The perplexing outcome of the morphological screen in both wild type and mutant strains could be due to several plausible reasons. Firstly, the non-specificity of Congo red dye might give a false indicator of cellulose production. It has been previously reported in a Congo red binding assay of Pseudomonas isolates whereby 100% of the Congo red binding material produced was found to be non-cellulosic (Spiers et al., 2003). This is further supported by Calcofluor staining whereby only minimal fluorescence intensity was visualized, suggesting low level of cellulose production in both wild type and ∆bcsA strains. Therefore, the slight change in exopolysaccharide composition might not be sufficient per se to cause pronounce and discernible differences in colony morphology. Furthermore, previous analysis of colony morphology has concluded it to be a scanty indicator of biofilm formation and cellulose production (Ude et al., 2006). Besides that, the small amount of cellulose detected from ∆bcsA strain through calcofluor staining strongly indicates that bcsA gene might not be the only gene responsible for cellulose biosynthesis in B. pseudomallei UKMS-01.

4.2.2 Electron microscopy studies on B. pseudomallei wild type and ∆bcsA strains

Scanning and transmission electron microscopy studies were employed in an attempt to visualize cellulose on the surface of B. pseudomallei UKMS-01 and to investigate cellulose production in both wild type and ∆bcsA strains as described in Materials and methods ( and For scanning electron microscopy, the wild type strain showed a very densely packed gum-like matrix that wrap on the bacterial surface and between cells (Figure 4.10). This matrix appears to be relatively dense though unstructured, in which the adhering bacteria were embedded. On the other hand, the bcsA knockout mutant is characterized by less dense and more fibrous structures between cells (Figure 4.11). From the images, two distinct populations of cells either aggregated or nonaggregated were observed. The aggregated cells are mostly covered by extracellular matrix while planktonic cells appear as dispersed individuals (Figure 4.12). This observation is consistent with the finding that the expression of extracellular matrix is not homogenous within a single population of bacterial cells (White et al., 2008) and extracellular matrix is a shared among bacterial cells. Despite differences in the structure of the exopolysaccharide network as shown in the two strains, the nature of the matrix, whether it is consists of curli or cellulose or both and other components remains elusive. However, a preliminary hypothesis can be drawn whereby the knockout of bcsA gene altered the expression patterns of extracellular matrix, revamp the exopolysaccharide composition and subsequently leads to modification of the physical properties of extracellular matrix (biofilm formation).

For transmission electron microscopy, the wild type and ∆bcsA strains were both covered with an extensive layer of exopolysaccharide (Figure 4.13 and Figure 4.14). No significant difference was observed from the micrographs. Apart of that, the micrographs also revealed that the cell population of both strains has a tendency to live as a multicellular unit. Comparatively, scanning electron microscopy is a more appropriate technique to investigate the cellulose production in both strains since it allows the visualization of the architecture of extracellular matrix.

4.2.3 Quantification of cellulose production by digestion of extracellular matrix with cellulase

The binding of Congo red and Calcofluor to cellulose is not literally specific and therefore cellulose production in both wild type and ∆bcsA strains were quantified by cellulase digestion. Firstly, cellulose was extracted from lyophilized cells of plate-grown colonies by incubation with acetic-nitric reagent. The addition of acetic-nitric reagent destroys all other polysaccharide materials except for crystalline cellulose. The isolated polymer was then treated with cellulase produced by Aspergillus niger. Endoglucanases are the major cellulolytic components in this enzyme complex (Vidmar et al., 1984) which is capable of catalyzing the hydrolysis of endo-1,4-β-D-glycosidic linkages in cellulose, lichenin and barley glucan. Consequently, the amount of glucose released by cellulolytic activity in addition to glucose present along with the enzyme was determined by colorimetric method using glucose assay kit.

By using glucose assay kit, enzymatic determination of glucose was achieved in a series of reactions. Firstly, glucose released from digestion was oxidized to gluconic acid and hydrogen peroxide by glucose oxidase. Then, hydrogen peroxide reacted with o-dianisidine in the presence of peroxidase to form a coloured product. Upon oxidation, o-dianisidine changed from colourless to brownish. Sulfuric acid was added to convert oxidized o-dianisidine to a more stable pink-coloured product and the intensity of pink colour was measured by absorbance at 540 nm. The absorbance reading is proportional to the glucose concentration from standard curve (Appendix).

The amount of released glucose was determined after one hour incubation with cellulase in citrate buffer at 55°C. The respective glucose concentrations were summarized in Table 4.2. It can be observed that approximately 109.3 μg of glucose was detected from cellulase solution only suggesting that cellulase enzyme alone carries high residual of glucose molecules. Therefore, the estimated concentrations of released glucose in wild type sample treated with cellulase and ∆bcsA sample treated with cellulase are 2.45 μg and 3.1 μg respectively. The quantitative assay confirmed the findings by Calcofluor staining whereby i) cellulose is produced at very low level by both strains as proven by the breakage of β(1→4)-glycosidic bonds and the released of glucose units. ii) The mutant strain does produced cellulosic polymer despite sequencing confirmed the deletion of bcsA gene which indicates that bcsA gene is not the only gene responsible for cellulose production.

4.2.4 Roles of cellulose in biofilm formation

A positive correlation between cellulose and biofilm formation was first established entailing findings in which cellulose is another structural component of the extracellular matrix of E. coli, Salmonella spp., and Pseudomonas spp. biofilms (Zogaj et al., 2003). It has been previously reported in Pseudomonas spp. that cellulose matrix-based biofilms were generally sturdier than non-cellulosic biofilms (Ude et al., 2006). The increased strength of cellulose matrix-based biofilm may lies in the fact that β(1→4)-linked hexose are on average more rigid and it's hydrophobic nature tends to exclude water to form larger structures (Sutherland, 2001a). In a separate study, cellulose together with other polysaccharide, short adhesins and pili were shown to contribute to biofilm maturation, a critical step towards biofilm expansion (Sandra et al., 2007 ).

In this study, we investigated the ability of both wild type and mutant strains to display biofilm behavior (cell aggregation, adherence to glass and plastic surfaces) in four different medium (M9 minimal medium supplemented with glucose, LB without NaCl broth, LB without NaCl broth supplemented with 0.2% glucose and LB broth supplemented with 0.2% glucose). Biofilm formation was measured by assessing the ability of both strains to attach to abiotic surfaces such as glass and polyvinylchloride whilst cell-cell interaction was evaluated by a cell aggregation sedimentation test. Besides that, the biofilm formation in glass tubes containing different medium were also visually inspected.

Attachment assays of the two bacterial strains were conducted on 96-well polystyrene microtiter plates. An overnight culture of bacterial strain was diluted 1:10 in 200 μl fresh medium. Biofilm formation and attachment were allowed to occur for overnight at 37°C. The attachment of cells to plate surface was quantified by dissolving crystal violet-bound bacteria in DMSO and absorbance at 595 nm was measured. Surface attachment of both strains in all four different medium was depicted in Figure 4.15a and 4.15b. In LB medium supplemented with 0.2% glucose and LB without NaCl medium supplemented with 0.2% glucose, both strains formed an adherent biofilm whereby bacterial cells are found adhered to the bottom of the wall. The adherent ability was not affected in the ∆bcsA mutant strain (P > 0.05; paired t tests). In LB without NaCl medium, pellicle was observed at the air-liquid interface in each well of both strains. However, no significant different in the adherence pattern was detected between wild type and mutant strains (P > 0.05; paired t tests). In M9 minimal medium, bacterial adherence to the wall seems more brittle and the biofilm structure was seen floating in the well during washing steps. Although both strains showed a similar adherence pattern, paired t tests revealed that the adherence ability of wild type strain was significantly higher than mutant strain (P < 0.05; paired t tests). This could be possibly due to nutrient starvation signals during bacterial growth in defined medium that altered the adherence behavior of bacteria (Gerstel and Romling, 2001). Under nutrient starvation conditions, bcsA gene seems to play a role in the adherence of bacterial cells to surfaces. Therefore, deletion of bcsA gene in B. pseudomallei UKMS-01 leads to reduced ability of bacterial adherence to surfaces specifically under starvation stress.

Next, the phenotypes of the evolved biofilms in different medium were visually inspected. Generally, there are four common types of air-liquid (A-L) interface biofilms including waxy aggregates (WA), viscous mass (VM), floccular material (FM) and physically cohesive (PC) biofilms (Spiers et al., 2006). Wild type and mutant strains grown in M9 minimal medium supplemented with 0.2% glucose produced WA-type biofilms characterized by brittle and waxy looking which covered the entire A-L interface in a thin piece (Figure 4.16). VM-type biofilms were not observed in this experiment because this type of biofilm is dominantly produced by alginate-expressing strains (Ude et al., 2006). FM-type biofilm is characterized by the presence of non-viscous, discrete flocs at the surface and the flocs are capable of maintaining their structure even after mechanical disturbance (Ude et al., 2006). In contrast to FM-type biofilms, PC-type biofilms are consists of a single, cohesive layer spanning the entire A-L interface and often seen attached to the walls in the meniscus region (Ude et al., 2006). The growth of both wild type and ∆bcsA strains in rich medium be it with or without the addition of glucose and salt produced PC-type biofilms with moderate to high viscosity (Figure 4.16). In summary, the knock out bcsA gene in B. pseudomallei does not significantly affect the bacterial biofilm formation capacity.

To further characterize phenotypes associated with bcsA expression, we investigated the capacities of strain UKMS-01 and ∆bcsA to aggregate. It was previously demonstrated that aggregated cells confers enhance resistance towards sodium hypochlorite (chlorine) treatment as compared to non aggregated cells and cellulose has been shown to play an important role in bacterial resistance towards chlorine treatment. These findings suggested that cellulose was produced by the aggregated clumps of bacterial cells. Hence, we are interested as in whether the deletion of bcsA gene in B. pseudomallei UKMS-01 would downregulate the ability of cells to form aggregates. For all four different medium, the amount of planktonic cells was quantified by determining the optical density of cells from the top layer of the cultures in standing cuvettes after sedimentation of bacterial cells. The measurements were taken every 30 minutes for five hours. The growth rate of the mutant strain was not differed significantly from the value for UKMS-01 proposing that any changes in the bacterial cell aggregation behavior are not due to cellular stress. The growth curves of both strains are depicted in Figure 4.17. As shown in Figure 4.18a and 4.18b, the cells of the two strains displayed similar aggregated behavior suggesting that deletion of bcsA gene does not significantly influence the aggregative phenotype.

4.2.5 Roles of cellulose in resistance to desiccation

In some members of Enterobacteriaceae, cellulose has been shown to be the contributing factor for long term resistance to desiccation (White et al., 2006). In Pseudomonas spp., the production of cellulose together with alginate is speculated to protect bacterial cells against ethanol treatment which has a deleterious dehydrating and membrane-perturbing effect on bacterial cells (Clark and Beard, 1979; Gerstel and Romling, 2001). With the aim to better understand to which extent cellulose enhances bacterial resistance to desiccation, we compared the percentage of survival in UKMS-01 with that of ∆bcsA mutant strain after exposure to dry conditions.

Suspension of bacteria was allowed to air dry at room temperature at one hour intervals for five hours before plating onto agar medium. The percentage of cells surviving drying was calculated as:

Recovered cells (c.f.u./ml)

Number of cells (c.f.u./ml) before exposure to dry conditions

As shown in Figure 4.19a and 4.19b, both strains exhibited highest survival rate in LB medium supplemented with 0.2% glucose even after two hours of exposure to dry conditions. On the other hand, both strains when grown on M9 minimal medium seem to struggle hard for survival after desiccation. Still, the survival percentage of both strains plummeted after being air-dried for the second hour. There was visibly no survivor detected after three hours of exposure to dryness. In LB medium without NaCl, wild type strain was better adapted to desiccation after one hour of exposure to dry conditions as compared to ∆bcsA mutant strain contributing to a minor variance of 3%. As a whole, the survival rates of both strains fall in a considerably coherent trend indicating that the lost of bcsA gene does not resulted in any significant changes in sensitivity to desiccation. Thus, our results would suggest that in B. pseudomallei UKMS-01, cellulose function is not correlated with desiccation.

Chapter 5.0 Discussion









catalytic subunit


hypothetical protein similar to bcsG




No significant matches


hypothetical protein similar to bcsC


bcsZ endoglucanase



regulatory subunit


15873The complete genome sequence of B. pseudomallei K96243 was made available in 2004 and in turn allowed for extensive bioinformatics analysis. An in silico data mining of B. pseudomallei K96243 genome database revealed that two copies of cellulose synthase gene were present on chromosome II, one at loci BPSS0735 and another being located at loci BPSS1577. Our sequence analysis showed that both genes are grouped under CelA-like superfamily which is involved in the elongation of glucan chain of cellulose (cd06421, National Center for Biotechnology Information, NCBI). According to NCBI databases, BPSS1577 locus is annotated as bcsA (abbreviations of bacterial cellulose synthase) which encoded for the catalytic subunit of cellulose synthase. The databases also deduced that this gene has a 48.61% identity to the bcsA gene of E. coli O157:H7 (accession no.: YP_111583.1, NCBI). Others putative cellulose biosynthesis genes were also found adjacent to bcsA, suggesting that B. pseudomallei K96243 might harbor a putative cellulose biosynthesis gene cluster on its chromosome II (Figure 5.1).


hypothetical protein similar to


Figure 5.1: Genetic organization of putative bacterial cellulose biosynthesis gene cluster in B. pseudomallei K96243. (Region: 213810..2153980, minus strand). Percentages of amino acid identities are shown for E. coli and S. spp., except for BPSS1580 (C. violaceum)

In the case of BPSS0735, this gene encodes for a 655 residue that shows a 31.8% similarity with an internal region of A. xylinus cellulose synthase 1, from residue 119 to 352. However, the genes adjacent to BPSS0735 have no deducible linkage to any of the cellulose biosynthesis genes in the databases. The genetic organization of the region spanning BPSS0735 is illustrated in Figure 5.2.






outer membrance efflux protein


hypothetical protein similar to



similar to A. tumefaciens hypothetical protein


hypothetical protein

Figure 5.2: Genetic organization of region spanning BPSS0735 on chromosome II of B. pseudomallei K96243. (Region: 985782..992559, minus strand). BPSS0734, percentage of amino acid identity is shown for P. aeruginosa, BPSS0735, percentage of amino acid identity is shown for A. xylinus, BPSS0736, percentage of amino acid identity is shown for A. tumefaciens.

A ClustalW2 multiple sequence alignment tool provided by the European Bioinformatics Institute (EBI) was used to aligned the deduced amino acid sequences of BPSS0735 and bcsA. The alignment output is depicted in Figure 5.3. Similarity based on amino acids among the two cellulose synthase genes was 35%, suggesting a great diversity between BPSS0735 and bcsA (BioEdit software). In other words, B. pseudomallei K96243 harbors copies of chromosomally encoded cellulose synthase genes which appear to be of different genetic origins. Notably, our previous data also revealed that homologous copies of cellulose synthase gene are commonly found in a number of B. pseudomallei strains, signifying the importance of these genes.

Our finding is in accordance with the previously reported redundancy of paralogous proteins in the model organism of cellulose studies whereby G. xylinus was also found to carry a second copy of cellulose synthase gene, namely acsAII (abbreviations of Acetobacter xylinus cellulose synthase) (Saxena and Brown, 1995) along with acsAB (Saxena et al., 1994; Saxena and Brown, 1995). Furthermore, acsAII gene is also conserved among the G. xylinus strains. Sequence comparisons shown that the two cellulose synthase genes are similar but nonidentitical (68.7% similarity, 47.3% identity). The acsAII gene encodes for a 175-kDa polypeptide that exhibit significant cellulose synthase in vitro (Saxena and Brown, 1995). However, it is worth to mentioned that the expression of this gene is not necessary for cellulose biosynthesis when cells are grown under laboratory conditions (Saxena and Brown, 1995).

In this study, we investigated the role of bcsA gene in the cellulose production machinery of B. pseudomallei UKMS-01. It has been previously shown in many studies that bcsA which encodes for the catalytic subunit of cellulose synthase is the core structural gene involved in cellulose biogenesis (Matthysee et al., 1995b; Zogaj et al., 2001; White et al., 2003; Monteiro et al., 2009; Wouter et al., 2009). Over the years, extensive researches have been evolving around the manipulation of bcsA gene to generate cellulose deficient mutant for further characterization and exploration. Additionally, it is proven that disruption of bcsA gene alone is sufficient to switch off the biosynthesis machinery (Zogaj et al., 2001; Romling, 2002; Solano et al., 2002). Furthermore, bcsA is positioned within a putative cellulose biosynthesis gene cluster which has been extensively characterized over the years (Figure 5.1) while the flanking genes of BPSS0735 gave no clue of involvement in cellulose production. On top of all these, three highly conserved motives HAKAGN, QTP, FFCGS, which differentiate cellulose synthase from glycosyltransferases, are present in BcsA protein. BPSS0735 encoded protein also contains the aforementioned motives but with a slightly altered FFCGS motif. Though, the resulting effects due to the presence of FCIGT motif instead of FFCGS motif in BPSS0735 encoded protein have yet to be known. The motif, QRXRW, characteristic of processive glycosyltransferases, is present in both cellulose synthase. A gathering of these accumulative predictions with some that have been substantiated by previous studies, we believed that bcsA plays a more dominant role in cellulose biosynthesis as compared to its counterpart.

In order to examine the roles of bcsA and cellulose production in B. pseudomallei, we first generated a ∆bcsA mutant of B. pseudomallei namely, B. pseudomallei UKMS-01 ∆bcsA01 by using allelic replacement mutagenesis via homologous recombination. Allelic replacement was carried out by using pDM4 and its derivatives as the gene replacement vector. The plasmid pDM4 is a non-replicative vector that harbors mob gene which allows for the mobilization of this vector into the desired bacterial host. The selections of single and double cross over events were both relied on the selectable (catR) and counter-selectable (sacB) markers (Milton et al., 1996). Sequencing of the PCR amplified fragment derived from ∆bcsA01 mutant shown that legitimate recombination has occurred at the targeted region, namely BPSS1576 and BPSS1578 and it also displayed the deletion of bcsA gene from the chromosomal DNA as compared to that of wild type. Therefore, this genetic manipulation tool allows for targeted construction of markerless mutagenesis, including point mutations or in-frame deletions. It has several advantages over conventional gene disruption whereby only the gene of interest is disrupted or deleted with minimal perturbation of adjacent genes. Moreover, the excision of selectable and counter-selectable markers during the introduction of mutation permits their repeated use in subsequent manipulations (Moore and Leigh, 2004). Thus, it is feasible to use the vector pDM4 for generation of double knock out mutant of both bcsA and BPSS0735 genes that facilitates future combinative in-depth study (e.g. gene-gene interactions) of cellulose producing phenotype in B. pseudomallei UKMS-01.

Consequently, the resulting ∆bcsA mutant was further characterized to study the resulting effects of the deletion on B. pseudomallei UKMS-01. Notably, ∆bcsA mutant has displayed similar binding patterns of Congo red and Calcofluor dyes as observed in wild type. In addition to that, cellulase digestion assay also indicated the presence of cellulose in both ∆bcsA mutant and wild type strains albeit in low amount. These observations indicated that BcsA in not the only cellulose synthase responsible for cellulose biosynthesis in B. pseudomallei UKMS-01. Thus, the characterization studies which showed no significant deviations from wild type cells may lie in the fact that the occurrence of second copy of cellulose synthase gene in B. pseudomallei has probably neutralized the deletion effect of bcsA. Likewise, BPSS0735 encoded protein has either played a compensatory or dominant role in cellulose biosynthesis as opposed to our earlier postulation. However, our preliminary findings offer the possibility to further explore the molecular details of cellulose production in B. pseudomallei that entail the generation of single knock out mutant of BPSS0735 and double knock out mutant of bcsA and BPSS0735. It is envisaged that further characterization of these mutants will expedite the understanding of the theoretical framework of cellulose biosynthesis machinery in B. pseudomallei UKMS-01.

In a separate observation, B. pseudomallei UKMS-01 seems to express low level of cellulose whereby only 2.45 μg of release glucose unit was produced from 200 mg of lyophilized plate-grown wild type cells. This could probably reflect an underestimated level of cellulose expression taking into considerations that under laboratory conditions without any induction, the cells might want to maintain the expression of this high cost polymer at the basal level (Ude et al., 2006). This opinion is in agreement with a finding in P. fluorescens SBW25, whereby elevated expression of acetylated cellulose polymer was recorded in sugar beet rhizosphere (Gal et al., 2003). Indeed, the rhizosphere-induced wss (operon encoding acetylated cellulose polymer) locus has significantly contributed to the ecological performance of P. fluorescens SBW25 in colonizing the plant environment (Gal et al., 2003). Again, this supports our conjecture that bcs/acs-like locus is unlikely to be fully activated in the laboratory environment indicating that cellulose might be the determinants of adaptive fitness in bacterial natural habitats. To gain a better understanding and tangible evaluation on different aspect of cellulose production in B. pseudomallei UKMS-01, it is strongly suggested to adapt environmental stimuli in the mode of future studies.

The ability of B. pseudomallei to infect susceptible tomato plants but not rice plants was recently demonstrated (Lee et al., 2010). It pointed towards the notion that B. pseudomallei is also a potential plant pathogen apart of its established infection in human and animals. Additionally, E. coli O157:H7 pathogenic strain has also been widely associated with infections of alfalfa and radish sprouts, lettuce and spinach leaves (Brueuer et al., 2001; Lodato, 2002). Interestingly, cellulose synthase mutants of E. coli O157:H7 showed a significantly reduced binding ability towards their plant host cells (Matthysee et al., 2008). In this study, we have reported that under nutrient depletion conditions, ∆bcsA mutant showed a significant reduction in adherence ability to abiotic surfaces. Taken together all these and having established the ability of B. pseudomallei as a phytopathogen, we thus postulated that bcsA may act as contributing factor in the initial plant attachment and latter plant colonization of B. pseudomallei. For future studies, a threading model of single (bcsA or BPSS0735) or double knock out (bcsA and BPSS0735) mutant of cellulose synthase gene can be developed to study the correlation between cellulose and plant attachment that successively help us to gain further insight into the survival and persistence of B. pseudomallei in the natural environment.

Chapter 6.0 Conclusion

In summary, ∆bcsA mutant of B. pseudomallei UKMS-01 was successfully generated in this study via allelic replacement mutagenesis by using pDM4 and its derivatives as gene replacement vectors. The ∆bcsA mutant was constructed with the aim of investigating the role of bcsA in cellulose production and the resulting effects of bcsA deletion on biofilm formation and desiccation survivability in B. pseudomallei UKMS-01. Moreover, verification of the functionality of BPSS0735 in cellulose biosynthesis is possible following the loss of bcsA gene.

Evaluation of both wild type and ∆bcsA mutant colony morphology on Congo red and Calcofluor plates showed no significant difference. Both strains displayed a pink, dry and rough morphotype on Congo red plates. Fluorescens were also visible at the edge of the colony of both wild type and mutant strain. Though no cellulose could be directly visualized by electron microscopy, but extracellular matrix of presumably different structure and texture were observed from wild type and ∆bcsA mutant. The mutant cells were connected by a less dense and more fibrous extracellular matrix while the wild type cells showed a densely packed gum-like matrix in direct contact to the bacterial surface and between bacterial cells. Besides, cellulose was detected from the lyophilized cells of UKMS-01 and ∆bcsA01 via cellulase production assay. The estimated concentrations of released glucose in wild type sample treated with cellulase and ∆bcsA sample treated with cellulase are 2.45 μg and 3.1 μg respectively. The findings above strongly suggest that BcsA is not the only cellulose synthase responsible for cellulose production, thus indicated possible cellulose synthase activity by BPSS0735.

On another note, deletion of bcsA did not reflect any correlation to biofilm formation, adherence and long-term desiccation. One exception was observed during the adherence assay of ∆bcsA mutant grown on nutrient depleted medium whereby the adherence ability of mutant cells was found to be significantly reduced comparing to wild type cells.

In conclusion, bcsA is apparently not the only cellulose synthase gene involve in cellulose production of UKMS-01 as shown in this study. Whether BPSS0735 encoded protein plays a dominant or compensatory role in cellulose production and whether bcsA and BPSS0735 act synergistically or antagonistically in cellulose biogenesis are certainly something worth pondering upon. Most importantly,