B. amyloliquefaciens FZB42 is known as a plant growth promoting rhizobacterium because it offers not only protection towards the competitive plant-pathogenic microflora within rhizosphere by secretion of antifungal and antibacterial lipopeptides and polyketides (Koumoutsi et al. 2004; Chen et al. 2006) but also by production of plant hormones such as IAA (Idris et al. 2007). However, the molecular mechanisms behind this ability are not fully understood. In order to find out the beneficial action of this strain at molecular level, transposon mutagenesis was performed.
Transposon-based mutagenesis is a powerful technique for generating mutant libraries, and its use has led to the identification of gene functions in various bacterial systems. In bacteria, transposons are widely employed as random insertion mutagens both at a genome level or and in the analysis of the organization of individual genes (Hayes, 2003; Picardeau 2010). Mariner-family transposable elements are a diverse and taxonomically widespread group of transposons occurring throughout the animal kingdom. Among hundreds of different mariners, only two are known to be active, these are Mos1 and Himar1. Both require no host-specific factors for transposition and so have been advanced as generalized genetic tools (Lampe et al. 1999). Himar1 has been used as a prokaryotic genetic tool such as in Burcella melitensis (Wu et al. 2006), Leptospora interrogans (Bourhy et al. 2005), Leptospira biflexa (Louvel et al. 2005), Rickettsia prowazekii (Liu et al. 2007), Borrelia burgdorferi (Morozova et al. 2005), and Bacillus subtilis (Lebron et al. 2006).
In this research three different plasmids containing a mariner-based Himar1 tranposon namely pMarA, pMarB and control plasmid pMarC were used in transposon mutagenesis in B. amyloliquefaciens FZB42. Plasmid pMarA and pMarB differ in the promoters that drive the expression of the Himar1 transposase gene. pMarA has Himar1 under the transcriptional control of housekeeping σ factor σA of B. subtilis, while pMar B uses general stress response σ factor σB for transposase expression. pMarC has no transposase gene as well as its promoter and is used as a control (Le Breton et al.2006).
Transformation of these plasmids was done by modification of the method of Kunst, F and Rapoport, G. (1995). The same amount (1 μg) of plasmid DNA from pMarA, pMarB and pMarC was transformed into FZB42 (see material and methods). pMarA and pMarC have been successfully transformed into FZB42, however, pMarB failed. Because pMarA and pMarB contained the same Himar1 mariner transposase gene only differing in their respective promoters, we continued to use the pMarA as a source of transposon mutagenesis.
Transformants that contained plasmid pMarA had to be verified that they contained the original intact plasmid before being used for transposon mutagenesis. This was done by screening the transformants for the plasmid-associated properties, i.e. Kanr and Eryr at the permissive temperature for plasmid replication (30°C) and Kanr and Erys at the restrictive temperature (48°C). Then the plasmid was extracted from the transformants and transformed into E. coli DH5α. Next, plasmid DNA was extracted from E. coli DH5α and subjected to restriction endonuclease analysis with EcoRI. The restriction was then analysed through agarose gel electrophoresis to verify that the transformants contained the correct plasmid. Fig. 2 shows the restriction analysis of plasmid extracted from transformants E. coli DH5α and plasmid pMarA as a positive control.
|Figure 2. Restriction analysis of plasmid DNA cut with EcoRI|
|Lane 1- 4 from transformed E. coli, lane C from plasmid pMarA.|
After verifying that the plasmids pMarA and pMArC were correctly inserted in B. amyloliquefaciens FZB42, the transposon mutagenesis was done by growing the isolated clones overnight in liquid LB medium at 37°C. Then each culture was plated either on LB, LB plus 5 μg/ml Kan or LB plus 1 μg/ml Ery and incubated at the nonpermissive temperature for plasmid replication (48°C). Representative data that are the average data of two separate experiments are presented in Table 6. Kanr clones represented in this transposition events appeared at frequency ~ 10-2 which is significantly higher than that reported for transposons Tn917 and Tn10 (10-6 and 10-4, respectively), which are commonly used in B. subtilis. There are no antibiotic-resistant clones detected when B. amyloliquefaciens FZB42 carrying pMarC lacking of transposase coding sequence was plated in LB plus Kan and LB plus Ery at 48°C. Hence, the Emrr clones detected were likely a consequence of transposition event from plasmid multimers in which most plasmid sequence were inserted into the B. amyloliquefaciens FZB42 chromosome (Le Breton et al. 2006).
Viable cell count (CFU/ml)
LB KanR 48°C
LB ErmR 48°C
3.4 x 10 8
2.6 x 10 7
3.6 x 10 5
7.6 x 10 -2
2.5 x 10 8
Southern blot analysis was done to verify integration of the transposon and to test whether the insertions are likely to be random. In this analysis chromosomal DNA from B. amyloliquefaciens FZB42 and clones were isolated and digested with EcoRI. Digoxigenin-labeled DNA specific for the transposon was created by cutting TnYLB-1 region with PstI. Hybridization of this probe to EcoRI-digested DNA from clones gave the patterns illustrated in Fig. 3 A. In addition, PCR for the presence of the kanamycin resistance gene was also performed using primers flanking the kanamycin coding sequence in all clones including their complementation and retransformation (Fig. 3 C).
|Figure 3. TnYLB-1 transposition in B. amyloliqufaciens FZB42|
|A. Southern hybridization analysis of randomly chosen B. amyloliquefaciens FZB42 TnYLB-1 insertion mutants. Chromosomal DNA from B. amyloliquefaciens FZB42 (WT) and transposants (lanes 1-10) were digested with EcoRI and analyzed by Southern blotting using a hybridization probe specific for TnYLB-1. DNA fragment sizes (kbp) areindicated to the left and are based on DNA markers. B. PCR products of kanamycin gene, wild type FZB42 (lane 1) and the mutants (lane 2-20).|
To verify that transposition with TnYLB-1 is an efficient tool to get insertion mutations, 787 temperature-resistant Kanr clones were spotted onto glucose minimum medium to screen for auxotrophic mutations. Of these, seven (~1%) clones spotted failed to grow on the minimal medium, indicating that the transposon insertion was an effective way to create mutation. To identify the B. amyloliquefaciens FZB42 genes disrupted by insertion and to further characterize the insertion sites, chromosomal DNA was extracted from the auxotroph phenotype. Two DNA samples isolated from auxotroph phenotype were used in an inverse PCR protocol using primers oIPCR1 and oIPCR2 that allows amplifying the flanking region of the transposon containing the inverse terminal repeat. The amplified DNAs were then sequenced using primer oIPCR3 and the sequences were characterised by BLAST analysis (Le Breton et al. 2006). Each of the two auxotroph mutants that were examined yielded in an insertion at an unique location on the B. amyloliquefaciens FZB42 chromosome. The two putative insertions were found in pabB encoding para-aminobenzoate synthase (subunit A) and hisJ encoding histidinol phosphate phosphatase. Additions of histidin and para aminobenzoic acid in the minimal medium restore the growth of hisJ and pabB auxotroph, confirming the need of those compounds for growth of both auxotrophic mutants.
B. amyloliquefaciens FZB42 which contains plasmid transposon TnYLB-1 was then used to create mutant library of the different phenotype, i.e. mutant in biofilm production, swarming, plant growth promotion, nematocidal production and antibiotic. Three biofilm mutants (pabB, yusV and degU), one swarming mutant (degU), three plant growth promotion mutants (nfrA, abrB and RBAM_017410), five nematocidal mutants (prkA, yhdY, ywmL, mlnD and RBAM_007470) and five antibiotic mutants (degU, degS, nrsB, yaaT and RBAM_029230) were found in this screening (Romy Scholz and Quisheng He, unpublished data). These insertion sites are mapped onto a circular representation of the B. amyloliquefaciens FZB42 genome in Fig. 4, revealing a genome-wide distribution of insertion sites that are found in open reading frames. Collectively, the data shown in Fig. 3 and 4 indicate that transposition of the TnYLB-1 element in B. amyloliquefaciens FZB42 produces predominantly mutants with single insertions that are distributed around the chromosome in an apparently random fashion.
|Figure 4. Random distribution of TnYLB-1 insertions in the B. amyloliquefaciens FZB42 chromosome|
|Brown color represents biofilm mutants, blue color represents PGPR mutants, green color represents nematocidal mutants, red color represents antibiotics mutants and grey color represents auxotrophic mutants.|
Bacteria motility mechanisms, including swimming and swarming, can have a profound impact on the colonization of surfaces, the first step in the formation of adherent microbial assemblies called biofilms. Both features are known to play important role on how prokaryotes interact with and adapt to surface environments (Mireles II et al. 2001; Meritt et al. 2007). To discover the genes involved in swarming motility and biofilm formation, I carried out insertional mutagenesis in B. amyloliquefaciens FZB42 using the transposon TnYLB-1. Approximately 6000 colonies from the transposon library were screened for being impaired in swarming motility on swarming agar plates (LB solidified with 0.7% agar). Swarming motility of B. subtilis has previously been evaluated with this approach (or slight variations thereof) by several groups (Senesi et al. 2002; Kearns et al. 2003; Calvio et al. 2005). One mutant in swarming motility was found in this screening which was impaired in its ability to cover the whole area of swarming agar plate (Fig. 8). Later, on biofilm screening this mutant also showed a defect in biofilm formation.
I screened the TnYLB-1 insertion library using MSgg media to find biofilm mutants. Defect in biofilm formation of B. subtilis has previously been screened with this approach (Brenda et al. 2004). Using this assay, 6000 mutants were screened to identify mutants exhibiting a reduced ability to form biofilms. Three unique mutants that were impaired in biofilm production in MSgg media were isolated. To identify the B. amyloliquefaciens FZB42 genes disrupted by the insertions, chromosomal DNA from these three isolates was extracted and used in an inverse PCR protocol that amplified the chromosomal DNA abutting the transposons’s inverse terminal repeats (ITRs). The amplified DNA was then sequenced using primer oIPCR3. BLAST analysis of the DNA sequences revealed that these mutants carried TnYLB-1 insertions at distinct genes. The insertion of TnYLB-1 transposon in two mutants that showed a reduced ability to form biofilm was in the yusV and pabB genes, whereas the mutant that exhibited defect in swarming ability and biofilm formation was in the degU gene. The yusV gene has unknown function, but it is similar to iron (III) dicitrate transport permease and pabB gene encodes for para-aminobenzoate synthase (subunit A). These three mutants showed different phenotypes in respect to biofilms formation. The mutant of B. amyloliquefaciens FZB42 harboring degU::TnYLB-1 formed thin pellicle when grown in standing liquid MSgg medium (Fig. 9 B), the mutant harboring yusV::TnYLB-1 formed pellicles with flat surface (Fig. 13 B), whereas the mutant harboring pabB::TnYLB-1 formed flat and thin pellicles with granules on the top of the surface (Fig. 17 B).
Nucleotide sequence analysis revealed that the transposon insertion in the mutant which was defect in biofilm formation and swarming motility took place within degU gene. The degU has a key role in regulating several post-exponential phase processes in B. subtilis, including the activation and inhibition of genetic competence, the inhibition of flagellar based motility, the activation of degradative enzyme production and the activation of poly-γ-glutamic acid production (Verhamme et al. 2007). Therefore, the degU mutant is a regulatory mutant, because it involves in regulating many bacterial cell activities. The genetic organizations of the degU regions carrying transposon insertion are shown in Fig. 5.
|Figure 5. Genomic organization of degU region carrying the TnYLB-1 insertion and its flanking regions.|
In order to confirm that insertion of TnYLB-1 mariner-based tranposon in degU gene was responsible for impairing swarming motility in B. amyloliquefaciens FZB42, complementation of the degU mutant by inserting the intact degU was done. To complement the degU gene, the coding region plus 178 bp of upstream sequence and 183 bp of downstream sequence were amplified using primers degU-dw-ClaI and degU-up-Eco88I containing Eco91I and SacII site respectively (Fig. 6). The fragment (889 bp) was cloned into linearized ClaI/Eco88I pUC18 plasmid which had an Amy cassette (pVBF). The ligated DNA was then transformed into CaCl2-competent DH5α cells. To confirm the clones containing the degU gene fragment, PCR using degU-dw-ClaI and degU-up-Eco88I primers was performed. The correct plasmid pVBF-∆degU cassette was transformed into degU mutant. The transformants were then grown in LB plus Kan (5 ug/ml), Ery (1 ug/ml) and 1% amylum. The transformants without clear zone when the medium was added with iodine solution were the the correct transformants.
|Figure 6. Strategy for construction of pUC18-∆degU cassette.|
To validate the correct insertion of the complementation and retransformation transformants, chromosomal DNA was amplified by PCR using primers degU-dw-ClaI and degU-up-Eco88I and compared with the B. amyloliquefaciens FZB42 and degU mutant (Fig. 7). In the complementation transformant, PCR fragment of the degU gene showed a similar fragment length as a wild type suggesting that the complementation had occurred, whereas in in retransformation transformant, the length of the degU gene was longer than the wild type indicating that degU has been inserted with TnYLB-1 segment as in degU mutant.
|Figure 7. PCR product of degU gene|
|Wild type FZB42 (lane1), degU mutant (lane 2), complementation of degU (lane 3) and retransformation of degU (lane 4).|
Insertion of intact degU gene into the degU mutant restored the swarming motility of the mutant (Fig. 8 C). An additional approach to demonstrate that degU plays a role in swarming motility was done by retransformation. In retransformation, chromosomal DNA of degU mutant that has been disrupted by insertion of TnYLB-1 was transformed into the wild type of B. amyloliquefaciens FZB42. The phenotype of transformants from retransformation showed an impaired swarming motility as in degU mutant confirming that the degU gene had a role in swarming motility (Fig. 8).
|Figure 8. Phenotype of swarming motility in degU mutant|
|B. amyloliquefaciens FZB42 (A), degU mutant (B), complementation degU (C) and retransformation of mutant degU (D).|
The degU gene is also responsible for biofilm formation. The degU mutant was impaired in its ability to form robust pellicles as in the wild type when grown in MSgg medium. This mutant formed thin pellicles with little granules in the center (Fig. 9 B). Complementation degU mutant with intact degU from B. amyloliquefaciens FZB42 repaired the mutant ability to produce biofilm in standing MSgg medium (Fig. 9 C). On the contrary, retransformation of degU::TnYLB-1 fragment to the wild type of B. amyloliquefaciens FZB42 resulted in transformants impaired in biofilm production with phenotype relatively similar to the degU mutant (Fig. 9 D). In Bacillus subtilis JH642 degU was shown to enhance biofilm formation through activating poly-γ-glutamic acid production (Stanley and Lazazzera, 2005). However the role of poly-γ-glutamic acid in B. subtilis wild strain 3610 and B-1 in the formation of sessile communities had an inconsistent role (Brenda et al. 2006), thus raising the question whether degU had a variable role in controlling biofilm formation.
|Figure 9. Phenotype of biofilm formation in degU mutant|
|B. amyloliquefaciens FZB42 (A), degU mutant (B), complementation of degU (C), and retransformation of degU (D).|
The insertion of TnYLB-1 tranposon found in the second biofilm mutant was in the yusV gene. Its function is unknown but it is similar with iron (III) dicitrate transport permease. The yusV gene lies in the yus operon which still has unknown function (yusU, yusW, yusX and yusZ). Genetic map of insertion TNYLB-1 in yusV gene and its neighboring region is shown in Fig. 10.
|Figure 10. Genomic organization of yusV region carrying the TnYLB-1 insertion and its flanking regions.|
The intact yusV gene plus 77 bp of upstream sequence and 170 bp of downstream sequence was amplified using primers yusV-up-SacII and yusV-dw-Eco91I containing Eco91I and SacII site respectively. The fragment (1072 bp) was then cloned into pUC18 plasmid which had an Amy cassette (pVBF) and linearized with SacII and Eco91I restriction enzymes (Fig. 11). The ligated plasmid DNA was transformed into CaCl2-competent DH5α cells. To select the clones with yusV insertion, the transformants were grown in the medium containing Ampr and confirmed with PCR using plasmid DNA as a template with yusV-up-SacII and yusV-dw-Eco91I primers. The plasmid with the correct insertion of yusV gene was then transformed into the yusV mutant. The transformants were selected for Kan r and Eryr with no clear zone in media when it was added with iodine solution. Retransformation of the yusV::TnYLB-1 fragment into the wild type was also done in order to confirm that the yusV gene was responsible for the defect in biofilm formation in yusV mutant.
|Figure 11. Strategy for construction of pUC18-∆yusV cassette.|
To validate that the transformants contained the right insertion, the chromosomal DNA of transformants was amplified using the appropriate primers and compared to the wild type and the mutant. PCR that amplified yusV gene in yusV mutant showed a longer PCR fragment to the wild type indicating the insertion of transposon TnYLB-1. Two bands of PCR product of yusV gene in the complementation confirming that the intact yusV gene and the yusV::TnYLB-1 fragment were in the chromosome of yusV mutant. Retransformation of yusV::TnYLB-1 fragment to B. amyloliquefaciens FZB42 resulted in a longer PCR fragment in contrast to the wild type (Fig. 12).
|Figure 12. PCR product of yusV gene|
|wild type FZB42 (lane1), yusV mutant (lane 2), complementation of yusV (lane 3) and retransformation of yusV (lane 4).|
The finding that the yusV gene is involved in biofilm formation has never been reported in any publication (Branda et al. 2006; Kobayashi 2007; Chai et al. 2008), hence adds a new insight in biofilm formation. Insertion of TnYLB-1 in yusV gene impaired the ability of the mutant to produce biofilm. The yusV mutant produced very thin pellicles when grown in MSgg medium (Fig. 13 B), whereas the wild type formed thick pellicles (Fig. 13 A). When complemented with intact yusV gene, the insertion restored the ability of the yusV mutant to form biofilm (Fig. 13 C). Retransformation of yusV::TnYLB-1 fragment to the wild type generated the inability of the wild type phenotype to form a thick pellicles (Fig. 13 D).
|Figure 13. Phenotype of biofilm formation in yusV mutant|
|B. amyloliquefaciens FZB42 (A), yusV mutant (B), complementation of yusV (C), and retransformation of yusV (D).|
The pabB gene is one of the folate biosynthesis operon genes, that encodes for para-aminobenzoate synthase. Insertion of TnYLB-1 tranposon in this region generated auxotroph and biofilm mutant. The genomic organisation of the chromosomal region carrying the TnYLB-1 insertion in the pabB gene and its flanking region is depicted in Fig. 14.
|Figure 14. Genomic organization of pabB region carrying the TnYLB-1 insertion and its flanking regions.|
To complement the pabB gene that has been disrupted by the insertion of TnYLB-1 transposon, the plasmid harboring the intact pabB gene from B. amyloliquefaciens FZB42 was constructed. The construct contained the gene pabB/pabB-dw-Eco91I/pabB-up-SacII fragment (1695 bp), which was inserted into linearized Eco91I/SacII pUC18 containing Amy cassette (pVBF) (Fig. 15). The plasmid bearing the wild type allele was transformed into CaCl2-competent E. coli DH5α and selected for Ampr, which was associated with plasmid transformation. To verify the ligation of the wild type gene in the pVBF plasmid, the wild type gene was amplified by PCR using plasmid DNA extracted from E. coli DH5α as the template with pabB-dw-Eco91I and pabB-up-SacII primers. The plasmid containing the correct insertion of the wild type gene was transformed into pabB mutant and selected for Eryr.
|Figure 15. Strategy for construction of pUC18-∆pabB cassette.|
The length of PCR fragment which amplified the pabB gene in the wild type was different from the pabB mutant, complementation and retransformation. The pabB mutant, its complementation and retransformation showed the same band of PCR product confirming that the insertion of TnYLB-1 in the pabB gene had occured. However, the complementation showed also an extra PCR fragment which has the same size as the wild type indicating that the intact pabB gene has been successfully inserted to the mutant (Fig. 16).
|Figure 16. PCR product of pabB gene|
|wild type FZB42 (lane1), pabB mutant (lane 2), complementation of pabB (lane 3) and retransformation of pabB (lane 4).|
Compared to the wild type, biofilms in pabB mutant showed distinct features such as flat and thin pellicles with little granules on the top of the covered surface of MSgg media (Fig. 17 B). Complementation of the pabB gene by insertion of the intact pabB gene into the pabB mutant restored its ability to produce biofilm. The feature of biofilm after complementation showed thick pellicles as the wild type (Fig. 17 C). However, impairment in biofilm formation in this mutant was not directly related to the gene that was involved in biofilm formation but due to the growth defect in the bacteria as pabB mutant was an auxotroph. This was confirmed by the addition of 0.1 mM para-aminobenzoic acid (PABA) in MSgg medium which restored the ability of the pabB mutant to form biofilm (Fig. 17 E). Retransformation of pabB mutant to B. amyloliquefaciens FZB42 impaired its ability to produce biofilm (Fig. 17 D), indicating that the pabB fragment has transformed into the wild type.
|Figure 17. Phenotype of biofilm formation in pabB mutant|
|B. amyloliquefaciens FZB42 (A), pabB mutant (B), complementation of pabB (C), retransformation of pabB (D) and addition of 0.1mM PABA (E).|
To investigate the genes that are involved in plant growth-promoting in B. amyloliquefaciens FZB42, non-growth-promoting mutants generated by mariner based transposon TnYLB-1 were screened by the Lemna biotest system (Idris et al. 2007). In short, 1.25 μl of fresh growing mutant cells at OD600 1 were added to the plant in microtiter plate wells containing 1.25 ml Steinberg growth medium. Each treatment was repeated six times. The plates were kept at 22°C with continuous 24 hours light. After ten days plants were harvested and growth was measured in terms of dry weight. Of 3000 transformants screened, three transformants exhibited impairment in plant growth promoting activity. To identify the genes in these mutants, inverse PCR products were sequenced and searched using BLAST analysis. Blast results showed that the insertion of the mutants were in the nfrA, abrB and RBAM_017410 gene.
Since the nfrA mutant showed reduction of plant growth-promoting activity of wild type FZB42, I examined whether the complementation of the gene could restore the promoting ability of the mutants. The nfrA gene encodes for FMN-containing NADPH-linked nitro/flavin reductase, which is regarded as an essential gene. The region of the nfrA gene inserted in TnYLB-1 transposon and its neighboring area is depicted in Fig. 18.
|Figure 18. Genomic organization of nfrA region carrying the TnYLB-1 insertion and its flanking regions.|
Complementation of the nrfA mutant was done by amplifying the nfrA coding region plus 241 bp upstream sequences and 278 bp of downstream sequences using nfrA-dw-Eco88l and nfrA-up-ClaI primers, which contained Eco88l and Clal site, respectively. The fragment of nfrA/nfrA-dw-Eco88l/nfrA-up-ClaI (1269 bp) was cloned into linearized ClaI/Eco88I pUC18 plasmid which contained an Amy cassette (pVBF) (Fig. 19).
|Figure 19. Strategy for construction of pUC18-∆nfrA cassette.|
The pVBF containing fragment of nfrA/nfrA-dw-Eco88l/nfrA-up-ClaI was then transformed into CaCl2-competent DH5α cells. The clones were selected for Ampr and confirmed with PCR using the appropriate primers. The plasmid with the appropriate fragment was then transformed to the nfrA mutant. Retransformation of the nfrA::TnYLB-1 fragment into wild type was done to confirm that the transposon was inserted in this gene. The correct transformants were confirmed with PCR and analysed by gel electrophoresis. Fig. 20 shows the insertion of intact nfrA gene into the nfrA mutant and insertion of the nfrA::TnYLB-1 fragment into the wild type confirming that the complementation and retransformation were successful.
|Figure 20. PCR product of nfrA gene|
|Wild type FZB42 (lane1), nfrA mutant (lane 2), complementation ofnfrA (lane 3) and retransformation of nfrA (lane 4).|
In the Lemna biotest system, the nfrA mutant showed significant reduction (p ≤ 0.05) in plant dry weight compared to the wild type. Complementation of nfrA gene in mutant restored the plant growth-promoting ability as showed by a significant increase (p ≤ 0.05) compared to mutant. Retransformation of mutant which caused insertion of nfrA::TnYLB-1 into wild type impaired the growth stimulation as indicated by significant reduction (p ≤ 0.05) compared to wild type (Fig. 21 and 22).
|Figure 21. Influence of nfrA mutation on plant growth promoting ability of B. amyloliquefaciens FZB42 on L. minor|
|The figure shows the representative plant response to the bacterial treatment at 10 days. A. B. amyloliquefaciens FZB42, B. Control, C. nfrA mutant, D. Complementation, E. Retransformation|
|Figure 22. Growth stimulating effects of nfrA mutation on L. minor|
|FZB42 : B. amyloliquefaciens FZB42, AB106 : nfrA mutant, AB106C : Complementation, AB106R : Retransformation. Values shown represent the mean of dry weight ± standard error. Different letters indicates means that differ significantly (P ≤ 0.05).|
In order to examine whether or not the plant growth promotion mutants impaired their ability to promote the growth of other plant than L. minor, I tested toward the growth of Arabidopsis thaliana. A gnotobiotic system was applied to study the interactions between B. amyloliquefaciens FZB42 and A. thaliana to exclude other uncontrolled experimental factors that would interfere with this interaction as in natural environment. In this system, the surface sterilized A. thaliana seeds were germinated for seven days and the roots were immersed in bacterial culture (1 x 10-5 CFU/ml) for five minutes. The plants were then grown in a growth chamber for three weeks. In this assay, the mutant resulted in significant (p ≤ 0.05) reduction of plant fresh weight compared to wild type. No significant differences were found among control, mutant and retransformation. The wild type and complementation exhibited significant increase (p ≤ 0.05) compared to control and mutant. These data suggest that complementation restored plant growth promoting activity of the wild type (Fig. 23 and 24).
|Figure 23. Influence of nfrA mutation on plant growth promoting ability of B. amyloliquefaciens FZB42 on A. thaliana|
|The figure shows the representative of plant growth to the bacterial treatment at 21 days. A. B. amyloliquefaciens FZB42, B. Control, C. nfrA mutant, D. Complementation of nfrA, E. Retransformation of nfrA.|
|Figure 24. Growth stimulating effects of nfrA mutation on A. thaliana|
|FZB42 : B. amyloliquefaciens FZB42, AB106 : nfrA mutant, AB106C : Complementation, AB106R : Retransformation. Values shown represent the mean of fresh weight ± standard error. Different letters indicate means that differ significantly (P ≤ 0.05).|
The abrB gene plays a role in regulation of transition state genes which are involved in expression of numerous genes functions such as the formation of biofilm (Hamon et al. 2004 and Branda et al. 2001), production of extracellular degradative enzymes (Makarewicz et al. 2008), initiation of sporulation (Perego and Hoch, 1991) and production of antibiotics (Strauch et al. 2008). A mutant with insertion of TnYLB-1 transposon in the abrB gene showed reduction in plant growth promoting activity of B. amyloliqufaciens FZB42 as detected in screening with L. minor. Insertion of TnYLB-1 in the abrB coding region and its flanking region is shown in Fig. 25.
|Figure 25. Genomic organization of abrB region carrying the TnYLB-1 insertion and its flanking regions.|
To complement the abrB mutant, the abrB gene plus a 373 bp upstream region and a 396 bp downstream region was amplified using abrB-dw-SacII and abrB-up-Eco91l primers which contained SacII and Eco91l restriction site, respectively. The fragment (1045 bp) was then cloned into linearized SacII/Eco91I pUC18 plasmid which had an Amy cassette (pVBF) (Fig. 26). The ligated DNA was transformed into CaCl2-competent DH5α cells and selected for Ampr. The clones containing the abrB insertion were confirmed with PCR using abrB-dw-SacII and abrB-up-Eco91l primers. The plasmid was then transformed into abrB mutant and selected for Kanr and Eryr.
|Figure 26. Strategy for construction of pUC18-∆abrB cassette.|
Amplification of the abrB gene in the wild type resulted in a fragment which had a length around 900 bp, whereas in the mutant it was around 2000 bp. Complementation of the abrB mutant that has been disrupted by tranposon insertion was done by inserting an intact abrB gene from the wild type. The PCR product of abrB gene amplification from complementation showed two fragments which indicated that the abrB intact gene has been successfully inserted back to the mutant (Fig. 27). Retransformation of abrB::TnYLB-1 fragment from the abrB mutant into the wild type FZB42 was also done in order to confirm that the mutation was in the abrB gene. The PCR product of retransformation indicated that abrB gene containing transposon insertion has replaced the intact abrB gene in the wild type (Fig. 27).
|Figure 27. PCR product of abrB gene|
|Wild type FZB42 (lane1), abrB mutant (lane 2), complementation of abrB (lane 3) and retransformation of abrB (lane 4).|
Bacterial culture of FZB42 wild type, abrB mutant and its complementation and retransformation were applied to Lemna biotest system in order to compare their effects on the growth of L. minor (Fig. 28 and Fig. 29). L. minor treated with the wild type bacterial culture revealed significant increase (p ≤ 0.05) in growth parameter indicated by dry weight compared to control, mutant and retransformation. As predicted, complementation of the abrB mutant showed similar plant growth promoting activity to wild type FZB42, however it showed no significance difference compared to retransformation. The growth of plants treated with complementation cells increased significantly (p ≤ 0.05) compared to control and mutant.
|Figure 28. Influence of abrB mutation on plant growth promoting ability of B. amyloliquefaciens FZB42 on L. minor|
|The figure shows the representative plant response to the bacterial treatment at 10 days. A. B. amyloliquefaciens FZB42, B. Control, C. abrB mutant, D. Complementation of abrB, E. Retransformation of abrB.|
|Figure 29. Growth stimulating effects of abrB mutation on L. minor|
|FZB42 : B. amyloliquefaciens FZB42, AB108 : abrB mutant, AB108C : Complementation, AB108R : Retransformation. Values shown represent the mean of fresh weight ± standard error. Different letters indicate means that differ significantly (P ≤ 0.05).|
The effect of the bacterial culture of the abrB mutant towards the growth of A. thaliana was shown in Fig. 30 and Fig 31. Immersion of the roots of A. thaliana into the bacterial culture of FZB42 improved the growth of the plant significantly (p ≤ 0.05) when compared to the control and the retransformation. Application of the mutant slightly reduced the fresh weight of A. thaliana compared to FZB42. The treatment with complementation cells showed no significant differences compared to the control, mutant and retransformation altough the fresh weight of A. thaliana in this treatment showed an increase tendency.
|Figure 30. Influence of abrB mutation on plant growth promoting ability of B. amyloliquefaciens FZB42 on A. thaliana|
|The figure shows the representative of plant growth to the bacterial treatment at 21 days. A. B. amyloliquefaciens FZB42, B. Control, C. abrB mutant, D. Complementation of abrB, E. Retransformation of abrB.|
|Figure 31. Growth stimulating effects of the abrB mutation on A. thaliana|
|FZB42 : B. amyloliquefaciens FZB42, AB108 : abrB mutant, AB108C : Complementation of abrB, AB108R : Retransformation of abrB. Values shown represent the mean of fresh weight ± standard error. Different letters indicate means that differ significantly (P ≤ 0.05).|
The RBAM_017410 is a small gene of 185 bp where its function is still unknown. However, it is homologous (76%) to ribonucleoside-diphosphate reductase small subunit in B. subtilis. The region of RBAM_017410 inserted with TnYLB-1 and its flanking region is shown Fig. 32.
|Figure 32. Genomic organization of RBAM_017410 region carrying the TnYLB-1 insertion and its flanking regions.|
To complement the RBAM_017410 mutant, the RBAM_017410 gene (186 bp) plus 261 bp upstream region and 123 downstream region was amplified using 410-up-SacII and 410-dw-Eco91I primers, containing SacII and Eco91l restriction site, respectively. The resulting fragment (586 bp) was cloned into linearized SacII/Eco91I pUC18 plasmid which had an Amy cassette (pVBF) (Fig. 33). The ligated DNA was transformed into CaCl2-competent DH5α cells. To verify the insertion site, the plasmid DNA extracted from DH5α was amplified using 410-up-SacII and 410-dw-Eco91I primers. The plasmid with correct insertion fragment was then transformed into RBAM_017410 mutant.
|Figure 33. Strategy for construction of pUC18-∆ RBAM_017410 cassette.|
Amplification of FZB42 chromosomal DNA by PCR using primer 410-up-SacII and 410-dw-Eco91I which amplified RBAM_017410 gene showed a fragment around 500 bp length, whereas the mutant revealed a fragment 2000 bp in length, indicating that there was insertion of the TnYLB-1 transposon (Fig. 34). Complementation to replace the disrupted RBAM_017410 gene was done by inserting the intact gene from the wild type. Two fragments were obtained in complementation when its chromosomal DNA was amplified with primer 410-dw-Eco91I and 410-up-SacII, suggesting the intact gene has been successfully inserted. Retransformation was also done to verify that the RBAM_017410 gene was interrupted by transposon insertion. This was done by transforming chromosomal DNA mutant to the FZB42 wild type. Amplifying the chromosomal DNA of retransformation showed that the gene from mutant has been transformed into wild type (Fig. 34).
|Figure 34. PCR product of RBAM_017410 gene|
|Wild type FZB42 (lane1), RBAM_017410 mutant (lane 2), complementation of RBAM_017410 (lane 3) and retransformation of RBAM_017410 (lane 4).|
In the preliminary screening, the RBAM_017410 mutant showed reduction in plant growth promoting activity. Application of bacterial culture of the mutant reduced significantly (p ≤ 0.05) the growth of L. minor compared to wild type. The complementation of RBAM_017410 significantly improved (p ≤ 0.05) the dry weight of the plants compared to control and mutant, however it showed no significant difference to the retransformation. Bacterial culture of FZB42 revealed significant (p ≤ 0.05) increase compare to all treatment in Lemna biotes system (Fig. 35 and 36).
|Figure 35. Influence of RBAM_017410 mutation on plant growth promoting ability B. amyloliquefaciens FZB42 on L. minor.|
|The figure shows the representative plant response to the bacterial treatment at 10 days. A. B. amyloliquefaciens FZB42, B. Control, C. RBAM_017410 mutant, D. Complementation of RBAM_017410 , E. Retransformation of RBAM_017410.|
|Figure 36. Growth stimulating effects of RBAM_017410 mutation on L. minor|
|FZB42 : B. amyloliquefaciens FZB42, AB107 : RBAM_017410 mutant, AB107C : Complementation of RBAM_017410 , AB107R : Retransformation of RBAM_017410. Values shown represent the mean of dry weight ± standard error. Different letters indicate means that differ significantly (P ≤ 0.05).|
Immersion the roots of A. thaliana in bacterial cultures of the RBAM_017410 mutant reduced the growth of the plant significantly (p ≤ 0.05) compared to wild type. The application of the complementation revealed no significant increase of plant fresh weight compared to control and mutant, but showed significance increase (p ≤ 0.05) compared to retransformation. The treatment of wild type and complementation showed no significant differences in plant growth promotion capability (Fig. 37 and 38).
|Figure 37. Influence of RBAM-017410 mutation on plant growth promoting ability of B. amyloliquefaciens FZB42 on A. thaliana|
|The figure shows the representative of plant growth to the bacterial treatment at 21 days. A. B. amyloliquefaciens FZB42, B. Control, C. RBAM_017410 mutant, D. Complementation of RBAM_017410 , E. Retransformation of RBAM_017410.|
|Figure 38. Growth stimulating effects of RBAM_017410 mutation on A. thaliana|
|FZB42 : B. amyloliquefaciens FZB42, AB107 : RBAM_017410 mutant, AB107C : Complementation of RBAM_017410, AB107R : Retransformation of RBAM_017410. Values shown represent the mean of dry weight ± standard error. Different letters indicate means that differ significantly (P ≤ 0.05).|
It is generally assumed that colonization of plant tissue is a crucial step in plant- bacteria interactions. Here I showed the ability of B. amyloliquefaciens FZB42 and its mutants to colonize the roots of A. thaliana in a gnotobiotic system using a scanning electron microscope (SEM) and a confocal laser scanning microscope (CLSM). Labeling of B. amyloliquefaciens FZB42 with GFP protein was performed via homologous recombination (Ben et al. 2011). The GFP-tagged chromosomal DNA of FZB42 was then used to transfer the GFP into the mutants generating GFP-labeled mutants. The colonization of FZB42 and its mutants tagged with GFP were studied using CLSM in A. thaliana roots.
After seven days of incubation in growth chamber, colonization of B. amyloliquefaciens FZB42 and its mutants on the roots of A. thaliana was observed under SEM and CLSM. The results from CLSM showed that the wild type completely colonized the whole root surface including the root tip and the shed border cells (Fig. 39 A), the growing side root (Fig. 39 B) as well as the junctions between epidermal cells and root hairs (Fig. 39 C). Further observation by SEM revealed morphology of the wild type which formed an extracellular polymeric matrix around the bacterial cells (Fig. 40), which seemed to involve in adhering bacteria to the surface of the roots, besides it enabled the bacteria to form multicelullar aggregation on the roots. The shape of the wild type cell appeared as a compact rod, covered with polymeric matrix when colonizing the root.
|Figure 39. CLSM image of B. amyloliquefaciens FZB42 on A. thaliana root|
|7 days after inoculation. Left images show an overlay for orientation. A) Root tip of the main root; FZB42 colonizes the whole root surface including the root tip as well as the border cells. B) Growing side root; FZB42 colonizes the side root and the whole root surface; C) Epidermal cells with root hairs; FZB42 colonizes in junctions between epidermal cells and on root hairs (by courtesy of Barbara Beator).|
|Figure 40. SEM of B. amyloliquefaciens FZB42 colonizing A. thaliana roots|
|Note the polymeric matrix surround bacterial cells as indicated by the arrows A. Cells with 25.00K magnification, B. Cells with 10.00K magnification.|
Contrary to the wild type which colonized the whole root surface, all mutants impaired their ability to colonize the root of A. thaliana. Two biofilm mutants used in this experiment (yusV and degU mutants) showed ability to colonize only in some part of the root. The yusV mutant was found on the border cells but not in the root tip, surface of the root or side root (Fig. 41 A). However, there was a clump of bacterial cells opposite the cells (Fig. 41 B). This clump might originate from colonized border cells which detached during root preparation. The cells of yusV mutants showed significantly reduction in extracellular polymeric matrix production and most of the cells were not encased by this matrix as in FZB42 (Fig. 42 A). There was a clear shape difference between the FZB42 and yusV mutant, where the mutant appeared slender rod instead of compact rod.
|Figure 41. CLSM image of yusV mutant on A. thaliana roots|
|7 days after inoculation. Left images show an overlay for orientation. A) Root tip; The root surface and the root tip are not colonized. yusV mutant colonizes the border cells. B) Side root; The side root and the root surface are not colonized (by courtesy of Barbara Beator).|
|Figure 42. SEM of yusV mutant colonizing A. thaliana roots|
|A. Cells with 25.00K magnification, B. Cells with 10.00K magnification.|
As for the yusV mutant, the biofilm degU mutant showed similar root colonization, where it was found in small amount in some border cells but not in the root tip, the root surface or side root (Fig. 43 A-B). The shape of the degU mutant cell appeared as slim rods when colonizing the root and the polymeric matrix connected bacterial cells (Fig. 44 A-B).
|Figure 43. CLSM image of degU mutant on A. thaliana roots|
|7 days after inoculation. Left images show an overlay for orientation. A) Root tip; The root surface and root tip are not colonized but degU mutant colonizes the border cells. B) Side root; The root surface and the side root are not colonized (by courtesy of Barbara Beator).|
|Figure 44. SEM of degU mutant colonizing A. thaliana roots|
|Note the polymeric matrix connecting bacterial cells as indicated by the arrow A. Cells with 25.00K magnification, B. Cells with 10.00K magnification.|
Unlike the wild-type strain, the PGPR nfrA mutant failed to colonize the entire root surface (Fig. 45 A-B). It only colonized a small region of the roots that were root surface and border cells. SEM of nfrA mutant colonizing plant root showed reduction of extracellular polymeric matrix formation. The shape of the cells was slender rod and there were many dead bacteria cells among the colony (Fig. 46 A-B).
|Figure 45. CLSM image of nrfA mutant on A. thaliana roots|
|7 days after inoculation. Left images show an overlay for orientation. A) Root tip; The root surface and root tip are not colonized but nfrA mutant colonizes the border cells. B) Side root; The root surface is colonized but not the side root (by courtesy of Barbara Beator).|
|Figure 46. SEM of nfrA mutant colonizing A. thaliana roots|
|Note the dead bacterial cells as indicated by the arrows A. Cells with 25.00K magnification, B. Cells with 10.00K magnification|
Only few and small regions of the roots were colonized by the PGPR abrB mutant and significantly reduced extracellular polymeric matrix formation was observed compared with the wild type. Colonization was only detected in the border cells but not in the root tip or the root surface next to the tip (Fig. 47). The shape of the abrB mutant cell when colonizing the root was slender rod and many empty cells appeared in the colony (Fig. 48).
|Figure 47. CLSM image of abrB mutant on A. thaliana roots|
|7 days after inoculation. The left image shows an overlay for orientation. Root tip; The root surface and root tip are not colonized but abrB mutant colonizes the border cells (by courtesy of Barbara Beator).|
|Figure 48. SEM of abrB mutant colonizing A. thaliana roots|
|Note the dead bacterial cells as indicated by the arrows A. Cells with 25.00K magnification, B. Cells with 10.00K magnification|
The PGPR RBAM_017410 mutant was not able to be seen with CLSM when colonizing the root of A. thaliana (Fig. 49). This might be due to low emission of the GFP protein in this mutant. However, I could characterize the RBAM_017410 mutant by SEM while colonizing the root showing the slender rod shape of the cell with reduction of extracellular polymeric matrix production (Fig. 50).
|Figure 49. CLSM image of RBAM_017410 mutant on A. thaliana roots|
|7 days after inoculation. Left|
|Figure 50. SEM of RBAM_017410mutant colonizing A. thaliana roots|
|A. Cells with 25.00K magnification, B. Cells with 10.00K magnification.|
The interesting finding from this colonization experiment was a distinct mode of colonization between the wild type and the mutants. Whilst the wild type colonized the whole root surface including the root tip and the border cells as well as the growing side root, the mutants only colonized small region of the root. When comparing the cell shape of the wild type and the mutants colonizing the root of A. thaliana, clear differences became visible. The shape of the wild type cell was dumpy rod and extracellular polymeric matrix encased cells (Fig. 40), whereas the cell morphology of all the mutants were mostly slender rod with reduced extracellular polymeric matrix formation. There was not much difference in the cell shape among mutants, except that in nfrA mutant and abrB mutant many empty cells appeared (Fig. 46 and 48).
In order to determine which metabolites released by B. amyloliquefaciens FZB42 when interacted with L. minor, I decided to detect the metabolites released by mass spectrometric analysis. MALDI-TOF MS was carried out in Steinberg medium and L. minor inoculated with B. amyloliquefaciens FZB42 and the mutants (incubated for seven days at 22ºC). In this analysis I found production of surfactin in the Steinberg medium and in the L. minor when inoculated with wild type and two other mutants (degU and auxotroph mutant). Surfactin is a biosurfactant which displays an array of amazing activities, such as hemolytic, hypocholesterolemic agent, antitumoral, antimicrobial and antiviral. The biological role of surfactin is thought as supporting colonization of surfaces and acquisition of nutrients through their surface-wetting and detergent properties (Peypoux et al. 1999).
Analysis of the Steinberg medium containing L. minor and wild type showed the peaks that corresponded to surfactin production (m/z 1046.8, 1060.8 and 1074.8) (Fig. 51 A). Detection of L. minor extract grown in Steinberg medium inoculated with wild type revealed production of the similar metabolite (Fig. 51 B). Analysis of the Steinberg medium and the L. minor extract inoculated with auxotroph mutant indicated that surfactin was also produced in this treatment (Fig. 51 C-D). Inoculation of the degU mutant into the Steinberg medium planted with L. minor showed that Steinberg medium as well as L. minor extract also contained surfactin even though in the latter only two peaks of surfactin were detected (m/z 1060.8 and 1074.8) (Fig. 51 E-F).
B. amyloliquefaciens FZB42 is a potent producer of polyketides, such as bacillaene, difficidin, and macrolactin, as well as cyclic lipopeptides, such as surfactins, fengycins, and Bacillomycin D. These compounds are biosynthesized in a 4’-phosphopantetheine transferase (Sfp)-dependent fashion, while the production of the antibacterial dipeptide bacilysin is independent of Sfp (Koumoutsi et al. 2004; Chen et al. 2006; Chen et al. 2007; Chen et al. 2009c). From the genome analysis of B. amyloliquefaciens FZB42 there was no evidence for ribosomally produced antibacterial substances such as lantibiotics and other bacteriocins (Chen et al. 2007), however, Butcher and Helmann (2006) showed that the sfp mutant CH3, that can not produce polyketides and lipopeptides, produced at least one substance that was effective against Bacillus subtilis CU1065 and particularly against its sigW mutant HB0042.
Using tranposon TnYLB-1, I tried to investigate the capacity of B. amyloliquefaciens FZB42 for the production of secondary metabolites of ribosomal origin. For this purpose, I have used the sfp mutant strains CH5 which lost the ability to produce all nonribosomally formed metabolites of FZB42 that were synthesized according to the thiotemplate mechanism by knocking-out the sfp-4’-phosphopantetheine transferase gene responsible for cofactor loading to the thiotemplate reaction centers (Chen, 2009). The B. amyloliquefaciens FZB42 mutant strain CH5 was transferred with transposon TnYLB-1 using the same method as for wild type transformation. All mutants were then screened in subsequent spot-on-lawn tests on Bacillus subtilis HB0042 for the loss of antibacterial activity.
In this screening, several mutants were found unable to produce antibacterial compounds. One of the mutants, WY01 (Fig 52), was selected and studied further. Sequencing the WY01 mutant revealed a gene cluster around RBAM_029230 with hypothetical proteins. Based on MALDI-TOF mass spectrometry, the product of this gene cluster was bacteriocin. Extensive blast of known bacteriocin genes against the cluster uncovered the similarity to the circular bacteriocin uberolysin. For this reason it was assumed that the product of the gene cluster around RBAM_029230 is a circular peptide, and this is the antibacterial substance against Bacillus subtilis. This new antibacterial peptide is named “amylocyclin A” (Scholz, 2011). The insertion of TnYLB-1 transposon in the degU gene of the CH5 strain (RSpMarA2) produced a substance which was identified as a thiazole/oxazole-modified microcin (TOMM). This novel metabolite was also a ribosomally synthesized antibacterial substance and named as “plantazolicin” (Scholz et al. 2011). Both new antibiotics are narrow spectrum antibacterial compounds which have antagonistic activity towards closely related Gram-positive ba cilli.
|Figure 52. Spot on lawn test of WY01 mutant|
|A. CH5, B. WY01 mutant. Note clear zone was lost in the mutant (By courtesy of Zhiyuan Wang).|
Many bacteria such as Pseudomonas aeruginosa, P. fluorescens CHA0, Bacillus thuringiensis and Paenibacillus macerans (Ali et al. 2002; Siddiqui et al. 2005 ; Wei et al. 2003 and Oliveira et al. 2009) have the ability to produce active nematocidal metabolites against nematode pests. Here the nematocidal activity of B. amyloliquefaciens FZB42 was tested and the mutants were screened to discover the gene sresponsible for nematodical production. Transposon TnYLB-1 mutagenesis was used to generate the mutants of B. amyloliquefaciens FZB42 and the transposant were then screened to find the mutants that impaired nematocidal activity.
The screening was performed in China (Quisheng He – Yunan University, Kunming - China), four mutants were found which had reduction of nematocidal activity compared to FZB42 (Table 7). In this screening the survival rate of the nematode Caenorhabditis elegans was counted every four, eight and twenty hours and then compared with the wild type. Blast analysis of the DNA sequence form the mutants showed that the transposon disrupted the prkA, yhdY, RBAM_007470, ywmC and mlnD genes.
|© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.|
|DiML DTD Version 4.0||Zertifizierter Dokumentenserver|
der Humboldt-Universität zu Berlin
|HTML-Version erstellt am: