Construction of chromosomal gene deletion mutants

Strain AK1602, which has a deletion of the rcsB gene and a Cm^R cassette, was constructed by the one-step inactivation method 13 using primers A2414 and A2415 with pKD3 as the template.

Strain AK1603, which has a deletion of the klcA1 gene and a tetA, was constructed by the one-step inactivation method 13 using primers A2418 and A2419 with the phoQ::Tn10 chromosomal DNA as the template.

Strain AK1599, which has an insertion of the Km^R cassette behind the wzc gene, was constructed by the one-step inactivation method 13 using primers A2416 and A2417 with pKD4 as the template.

Strain AK1604, which has a deletion of the klcA2 gene and a Km^R cassette, was constructed by the one-step inactivation method 13 using primers A2420 and A2421 with pKD4 as the template.

Strain AK1605, which has a deletion of the dgeC gene and a Km^R cassette, was constructed by the one-step inactivation method 13 using primers A2424 and A2425 with pKD4 as the template.

Plasmid construction

Plasmid RSFRedTER-Sp for λ Bet Exo Gam expression was constructed as a spectinomycin-resistant version of RSFRedTER 18 by the one-step inactivation method 13 using primers A1190 and A1191 with EG16468 chromosomal DNA 23 as the Sp^R marker template.

All strains constructed using PCR reactions were analyzed by DNA sequencing to confirm that the PCR-generated DNA regions had the predicted sequences.

Transformation efficiency analysis

Bacterial cells cultured overnight in LB (Lennox) were added to 10 ml of fresh medium diluted 1:50 in test tubes and shaken at 37°C until the OD₅₉₅ value reached 0.3~0.8. Cells were collected by centrifugation and the supernatant was removed. The cell pellet was washed twice with 1 ml of 10% glycerol, and then the cells were suspended in 400 µl of 10% glycerol. 1-3 µl of plasmid DNA was mixed with 100 µl of competent cells (the cell suspension) and electroporated using a 2 mm cuvette and a Gene Pulser II (Bio-Rad) at 2.5 kV, 200 Ω, and 25 µF. Immediately after the pulse, SOB medium was added to the cuvette and collected in 1.5 ml tubes. Cells were spread on LB (Lennox) plates containing kanamycin (pAK1001 and pACYC177) or tetracycline (pCF430) and incubated at 37°C overnight. Colonies were then counted.

Genome sequencing, assembly and annotation

Whole genome sequencing was performed using DNBSEQ-G400 (MGI Tech Co., Ltd.). Libraries were prepared using IMGIEasy FS DNA Library Prep Set (MGI Tech Co., Ltd.) and fragments were validated using Fragment Analyzer and dsDNA 915 Reagent Kit (Agilent Technologies). Sequences were de novo assembled using SPAdes assembler version 3.15.5 42. Genome annotation was performed using DFAST.

Possible genes involved in plant growth promoting traits in the E. asburiae i6 genome

Although strain i6 was not originally isolated as PGPB, a list of candidate genes that may contribute to plant growth was compiled from the genome sequence of strain i6 (Table ​(Table5).5). Soil beneficial bacteria can promote plant growth by synthesizing molecules similar to plant hormones 26, 27. Auxin like indole acetic acid (IAA) is quantitatively the most plant hormones secreted by Phytophthora rhizobacteria 28, 29, suggesting that auxin is a signaling molecule in microorganisms 30. Similar to the genome features previously reported for Enterobacter sp. J49 30, genes responsible for the synthesis of indole-pyruvate decarboxylase and indole-3-acetaldehyde dehydrogenase were detected in the i6 genome, and no other IAA pathway-related genes were detected. The bacterial volatiles 2,3-butanediol and acetoin are also known to induce plant growth promotion 31. Acetolactate synthase BudB converts pyruvate to acetolactate, which is subsequently converted to acetoin by acetoin decarboxylase BudA. The (S)-acetoin-forming diacetyl reductase BudC catalyzes the conversion of acetoin to 2,3-butanediol, which is reversible. All genes for butA, butB and butC were found, as well as other genes encoding acetolactate synthases (ilvB, ilvN_2, ilvG etc.). Similar to the J49 genome 30, the gcd gene encoding GDH synthesis, responsible for the production of gluconic acid, the major organic acid in the phosphate solubilization mechanism most widely used by soil bacteria, was detected in the i6 genome, but the pqq gene cluster, required for the biosynthesis of the PQQ cofactor, was not detected. Redundant genes for siderophore-production, iron ABC transporters, and type VI secretion systems which may compete with (plant) pathogens for low iron levels and cell growth, were also found in the i6 genome (data not shown).

Genes involved in restriction and anti-restriction in the E. asburiae i6 genome

Interestingly, KlcA1 and KlcA2, two homologues of KlcA, which have been reported to function as an anti-type I restriction modification (RM) system in other species 32, were predicted on this genome and plasmid respectively, but only one type IV R enzyme was predicted to be encoded and no type I RM system (Table ​(Table6).6). This is in contrast to other E. asburiae strains ATCC 35953, FDAARGOS_892, and RHBSTW-01009, which encode two type I RM systems, one type IV restriction enzyme, and KlcA1 (Table ​(Table6).6). L1 also encodes a type I RM system, a type IV R enzyme, and other restriction systems, but no KlcA homologs. Because it is important for successful genetic manipulation to suppress restriction enzymes and allow bacteria to introduce foreign plasmids, the transformation efficiency of several plasmid DNAs was tested by electroporation using the i6 and ATCC 35953 strains (Fig. ​(Fig.11).

Open in a separate window

Figure 1

Transformation efficiency of E. asburiae strains i6, i6 klcA1 (AK1603), i6 klcA2 (AK1604), and ATCC 35953 (NBRC 109912 ^T) and E. cloacae ATCC 1304 (ENC) strain with pACYC177 (dark bar), pAK1001 (gray bar), and pCF430 (white bar). Transformation efficiency (colony count/µg plasmid DNA) was calculated by counting colonies after electroporation, recovery in SOB for 1 h, and overnight growth at 37°C on LB (Lennox) plates containing kanamycin (pACYC177 and pAK1001) or tetracycline (pCF430).

The results showed that the ATCC 35953 strain could not be transformed by the relatively large plasmid DNAs pCF430 (10 kb) 33 and pAK1001 (8.4 kb) 34, whereas i6 formed colonies, showing transformation efficiency at least four orders of magnitude higher than that of ATCC 35953 especially with pCF430 (Fig. ​(Fig.1).1). Using a relatively small plasmid DNA, pACYC177 (3.9 kb) 35, the ATCC 35953 strain also formed colonies on LB plates containing kanamycin, and the transformation efficiency of i6 was two orders of magnitude higher than that of ATCC 35953 (Fig. ​(Fig.11).

Application of the one-step gene inactivation method to the i6 strain

Unlike the ATCC 35953 strain, the i6 strain was able to accept foreign DNA with high efficiency (Fig. ​(Fig.1),1), so I attempted to apply the one-step gene inactivation method 36. Prior to this, RSFRedTER-Sp was constructed by replacing the Cm^R marker of the RSFRedTER plasmid 18, which can express λ Red recombinase, with the Sp^R marker, so that the Cm^R marker could be used in addition to the Km^R and Tc^R markers in the i6 strain. (Ap-resistant pKD46 was not used because two β-lactamase genes were predicted in the i6 genome.) Gene deletions in klcA1, klcA2, and dgeC were constructed in the i6 strain expressing λ Red recombinase from RSFRedTER-Sp using Tc^R, Cm^R, and Km^R markers, respectively. Similarly, a deletion in rcsB and a Km^R insertion behind wzc were made in the i6 strain using Cm^R and Km^R markers, respectively. The recombinants of interest grew with high colony formation rates (Table ​(Table7).7). Colony PCR in the junction region of the recombination site confirmed that the desired recombinants were constructed with high accuracy (Table ​(Table7).7). The klcA1 and klcA2 strains were included in the transformation efficiency analysis because KlcA was originally reported as a restriction enzyme inhibitor in other bacteria, such as E. coli and Klebsiella pneumoniae 32, and may have contributed to the high transformation efficiency of the plasmid DNA of strain i6. However, contrary to expectations, these deletions increased rather than decreased the transformation efficiency of some plasmid DNA (Fig. ​(Fig.1).1). Yet, this result may have been confounded by the fact that the type I RM system, a potential target of KlcAs, is not encoded in the i6 genome. In conclusion, this report successfully applied the one-step gene inactivation method to the newly identified E. asburiae strain i6, whose transformation efficiency was much higher than that of ATCC 35953.

Among the ECC, E. cloacae and E. hormaechei are the most frequently isolated species in clinical infections, especially in immunocompromised patients and those admitted to intensive care units 37, whereas E. asburiae had lower survival rates against serum than E. cloacae, E. hormaechei, and E. ludwigii isolates 4. In this report, the E. asburiae i6 genome unexpectedly exhibited most, if not all, of the genetic features of PGPB previously reported in Enterobater sp. J49. This was also true for the genome closest to i6, ATCC 35953, FDAARGOS_892, L1, RHBSTW-01009 (data not shown). Thus, E. asburiae is a preferred PGPB candidate in the ECC. Even if the current form of the i6 strain is not optimal in promoting plant growth, it would be advantageous and useful as a starting material for testing and fine-tuning any beneficial and detrimental traits. This is because the i6 strain, with all its potential as a PGPB, can be successfully genetically engineered. First of all, potential risk factors such as genes responsible for antibiotic resistance, biofilm formation, and some of the type VI secretion apparatus/effectors, etc. should be deleted by the one-step gene disruption method or its derivative, scarless genome editing 38. Such a risk negative i6 derivative strain could then be chemically mutagenized to express PGPB factors at optimal levels. Alternatively, or in addition, the ability of strain i6 to be a PGPB can be further enhanced by incorporating plasmid clones harboring other PGPB factors, even from different species, but only as genetically modified organisms. On the other hand, the amenability of this strain must also be useful to identify the relevant genes by deleting candidate genes from this genome that could produce a signal molecule that activates RcsB in Salmonella or even within the i6 strains. The recently developed conjugation-mediated versatile site-specific single-copy luciferase fusion system 39, which is broadly applicable to Gram-negative bacteria, should also be helpful in detecting intercellular interactions between the i6 strain and Salmonella, etc., and even within the i6 strains.

I thank Dr. Yoshihiko Hara for providing the plasmid RSFRedTER.