Genetic Analysis of the Plasmid-Based Temperature-Lethal Mutant pa1792|lpxH(Ts) in Pseudomonas aeruginosa

Many enzymes in the Raetz pathway for lipid A biosynthesis in Escherichia coli are essential. A homologous protein Pa1792|LpxH in Pseudomonas aeruginosa is known to complement the loss of LpxH in E. coli. Genome-wide transposon-insertion sequencing analysis indicates that lpxH is essential in P. aeruginosa. However, genetic analysis of lpxH in P. aeruginosa has not been carried out, partly because the conditional alleles of essential genes are not readily constructed. In this study, we first constructed a plasmid-based temperature-sensitive mutant ΔlpxH/pTS-lpxH or lpxH(Ts) in P. aeruginosa PAO1. Spot-plating assay indicated that lpxH(Ts) was lethal at a restrictive temperature, confirming its essentiality for growth. Microscopic analysis revealed that lpxH(Ts) exhibited an oval-shaped morphology, suggesting that lpxH was required for rod-shape formation. SDS-PAGE and Western blotting analysis showed that lpxH(Ts) failed to synthesize lipid A, consistent with its function in lipid A biosynthesis. Strong expression of lpxH but not the non-homologous isoenzyme lpxI or lpxG impeded growth and caused cell lysis, implying that lpxH-specific cofactors were required for this toxic effect in P. aeruginosa. Together, our results demonstrate that lpxH is essential for lipid A biosynthesis, rod-shaped growth, and viability in P. aeruginosa. We propose that this plasmid-based conditional allele is a useful tool for the genetic study of essential genes in P. aeruginosa.


Introduction
Pseudomonas aeruginosa is a rod-shaped Gram-negative bacterium that is widely found in soil and water.It is an opportunistic pathogen that attacks immunocompromised patients.Due to its intrinsic antibiotic resistance, it has been a top killer in intensive care units [1].The World Health Organization has listed it as one of the three top pathogens critically required for research and the development of new antibiotics [2].
Lipopolysaccharide (LPS), also known as endotoxin, is a major non-proteinaceous component of the cell wall in most of the Gram-negative bacteria [3][4][5][6].LPS consists of three domains: lipid A, core oligosaccharide, and O-specific antigen polysaccharide or O-antigen, in which lipid A serves as an anchor on the outer leaflet of the outer membrane in the Gram-negative bacterial cell envelope [3][4][5][6].
The lipid A biosynthesis pathway also known as the Raetz pathway consists of nine enzymes involved in the biosynthesis of Kdo 2 -lipid A in Escherichia coli [3,7].All of the first six enzymes, namely LpxA, LpxC, LpxD, LpxH, LpxB, and LpxK, are essential for synthesizing the minimal structure lipid IV A needed for viability in many Gram-negative bacteria [3,7,8].It has been an attractive target for the development of antibiotics [9,10].
Many Gram-negative bacteria contain homologous sequences with LpxA, LpxC, LpxD, LpxB, and LpxK in E. coli, but not LpxH [11].It has been shown that some bacteria contain non-homologous isofunctional sequences of LpxI or LpxG, instead of LpxH [11,12].Genetic analysis indicates that the sequence of lpxI from Caulobacter crescentus [11] or lpxG from Chlamydia trachomatis is capable of complementing the loss of lpxH function in E. coli [12].However, genetic analysis of lpxH mutants in P. aeruginosa has not been performed, largely caused by conditional alleles that may not be readily obtained.To circumvent this issue, we have adopted the three-step protocol for the construction of the plasmid-based conditional lethal allele of essential genes in P. aeruginosa [13,14].
In this study, we show that the plasmid-based temperature-sensitive mutant ∆lpxH/pTS-lpxH or lpxH(Ts) that we constructed was lethal at the restrictive temperature, confirming its essentiality for growth.Genetic phenotypes of the lpxH(Ts) mutant are presented and their implications are discussed.Together, our results demonstrate that the plasmidbased conditional lethal allele is a useful tool for the genetic analysis of essential genes in P. aeruginosa.

Plasmid Construction
We used the same deletion plasmid and rescue plasmid (or TS plasmid) constructed in the previous study [13,14].For the construction of the lpxH deletion and rescue plasmids, the deletion cassette and rescue cassette of lpxH were cloned in a deletion plasmid and a rescue plasmid using the homologous recombinase cloning kit (ClonExpress II one-step cloning kit; Vazyme, Nanjing, China).For the construction of overexpression plasmids, the araC-P BAD promoter fragment and downstream gene fragment were cloned in the pBBR1MCS-5 plasmid [15] using the Vazyme cloning kit.

Plasmid-Based ts-Mutant Strain Construction
We used a three-step protocol that we developed previously [13,14] to construct the plasmid-based ts-lethal mutant strain ∆pa1792/pTS-pa1792.

Spot-Plating Assay
The spot-plating assay [16] was adopted to test sensitivities to temperature.In brief, 10-fold serial-diluted cultures were transferred using a 48-pin replicator (V&P Scientific, Inc., San Diego, CA, USA) onto LB plates supplemented with appropriate stress factors and incubated at 30 or 42 • C as indicated.

LPS Preparation
A standard method of LPS preparation by Hitchcock and Brown [17] was followed.Briefly, 1 mL of fresh culture (0.8 OD 600 ) was harvested and pelleted.The cell pellets were resuspended in 250 µL lysis buffer (1.0 M Tris, pH 6.8, 10% vol/vol glycerol, 2% wt/vol SDS, 4% β-mercaptoethanol) and boiled at 96 • C for 10 min.Proteinase K (Cat# 39450-01-6, Sinopharm) was added to each sample to a final concentration of 200 mg/mL prior to incubation at 55 • C for 4 h.Subsequently, the sample temperature was raised to 70 • C for 1 h to inactivate proteinase K and prepare it for SDS-PAGE analysis.

SDS-PAGE and Silver Staining
LPS samples were resolved on the discontinuous (12% and 15%) gradient SDS-PAGE gel (1.0 mm thickness, 15 wells) to detect O-antigen, core oligosaccharides, and lipid A. Gels were stained by the silver staining method of Fomsgaard et al. [18] before being imaged on a Gel Doc system (Bio-Rad, Hercules, CA, USA).

Spontaneous Mutagenesis Screening
Approximately 1.0 × 10 +09 ∆pa1792/pTS-pa1792 cells (i.e., colony forming units calculated using a method reported by Zhang et al. [19]) were plated via spreading onto LB plates and incubated at the semi-restrictive temperature of 40 • C for 2 weeks for suppressor cells through spontaneous mutations as previously described [13,14,20].Suppressor colonies were validated via streaking on fresh LB plates and incubation at 42 • C. PCR assays using sequence-specific primers for pa1972 alleles were performed for the validation of the pa1792 deletion allele.To investigate the genetic phenotype of the essential gene pa1792|lpxH in P. aeruginosa, we constructed the plasmid-based temperature-sensitive mutant by using a three-step protocol [13,14].In brief, a schematic map of the deletion plasmid pUC-sacB-gen R containing the lpxH deletion allele and permissive temperature rescue plasmid pTS-oriTS-tet R containing the native promoter controlled wild type allele of lpxH were shown in Figure 1A.PCR analysis confirmed that the deletion plasmid contained no lpxH wild type allele (Figure 1B, see lane pDel) and both the deletion allele ∆lpxH and wild type allele lpxH + were present in the lpxH(Ts) mutant (Figure 1B, see lanes ts) when primer pair F1/R1 was applied.On the other hand, the analysis confirmed that the deletion allele ∆lpxH was only present on the chromosome (Figure 1C, see lanes ts) when the primer pair F2/R2 was used.The spot-plating assay with serial-diluted cells indicated that the ts-mutant lpxH(Ts) but not the wild type failed to grow at the restrictive temperature of 42 • C (Figure 1D), confirming its essentiality for growth [21].The growth curve analysis was consistent with the spot-plating assay (Figure 1E), concluding that the plasmid-based temperature-sensitive allele of the essential genes is suitable for genetic analysis in P. aeruginosa.

lpxH(Ts) Exhibits an Oval-Shaped Morphology at the Restrictive Temperature
Subsequently, we performed microscopic analysis to assess the terminal morphological phenotype of lpxH(Ts) at the restrictive temperature.As a control, we showed that the rod-shaped morphology of wild type cells at 30 °C was hardly altered after growth at 42 °C (Figure 2A).The cellular morphology of the mutant lpxH(Ts) cells at the permissive

lpxH(Ts) Exhibits an Oval-Shaped Morphology at the Restrictive Temperature
Subsequently, we performed microscopic analysis to assess the terminal morphological phenotype of lpxH(Ts) at the restrictive temperature.As a control, we showed that the rod-shaped morphology of wild type cells at 30 • C was hardly altered after growth at 42 • C (Figure 2A).The cellular morphology of the mutant lpxH(Ts) cells at the permissive temperature resembled that of the wild type (Figure 2B, see top row).However, 3 h after growth at the restrictive temperature, lpxH(Ts) exhibited an oval-shaped morphology (Figure 2B, see mid row).We found that 6 h after growth at the restrictive temperature, some of the lpxH(Ts) cells displayed the nearly round-shaped morphology (Figure 2B, see bottom row).This result indicates that lpxH is required for rod-shape morphogenesis in P. aeruginosa.

Growth Defect of lpxH(Ts) at the Restrictive Temperature Is Rescued by the araC-PBAD Promoter-Controlled E. coli lpxH or ec.lpxH at the Leaky Expression Level
It was shown that the P. aeruginosa LpxH protein or Pa.LpxH shared 47% identity with that of Ec.LpxH [22,23], a UDP-2,3-diacylglucosamine hydrolase involved in the lipid A biosynthetic pathway [22].Babinski et al. [22] showed that the pa.lpxH sequence could compensate for the loss of E. coli lpxH.However, it was not known if the ec.lpxH sequence could complement the loss of P. aeruginosa lpxH.Hence, we investigated if the ec.lpxH sequence could compensate for the loss of P. aeruginosa lpxH.
To this end, the ec.lpxH coding sequence PCR-amplified from E. coli K12 strain was cloned under the control of the araC-PBAD promoter [24] in the multi-host pBBR1MCS-5 plasmid [15] to generate the pOE-ec.lpxHplasmid.The spot-plating assay indicated that under no inducer arabinose, the multi-host plasmid pOE-ec.lpxHat the leakage expression level was sufficient to rescue the growth defect of lpxh(Ts) at 42 °C, similar to the positive control plasmid pOE-pa.lpxH(Figure 3A, see arrow).The negative control plasmid pOE showed no effect.This result validated that P. aeruginosa lpxH was a functional ortholog of E. coli lpxH.On the other hand, we found that strong overexpression (i.e., with

lpxH at the Leaky Expression Level
It was shown that the P. aeruginosa LpxH protein or Pa.LpxH shared 47% identity with that of Ec.LpxH [22,23], a UDP-2,3-diacylglucosamine hydrolase involved in the lipid A biosynthetic pathway [22].Babinski et al. [22] showed that the pa.lpxH sequence could compensate for the loss of E. coli lpxH.However, it was not known if the ec.lpxH sequence could complement the loss of P. aeruginosa lpxH.Hence, we investigated if the ec.lpxH sequence could compensate for the loss of P. aeruginosa lpxH.
To this end, the ec.lpxH coding sequence PCR-amplified from E. coli K12 strain was cloned under the control of the araC-P BAD promoter [24] in the multi-host pBBR1MCS-5 plasmid [15] to generate the pOE-ec.lpxHplasmid.The spot-plating assay indicated that under no inducer arabinose, the multi-host plasmid pOE-ec.lpxHat the leakage expression level was sufficient to rescue the growth defect of lpxh(Ts) at 42 • C, similar to the positive control plasmid pOE-pa.lpxH(Figure 3A, see arrow).The negative control plasmid pOE showed no effect.This result validated that P. aeruginosa lpxH was a functional ortholog of E. coli lpxH.On the other hand, we found that strong overexpression (i.e., with 0.2% arabinose supplementation) of ec.lpxH or pa.lpxH impeded the growth of both the ts-mutant lpxH(Ts) and the wild type cells (Figure 3A, see rectangle).The growth curve analysis supported the observation that the leaky expression level of lpxH rescued the growth of the ts-mutant lpxH(Ts) at the restrictive temperature and strong overexpression impeded the growth of the wild type (Figure 3B).Microscopic analysis showed that the oval-shaped morphology of lpxH(Ts) at the restrictive temperature was restored to the rod-shaped morphology by the leakage expression level of ec.lpxH-OE (Figure 3C, see upper right penal).On the other hand, some rod-shaped cells appeared to be lysed upon the strong expression of ec.lpxH (Figure 3C, see bottom right penal).These results indicated that the level of lpxH was tightly controlled; without LpxH, it was lethal with oval-shaped terminal morphology and too much LpxH was also lethal, displaying lysed rod-shaped cells.

Lipid A Is Missing in lpxH(Ts) at the Restrictive Temperature in P. aeruginosa
Ec.LpxH, a UDP-2,3-diacylglucosamine hydrolase in E. coli, was essential for lipid A biosynthesis [22].To investigate whether or not lipid A was synthesized in lpxH(Ts) at the restrictive temperature in P. aeruginosa, we performed SDS-PAGE and Western blotting analysis (see Section 2).For this reason, the wild type and lpxH(Ts) mutants cultivated at 30 °C and 42 °C were collected for LPS preparation using the HB method [17].Based on our previous study [14], while O-antigen signals were found above the position of the 35 kD pre-stained protein marker, core oligosaccharide signals were positioned below the 20 kD protein marker.It should be noted that protein molecular weight markers were not meant for the estimation of the LPS molecular weight but for position alignment of LPS molecules in SDS-PAGE gels and Western blots.
To permit all LPS components such as O-antigen, core oligosaccharide, and lipid A to be displayed in a single gel, we utilized a discontinued (12% and 15%) gradient SDS-PAGE gel, in which the gel electrophoresis was stopped when the pre-stained 25 kD protein marker migrated to the boundary between the 12% and 15% SDS-PAGE gels.Subsequently, the gel was subjected to silver staining after electrophoresis (Figure 4A).We no- The growth curve analysis supported the observation that the leaky expression level of lpxH rescued the growth of the ts-mutant lpxH(Ts) at the restrictive temperature and strong overexpression impeded the growth of the wild type (Figure 3B).Microscopic analysis showed that the oval-shaped morphology of lpxH(Ts) at the restrictive temperature was restored to the rod-shaped morphology by the leakage expression level of ec.lpxH-OE (Figure 3C, see upper right penal).On the other hand, some rod-shaped cells appeared to be lysed upon the strong expression of ec.lpxH (Figure 3C, see bottom right penal).These results indicated that the level of lpxH was tightly controlled; without LpxH, it was lethal with oval-shaped terminal morphology and too much LpxH was also lethal, displaying lysed rod-shaped cells.

Lipid A Is Missing in lpxH(Ts) at the Restrictive Temperature in P. aeruginosa
Ec.LpxH, a UDP-2,3-diacylglucosamine hydrolase in E. coli, was essential for lipid A biosynthesis [22].To investigate whether or not lipid A was synthesized in lpxH(Ts) at the restrictive temperature in P. aeruginosa, we performed SDS-PAGE and Western blotting analysis (see Section 2).For this reason, the wild type and lpxH(Ts) mutants cultivated at 30 • C and 42 • C were collected for LPS preparation using the HB method [17].Based on our previous study [14], while O-antigen signals were found above the position of the 35 kD pre-stained protein marker, core oligosaccharide signals were positioned below the 20 kD protein marker.It should be noted that protein molecular weight markers were not meant for the estimation of the LPS molecular weight but for position alignment of LPS molecules in SDS-PAGE gels and Western blots.
To permit all LPS components such as O-antigen, core oligosaccharide, and lipid A to be displayed in a single gel, we utilized a discontinued (12% and 15%) gradient SDS-PAGE gel, in which the gel electrophoresis was stopped when the pre-stained 25 kD protein marker migrated to the boundary between the 12% and 15% SDS-PAGE gels.Subsequently, the gel was subjected to silver staining after electrophoresis (Figure 4A).We noticed that the LPS sample prepared from the lpxH(Ts) mutant at 42 • C hardly showed any silver staining signals.To ensure an unbiased LPS preparation, cell lysates with equal amounts of total proteins were utilized for LPS extraction.To ascertain that the levels of LPS components such as O-antigen, core oligosaccharide, and lipid A were high enough for detection in Western blotting analysis, 3-or 4-fold higher amounts of the samples derived from the cultures at 42 • C were loaded on the gel (Figure 4A, see lane 5 and lane 6).LPS components such as O-antigen, core oligosaccharide, and lipid A were high enough for detection in Western blotting analysis, 3-or 4-fold higher amounts of the samples derived from the cultures at 42 °C were loaded on the gel (Figure 4A, see lane 5 and lane 6).LPS on a duplicate SDS-PAGE gel was transferred to a PVDF membrane.Molecules on the upper portion of the membrane transferred from the 12% SDS-PAGE gel were hybridized with the monoclonal antibodies (mAb) MF15-4 against O-antigen (Figure 4B).On the other hand, molecules on the lower membrane portion transferred from the 15% SDS-PAGE gel were first hybridized with mAb 26-5 against lipid A (Figure 4C) and subsequently re-hybridized with mAb 5C7-4 against the inner core after stripping (Figure 4D).To compare the relative localization of the lipid A and inner core signals, the pseudo-colored blot images were superimposed based on the protein markers.The result showed that lipid A migrated in front of the inner core (Figure 4E).We found that lipid A, core oligosaccharide, and O-antigen were not detected in lpxH(Ts) at the restrictive temperature (Figure 4, see lane 4 and lane 6).This result suggests that lpxH in P. aeruginosa is essential for lipid A biosynthesis, similar to that of lpxH in E. coli [22].Additionally, without lipid A or a membrane anchor of LPS, other components such as core oligosaccharide LPS on a duplicate SDS-PAGE gel was transferred to a PVDF membrane.Molecules on the upper portion of the membrane transferred from the 12% SDS-PAGE gel were hybridized with the monoclonal antibodies (mAb) MF15-4 against O-antigen (Figure 4B).On the other hand, molecules on the lower membrane portion transferred from the 15% SDS-PAGE gel were first hybridized with mAb 26-5 against lipid A (Figure 4C) and subsequently re-hybridized with mAb 5C7-4 against the inner core after stripping (Figure 4D).To compare the relative localization of the lipid A and inner core signals, the pseudo-colored blot images were superimposed based on the protein markers.The result showed that lipid A migrated in front of the inner core (Figure 4E).We found that lipid A, core oligosaccharide, and Oantigen were not detected in lpxH(Ts) at the restrictive temperature (Figure 4, see lane 4 and lane 6).This result suggests that lpxH in P. aeruginosa is essential for lipid A biosynthesis, similar to that of lpxH in E. coli [22].Additionally, without lipid A or a membrane anchor of LPS, other components such as core oligosaccharide and O-antigen were also missing in lpxH(Ts) at the restrictive temperature.The non-homologous UDP-2,3-diacylglucosamine hydrolyses such as Cc.LpxI in Caulobacter crescentus and Ct.LpxG in Chlamydia trachomatis were found to be able to compensate for the loss of LpxH in E. coli [11,12].To investigate whether these non-homologous UDP-2,3-diacylglucosamine hydrolyses cc.lpxI and ct.lpxG could also compensate for the loss of lpxH in P. aeruginosa, we constructed the overexpression plasmids pOE-cc.lpxIand pOE-ct.lpxG,similar to that of pOE-ec.lpxH(see Figure 3).The spot-plating assay indicated that the expression of cc.lpxI and ct.lpxG with or without arabinose rescued the growth defect of lpxH(Ts) at the restrictive temperature (Figure 5A, see arrow).Unlike the homolog ec.lpxH, strong overexpression of the non-homologs cc.lpxI and ct.lpxG was unable to impede cell growth, suggesting a subtle difference between the non-homologous isofunctional enzymes LpxH and LpxI or LpxG.We found that growth curve analysis was consistent with the spot-plating assay (Figure 5B).The non-homologous UDP-2,3-diacylglucosamine hydrolyses such as Cc.LpxI in Caulobacter crescentus and Ct.LpxG in Chlamydia trachomatis were found to be able to compensate for the loss of LpxH in E. coli [11,12].To investigate whether these non-homologous UDP-2,3-diacylglucosamine hydrolyses cc.lpxI and ct.lpxG could also compensate for the loss of lpxH in P. aeruginosa, we constructed the overexpression plasmids pOE-cc.lpxIand pOE-ct.lpxG,similar to that of pOE-ec.lpxH(see Figure 3).The spot-plating assay indicated that the expression of cc.lpxI and ct.lpxG with or without arabinose rescued the growth defect of lpxH(Ts) at the restrictive temperature (Figure 5A, see arrow).Unlike the homolog ec.lpxH, strong overexpression of the non-homologs cc.lpxI and ct.lpxG was unable to impede cell growth, suggesting a subtle difference between the non-homologous isofunctional enzymes LpxH and LpxI or LpxG.We found that growth curve analysis was consistent with the spot-plating assay (Figure 5B).SDS-PAGE and Western blotting analysis showed that the expression of homolog ec.lpxH and non-homologs cc.lpxI and ct.lpxG restored the lipid A biosynthesis in lpxH(Ts) at the restrictive temperature (Figure 5C-E and Figure S1), indicating that the expression of homolog or non-homolog enzymes UDP-2,3-diacylglucosamine hydrolyses not only rescues the growth defect but also restores the biosynthesis of lipid A in lpxH(Ts) at the restrictive temperature.To investigate whether or not lpxH mutation could be suppressed, we performed a spontaneous mutagenesis by spreading the lpxH(Ts) mutant cells (>1.0 × 10 +09 ) on LB plates incubated at the semi-restrictive temperature of 40 • C as we did previously [13,14].The spot-plating assay confirmed that the sup mutant could grow at 42 • C as that of the wild type (Figure 6A).However, sup failed to growth at 30 • C, suggesting that mutations in an unknown suppressor gene were temperature sensitive.This could be explained that at the restrictive temperature, the rescue plasmid pTS-lpxH in lpxH(Ts) was lost and sup was defective, which rescued the lethality of ∆lpxH.In contrast, at the permissive temperature, sup was effective and failed to suppress the growth defect of ∆lpxH.
SDS-PAGE and Western blotting analysis showed that the expression of homolog ec.lpxH and non-homologs cc.lpxI and ct.lpxG restored the lipid A biosynthesis in lpxH(Ts) at the restrictive temperature (Figures 5C-E and S1), indicating that the expression of homolog or non-homolog enzymes UDP-2,3-diacylglucosamine hydrolyses not only rescues the growth defect but also restores the biosynthesis of lipid A in lpxH(Ts) at the restrictive temperature.

Lethality of lpxH(Ts) at the Restrictive Temperature Is Suppressed by a Sup Mutant Identified from Spontaneous Mutagenesis
To investigate whether or not lpxH mutation could be suppressed, we performed a spontaneous mutagenesis by spreading the lpxH(Ts) mutant cells (>1.0 × 10 +09 ) on LB plates incubated at the semi-restrictive temperature of 40 °C as we did previously [13,14].The spot-plating assay confirmed that the sup mutant could grow at 42 °C as that of the wild type (Figure 6A).However, sup failed to growth at 30 °C, suggesting that mutations in an unknown suppressor gene were temperature sensitive.This could be explained that at the restrictive temperature, the rescue plasmid pTS-lpxH in lpxH(Ts) was lost and sup was defective, which rescued the lethality of ΔlpxH.In contrast, at the permissive temperature, sup was effective and failed to suppress the growth defect of ΔlpxH.To test this possibility, we showed that only the ΔlpxH allele but no wild type allele was detected in the sup mutant based on the PCR assay (Figure 6B).Growth curve analysis supported the observation in the spot-plating assay.Microscopic analysis showed that sup To test this possibility, we showed that only the ∆lpxH allele but no wild type allele was detected in the sup mutant based on the PCR assay (Figure 6B).Growth curve analysis supported the observation in the spot-plating assay.Microscopic analysis showed that sup exhibited rod-shaped morphology at 42 • C, which restored the oval-shaped morphology in the ∆lpxH mutant (Figure 6D, see bottom row).In contrast, when sup cells were shifted from 42 • C to 30 • C, oval-shaped cells started to appear, suggesting that it failed to suppress ∆lpxH whose phenotype started to appear (Figure 6D, see upper row).
Subsequently, we investigated whether or not lipid A biosynthesis in sup was restored at 42 • C. For this reason, we performed SDS-PAGE and Western blotting analysis.The analysis indicated that the O-antigen, core, and lipid A were all detected in sup at 42 • C, indicating that lipid A biosynthesis was restored in the sup mutant (Figures 6E-G and  S2).However, we were unable to identify the sup gene sequence using referenced genome re-sequencing analysis.It was possible that the sup locus in the strain we studied was not identical to that in the reference genome of PAO1 [25].

Discussion
In this study, by using a plasmid-based ts-mutant lpxH(Ts) in P. aeruginosa PAO1, we showed that lpxH(Ts) failed to grow at the restrictive temperature with an oval-shaped terminal morphology (see Figures 1 and 2).It has been shown that not all UDP-2,3diacylglucosamine hydrolyses in Gram-negative bacteria are encoded by lpxH [11,12].Non-homologous lpxI in C. crescentus and lpxG in C. trachomatis are found to encode the isofunctional enzyme UDP-2,3-diacylglucosamine hydrolase [11,12].
Though both LpxH and LpxI catalyze UDP-Diacylglucosamine hydrolysis to form lipid X and UMP, the former attacks the α-phosphate and the latter targets the β-phosphate during the catalytic reaction [11].Like LpxH, LpxG also attacks α-phosphate to hydrolyze UDP-Diacylglucosamine [12].We show, in this study, the leaky expression level of the araC-P BAD promoter-controlled lpxH in P. aeruginosa and E. coli and their non-homologous isofunctional sequences; lpxI in C. crescentus and lpxG in C. trachomatis rescues the lethality of lpxH(Ts) at the restrictive temperature (see Figure 5).This result suggests that the release of lipid X from UDP-Diacylglucosamine is sufficient for continuing the Raetz pathway regardless of the α-phosphate or β-phosphate it attacks.
However, inconsistent with this, under the strong expression level (i.e., induced by 0.2% arabinose supplementation), lpxH, but not lpxI or lpxG, impedes cell growth (see Figure 5).It is known that the strong expression of lpxG impedes growth by cell lysis in C. trachomatis [12].These results indicate that the strong expression of lpxH and lpxG hinders growth and leads to cell lysis in P. aeruginosa and C. trachomatis, respectively (This study, [12]).We propose that this toxic effect by the strong expression of lpxH and lpxG requires lpxHand lpxG-specific cofactors in P. aeruginosa and C. trachomatis, respectively.This would explain why the strong expression of lpxG shows no deleterious effect in P. aeruginosa (see Figure 5).
It has been shown that disruption of the first two genes lpxA and lpxC in the Raetz pathway is suppressible by a mutation in fabZ that is involved in FASII [26].However, it is unclear whether or not other essential genes in the pathway are suppressible.In this study, we have tested the suppressibility of lpxH by suppression analysis using the ts-mutant lpxH(Ts) (see Figure 6).Though we isolated a putative sup colony which survived at 42 • C, we failed to identify the sup gene sequence currently.
In conclusion, by using the plasmid-based temperature-sensitive mutant lpxH(Ts), we experimentally confirmed that lpxH in P. aeruginosa is essential for viability.LpxH is required for lipid A biosynthesis and rod-shaped growth.The strong expression of lpxH but not the non-homologous isofunctional UDP-2,3-diacylglucosamine hydrolase impedes growth and causes cell lysis in P. aeruginosa.Our study provides insight into the genetic basis of lpxH function at various expression levels.We propose that the plasmid-based temperature-sensitive mutant is a useful tool for the genetic analysis of essential genes in P. aeruginosa.

1 .
Plasmid-Based Temperature-Sensitive Mutant lpxH(Ts) Is Lethal at the Restrictive Temperature in P. aeruginosa

Genes 2024 , 13 Figure 1 .
Figure 1.Analysis of plasmid-based ts-mutant lpxH(Ts) reveals its essentiality for growth.(A) A schematic map of the pa1797|lpxH locus in P. aeruginosa PAO1.Locations of CDS, primer, deletion allele, native promoter controlled wild type allele are shown.(B) PCR assay using primer pair F1/R1 specific to lpxH alleles in both chromosome and plasmid.(C) PCR assay using primer pair F2/R2 specific to lpxH alleles in chromosome only.(D) Spot-plating assay showing lpxH(Ts) growth defect at 42 °C.(E) Growth curve analysis showing lpxH(Ts) growth defect at 42 °C.

Figure 1 .
Figure 1.Analysis of plasmid-based ts-mutant lpxH(Ts) reveals its essentiality for growth.(A) A schematic map of the pa1797|lpxH locus in P. aeruginosa PAO1.Locations of CDS, primer, deletion allele, native promoter controlled wild type allele are shown.(B) PCR assay using primer pair F1/R1 specific to lpxH alleles in both chromosome and plasmid.(C) PCR assay using primer pair F2/R2 specific to lpxH alleles in chromosome only.(D) Spot-plating assay showing lpxH(Ts) growth defect at 42 • C. (E) Growth curve analysis showing lpxH(Ts) growth defect at 42 • C.

3. 3 .
Growth Defect of lpxH(Ts) at the Restrictive Temperature Is Rescued by the araC-P BAD Promoter-Controlled E. coli lpxH or ec.

Figure 3 .
Figure 3. Leaky but not strong expression level of E. coli lpxH rescues lpxH(Ts) at 42 • C in P. aeruginosa.(A) Spot-plating assay.Leaky expression level of lpxH rescues lpxH(Ts) at 42 • C (see arrow).Strong expression level of lpxH impedes cell growth (see rectangle).(B) Growth curve analysis.X-and Y-axes indicate time (h) and cell density (OD 600 ).(C) Cell morphology analysis.Arrow shows the lysed cells.

Figure 4 .
Figure 4. LpxH is required for lipid A biosynthesis in P. aeruginosa.(A) A silver-stained SDS-PAGE gel image.LPS components including lipid A, core oligosaccharide, and O-antigen are separated on a discontinued gradient SDS-PAGE gel.(B) Western blot image.Molecules derived from the upper 12% SDS-PAGE gel portion are probed with mAb MF15-4 against O-antigen.(C) Western blot images.Molecules derived from the bottom 15% SDS-PAGE gel portion are probed with mAb 26-5 against lipid A. (D) Image of the re-probed blot in (C).The blot is re-probed with mAb 5C7-4 against the inner core oligosaccharide after stripping.(E) The merge of the blot (C) in red and (D) in green.

Figure 4 .
Figure 4. LpxH is required for lipid A biosynthesis in P. aeruginosa.(A) A silver-stained SDS-PAGE gel image.LPS components including lipid A, core oligosaccharide, and O-antigen are separated on a discontinued gradient SDS-PAGE gel.(B) Western blot image.Molecules derived from the upper 12% SDS-PAGE gel portion are probed with mAb MF15-4 against O-antigen.(C) Western blot images.Molecules derived from the bottom 15% SDS-PAGE gel portion are probed with mAb 26-5 against lipid A. (D) Image of the re-probed blot in (C).The blot is re-probed with mAb 5C7-4 against the inner core oligosaccharide after stripping.(E) The merge of the blot (C) in red and (D) in green.

Figure 5 .
Figure 5. lpxH(Ts) is rescued by the leaky and strong expression of lpxI or lpxG and by the leaky but not strong expression of lpxH.(A) Spot-plating assay.Leaky expression level of ec.lpxH, cc.lpxI, and ct.lpxG rescues lpxH(Ts) at 42 °C (see arrow).Strong expression level of lpxH, but not lpxI or lpxG impedes cell growth (see rectangle).(B) Growth curve analysis.X-and Y-axes indicate time (h) and cell density (OD600).(C) Silver-stained SDS-PAGE gel image.(D) Image of Western blot probed with mAb MF15-4 against O-antigens.(E) Merged image of Western blot probed with mAb 26-5 against lipid A (red) and re-probed with mAb 5C7-4 against the inner core (green).

Figure 5 .
Figure 5. lpxH(Ts) is rescued by the leaky and strong expression of lpxI or lpxG and by the leaky but not strong expression of lpxH.(A) Spot-plating assay.Leaky expression level of ec.lpxH, cc.lpxI, and ct.lpxG rescues lpxH(Ts) at 42 • C (see arrow).Strong expression level of lpxH, but not lpxI or lpxG impedes cell growth (see rectangle).(B) Growth curve analysis.X-and Y-axes indicate time (h) and cell density (OD 600 ).(C) Silver-stained SDS-PAGE gel image.(D) Image of Western blot probed with mAb MF15-4 against O-antigens.(E) Merged image of Western blot probed with mAb 26-5 against lipid A (red) and re-probed with mAb 5C7-4 against the inner core (green).

3. 6 .
Lethality of lpxH(Ts) at the Restrictive Temperature Is Suppressed by a Sup Mutant Identified from Spontaneous Mutagenesis

Figure 6 .
Figure 6.Characterization of a putative suppressor sup of lpxH.(A) Spot-plating assay.(B) PCR assay showing the sup mutant lacking a wild type allele.(C) Growth curve analysis.(D) Cell morphology analysis.Oval-shaped morphology was restored to rod-shaped cells in sup at 42 °C.At 30 °C, sup exhibits oval-shaped morphology.(E) Silver-stained SDS-PAGE gel analysis.(F) Western blotting image.Molecules derived from the upper portion of the SDS-PAGE gel are probed with mAb MF15-4 against O-antigen.(G) Merged image of Western blots.Molecules derived from the lower portion of the SDS-PAGE gel are probed with mAb 26-5 against lipid A (red) and re-probed with mAb 5C7-4 against the inner core (green).

Figure 6 .
Figure 6.Characterization of a putative suppressor sup of lpxH.(A) Spot-plating assay.(B) PCR assay showing the sup mutant lacking a wild type allele.(C) Growth curve analysis.(D) Cell morphology analysis.Oval-shaped morphology was restored to rod-shaped cells in sup at 42 • C. At 30 • C, sup exhibits oval-shaped morphology.(E) Silver-stained SDS-PAGE gel analysis.(F) Western blotting image.Molecules derived from the upper portion of the SDS-PAGE gel are probed with mAb MF15-4 against O-antigen.(G) Merged image of Western blots.Molecules derived from the lower portion of the SDS-PAGE gel are probed with mAb 26-5 against lipid A (red) and re-probed with mAb 5C7-4 against the inner core (green).

Table 1 .
Oligonucleotides, plasmids, and strains used in this study.