The Effects of HP0044 and HP1275 Knockout Mutations on the Structure and Function of Lipopolysaccharide in Helicobacter pylori Strain 26695

Helicobacter pylori infection is associated with several gastric diseases, including gastritis, peptic ulcer, gastric adenocarcinoma and mucosa-associated lymphatic tissue (MALT) lymphoma. Due to the prevalence and severeness of H. pylori infection, a thorough understanding of this pathogen is necessary. Lipopolysaccharide, one of the major virulence factors of H. pylori, can exert immunomodulating and immunostimulating functions on the host. In this study, the HP0044 and HP1275 genes were under investigation. These two genes potentially encode GDP-D-mannose dehydratase (GMD) and phosphomannomutase (PMM)/phosphoglucomutase (PGM), respectively, and are involved in the biosynthesis of fucose. HP0044 and HP1275 knockout mutants were generated; both mutants displayed a truncated LPS, suggesting that the encoded enzymes are not only involved in fucose production but are also important for LPS construction. In addition, these two gene knockout mutants exhibited retarded growth, increased surface hydrophobicity and autoaggregation as well as being more sensitive to the detergent SDS and the antibiotic novobiocin. Furthermore, the LPS-defective mutants also had significantly reduced bacterial infection, adhesion and internalization in the in vitro cell line model. Moreover, disruptions of the HP0044 and HP1275 genes in H. pylori altered protein sorting into outer membrane vesicles. The critical roles of HP0044 and HP1275 in LPS biosynthesis, bacterial fitness and pathogenesis make them attractive candidates for drug inventions against H. pylori infection.


Introduction
Helicobacter pylori (H. pylori) is a Gram-negative, microaerophilic spiral-shaped bacterium which infects more than half of the world's population [1][2][3]. As a notorious human pathogen, H. pylori employs multiple virulence factors, such as CagA, VacA, UreA and lipopolysaccharide (LPS), for environmental adaptation and host colonization during the process of host-microbe interaction [4][5][6]. It has been considered as one of the major etiological factors for the development of gastritis, peptic ulcer, gastric cancer and mucosaassociated lymphoid tissue (MALT) lymphoma, and is ranked as a group one carcinogen by the World Health Organization (WHO) [7][8][9].
One of the major characteristics of Gram-negative bacteria is the presence of LPS on the bacterial outer membrane (OM) [10]. As a virulence factor, LPS is essential for the survival between wild-type strain 26695 and the knockout mutants, thus suggesting that the sorting of proteins into OMVs is also affected by the HP0044 and HP1275 mutations. The current findings reveal the function and effect of GMD and PMM proteins in H. pylori in great detail and add a new understanding for future drug development against H. pylori infection.

Bacterial Strains, Plasmids and Growth Conditions
The bacterial strains and plasmids used in this study are listed in Table 1. The H. pylori wild-type strain 26695 (ATCC 700392; CagA + , VacA + ) and the knockout mutants used in this study were grown on sheep blood (5% w/v) agar plates and incubated for 48 h to 72 h at 37 • C under microaerophilic conditions (5% O 2 , 10% CO 2 and 85% N 2 ). To change the growth conditions from solid to liquid media, bacteria were scraped off from agar plates and resuspended in Brucella broth media (BD Biosciences, Franklin Lakes, NJ, USA) with 10% fetal bovine serum (FBS, Biological Industries, Kibbutz Beit Haemek, Israel) and 1% IsoVitalex (Creative, Taipei, Taiwan). The flasks were then placed into an anaerobic chamber and cultivated for 48 h under a microaerophilic atmosphere at 37 • C with shaking at 140 r.p.m. Table 1. Bacterial strains and plasmids used in this study.

Strain/Plasmid Characteristics and Marker 1 Source
Strain E. coli

Top10
The host for construction of pGEM-T-HP0044 and pGEM-T-HP1275 clones Invitrogen, Carlsbad, CA, USA

DH5α
The host for construction of pGEM-T-HP0044 d ::Cm r and pGEM-T-HP1275 d ::Cm r clones

Construction of the HP0044 and HP1275 Knockout Mutants and the Corresponding Knockout Complementary Mutants
The HP0044 or HP1275 knockout mutants were constructed using a strategy based on gene splicing by overlap extension (SOEing) [36][37][38]. First, the upstream (582 base pairs (bps)) and downstream (583 bps) regions of the HP0044 gene in addition to the upstream (679 bps) and downstream (796 bps) regions of the HP1275 gene were amplified from the genomic DNA of H. pylori wild-type strain 26695 using primer pairs KO0044F1-KO0044R1/KO0044F2-KO0044R2 and KO1275F1-KO1275R1/KO1275F2-KO1275R2, respectively ( Table 2). The two sets of generated polymerase chain reaction (PCR) products were then each joined via SOEing PCR with primer pairs KO0044F2/KO0044R1 or KO1275F2/KO1275R1 with a BamHI restriction site in the PCR products for subsequent insertion of the antibiotic resistance cassette to disrupt the HP0044 and HP1275 genes. The amplified 1165 bps fragment of HP0044 and the amplified 1475 bps fragment of HP1275 were purified and cloned into the pGEM-T vector (Promega, Madison, WI, USA) to generate the pGEM-T-HP0044 plasmid or the pGEM-T-HP1275 plasmid. Later, a chloramphenicol resistance cassette (Cm r ) was inserted into the pGEM-T-HP0044 or the pGEM-T-HP1275 plasmid that had been digested with a BamHI restriction enzyme and treated with shrimp alkaline phosphatase (NEB, Ipswich, MA, USA). The resulting plasmid, pGEM-T-HP0044 d ::Cm r or pGEM-T-HP1275 d ::Cm r , was confirmed by restriction digestion analysis, followed by natural transformation into H. pylori wild-type strain 26695.  -TAAGGAGTTACGATATGAAAGAAAAAATCGCTTT  Com0044R2 2  KpnI  AAGGTACCTCAGTGGTGATGGTGATGATGTTCATAAAAATTCCT  KO1275F1 3  -AAACGCATGGCAAAATTTATGCG  KO1275R1 3  BamHI  CCCAGGATCCTCAGGATCCG  KO1275F2 3  BamHI  CTGAGGATCCTGGGAATTTC  KO1275R2 3  -GGGATATGAAATGAGATTATCCC  Com1275F1 4 NcoI CCATGGCTTGATTGGAAGCACTAGCCACG Com1275R1 4 -ATGCTAATGTCCATATCGTAACTCCTTAAGTGTT Com1275F2 4 -TAAGGAGTTACGATATGGACATTAGCATTTTTAG Com1275R2 4 KpnI AAGGTACCTTAGTGGTGATGGTGATGATGAAGTTTTTCTAATAA 1 Primers used in the construction of the HP0044 knockout mutant. 2 Primers used in the construction of the HP0044 knockout complementary mutant. 3 Primers used in the construction of the HP1275 knockout mutant. 4 Primers used in the construction of the HP1275 knockout complementary mutant. 5 The sequences of restriction enzyme sites are underlined.
To complement the aforementioned HP0044 and HP1275 knockout mutations, gene complementation was conducted by inserting the complementary genes into the locus of the RdxA (HP0954) gene. A 1561 bps fragment containing the promoter region of the HP1563 gene (343 bps) and the open reading frame (ORF) of the HP0044 gene (1173 bps) was PCR-amplified from H. pylori 26695 genomic DNA using the primer pairs Com0044F1-Com0044R1 and Com0044F2-Com0044R2, respectively (Table 2). Similarly, the 1719 bps fragment containing the promoter region of the HP1563 gene (343 bps) and the ORF of the HP1275 gene (1376 bps) was PCR-amplified from H. pylori wild-type strain 26695 genomic DNA using the primer pairs Com1275F1-Com1275R1 and Com1275F2-Com1275R2, respectively ( Table 2). The two PCR products generated for each set were then joined via SOEing PCR with primers Com0044F1 and Com0044R2 for HP0044, or with primers Com1275F1 and Com1275R2 for HP1275. The resulting 1516 bps (for HP0044) and 1719 bps (for HP1275) products were then digested with NcoI/KpnI and ligated into the pGEM-T-RdxA L -T7 ter -RdxA R plasmid, which had been NcoI/KpnI-digested to create the pGEM-T-RdxA L -P HP1563 -HP0044-T7 ter -RdxA R plasmid and pGEM-T-RdxA L -P HP1563 -HP1275-T7 ter -RdxA R plasmid. The resulting complementary plasmids were then transformed into the HP0044 and HP1275 knockout mutants, respectively, by natural transformation.

Isolation of OMVs
OMVs were isolated from the supernatant of bacterial culture using a method described by Wai et al., with minor modifications [39]. Briefly, bacterial suspensions obtained from different strains of H. pylori harvested at the stationary phase (about 60 h of bacterial growth, the value of optical density at 600 nm (OD 600 ) = 2.0) in a total volume of 12.5 mL were centrifuged at 4000× g for 10 min at 4 • C. The obtained supernatants were filtered with 0.45 µm membranes and then ultracentrifuged at 20,000× g for 2 h at 4 • C to collect the OMVs (Micro Ultracentrifuge CS150NX, Hitachi Instruments Inc., Tokyo, Japan). The pellets containing OMVs were washed twice with PBS and resuspended in 20 mM Tris-HCl (pH 8.0) for further analysis.

LPS Isolation and LPS Profile Analysis
LPS from the tested H. pylori strains was isolated by the hot phenol-water method, with minor modifications [40]. In brief, different strains of H. pylori cultured in 1.5 mL of Brucella broth media at the log phase (OD 600 = 1.0) were harvested by centrifugation (4000× g for 10 min at 4 • C), and the whole-cell lysates containing LPS were extracted from the pellets by adding 600~800 µL of lysis buffer. One hundred and fifty microliters of the whole-cell lysate was added with 10 µL of proteinase K (20 mg/mL), and the resulting suspension was incubated at 60 • C for 120 min. Then, 150 µL of the prewarmed 90% hot phenol solution (70 • C) was added to the suspension and incubated in a water bath at 70 • C for 15 min with constant shaking or vortexing. Thereafter, the mixture was allowed to cool down on ice for 10 min and then subjected to centrifugation (4000× g for 10 min at 4 • C). After centrifugation, LPS was recovered from the aqueous phase. For silver staining, the purified LPS was separated on 15% trincine gel. The obtained gel was first fixed overnight in 150 mL of the fixing solution (25% isopropanol and 7% acetic acid), followed by incubation with the oxidation solution containing 7.5% acetic acid and 0.7% periodic acid for 10 min. The gel was then washed three times for 30 min each with distilled water, and stained with 150 mL of a freshly prepared staining solution (28 mL of 0.1 N NaOH, 2 mL of ammonium hydroxide and 5 mL of 20% silver nitrate) for 25 min. After staining, the gels were washed three times for 10 min each with distilled water and developed by reduction in 200 mL of the developing solution (0.05% citric acid and 0.074% formaldehyde) until the signal of LPS became visible.

Growth Curve Analysis of H. pylori
Various H. pylori strains were first grown on blood agar plates for 48 h, and then switched to Brucella broth media with 10% FBS (Biological Industries) and 1% IsoVitale X (Creative) under microaerophilic conditions at 37 • C for 16 h. The bacterial cultures were later diluted to an OD 600 value of 0.01 and incubated with shaking at 140 r.p.m. under microaerophilic conditions for up to 3 days. The value of OD 600 was measured by spectrophotometer at various time points, and the experiment was repeated in triplicate to obtain the averaged results and standard deviation.

SDS and Novobiocin Sensitivity Assay
Bacteria were harvested and diluted to an OD 600 value of 0.6 in Brucella broth media and 100 µL of cell suspension was added to 1900 µL of Brucella broth medium supplemented with an appropriate amount of the detergent SDS or the antibiotic novobiocin, with concentrations of up to 0.02% for SDS and 100 µg /mL for novobiocin. The assays were executed in triplicate with independent cultures in 24-well plates with shaking at 140 r.p.m. under microaerophilic conditions at 37 • C, and the OD 600 values were measured after 72 h of bacterial growth. The survival rate was defined as follows: survival rate (%) = (OD 600 of treatment/OD 600 of control) × 100.

Bacterial Infection Assay
AGS cells (ATCC 1739, human gastric adenocarcinoma epithelial cell line) were seeded at 2 × 10 5 cells/well in 6 cm dishes containing Ham's F-12 media (Sigma-Aldrich) supplemented with 10% FBS under standard cell culture conditions. When the AGS cells reached approximately 80% confluency, suspensions from different strains of H. pylori harvested at the log phase (OD 600 = 1.0) were added to the AGS cells at a multiplicity of infection (MOI) of 100. After 6 h of co-culturing the morphological changes of the cells were recorded by using an Olympus IX71 research inverted system microscope (Olympus, Tokyo, Japan) with 20× magnification.

Adhesion Assay
Following the bacterial infection described above, the dish with the AGS cell-bacterium co-culture was washed three times with warm Ham's F-12 media without FBS and mildly shaken to remove non-adherent bacteria. The AGS cells were subsequently lysed with 1 mL of PBS buffer containing 0.1% saponin to release the cell-associated bacteria. The obtained lysate was serially diluted and spread on sheep blood agar plates. The adhered bacteria were measured by the viable plate counting method after 72 to 96 h of incubation.

Internalization Assay
Following the bacterial infection described above, the dish with the AGS cell-bacterium co-culture was washed with a PBS buffer and the medium was then replaced by Ham's F12 medium with 100 µg/mL gentamicin for 1 h of incubation to kill the rest of the surfaceassociated, non-internalized bacteria. After incubation, the dish was washed with a PBS buffer and the medium was then replaced by Ham's F12 medium with 10 µg/mL gentamicin. The infected cells were further incubated for 24 h under standard culturing conditions. After incubation, cells were lysed with a 0.5% saponin-containing PBS buffer for 5 min. The lysate was then serially diluted and spread on sheep blood agar plates. The internalized bacteria were measured by the viable plate counting method after 72 to 96 h of incubation.

The Surface Hydrophobicity and Autoaggregation Assays
Bacteria were grown in Brucella broth media with 10% FBS and 1% IsoVitale X under microaerophilic conditions at 37 • C for 48 h. For the surface hydrophobicity assay, bacteria were harvested and diluted to reach an OD 600 value of 1.0 in a tube containing 5 mL of phosphate urea magnesium sulfate buffer (PUM: 22.2 g/L K 2 HPO 4 -3H 2 O, 7.26 g/L KH 2 PO 4 , 1.8 g/L urea, 0.02 g/L MgSO 4 -7H 2 O; pH 7.1) supplemented with 400 µL of n-hexadecane. The mixture was vortexed for 2 min to homogenize the bacterial suspension, and the tube containing bacteria was then left on the bench and stood up for 24 h. At the start of the experiment, the OD 600 of the bacterial suspension was measured as A 0 . During the test, 500 µL of solution from the top of the bacterial suspension was collected at 6, 8 and 24 h, and followed by measuring the OD 600 value as A t . The hydrophobicity index was defined as follows: the hydrophobicity index = (A 0 − A t )/A 0. For the autoaggregation assay, bacteria were harvested and diluted to reach an OD 600 value of 1.0 in a tube containing 6 mL of PBS, and the tube was then left on the bench and stood up for 24 h. The collection of bacterial suspension and the OD 600 recording were executed as described for the surface hydrophobicity assay. The ratio of autoaggregation was defined as follows: the autoaggregation rate (%) = (A 0 − A t )/A 0 × 100 . Each experiment was repeated in triplicate to obtain the average and standard deviation.

Sequence Analysis of the HP0044 and HP1275 Proteins in H. pylori 26695
The HP0044 coding sequence is located in the plus strand from 42,105 bps to 43,250 bps of the H. pylori genomic DNA, and is flanked by the HP0043 and HP0045 genes ( Figure 2a). Similarly, the HP1275 coding sequence is located in the plus strand from 1,350,206 bps to 1,351,585 bps of the H. pylori genomic DNA, and is flanked by the HP1274 and HP1276 genes ( Figure 2c). The HP0044 gene is 1146 bps in length, and can be translated into a protein with 381 amino acid residues. As for HP1275 gene, it contains 1380 bps and can be translated into a protein with 459 amino acid residues. The conserved domain of the HP0044 protein is found in the members of the NADB_Rossmann superfamily, which have the Rossmann-fold NAD(P)H/NAD(P) + binding (NADB) domain, and the conserved domain of the HP1275 protein belongs to the phosphohexomutase superfamily, which catalyzes a reversible intramolecular phosphoryl transfer of the sugar substrates (Figure 2b,d). The amino acid sequences of the HP0044 and HP1275 proteins and the GMD, PMM and PGM from different species of Gram-negative bacteria were aligned by using the LALIGN program, and the results are summarized in Table 3. Based on the results obtained from the sequence alignment, the HP0044 protein displays a high degree of sequence homology with GMD in other Gram-negative bacteria (61.7% identity for Yersinia enterocolitica 8081, 60.1% identity for Escherichia coli K12, 59.6% identity for Salmonella typhimurium CT18 and 61.0% identity for Vibrio cholerae O1 biovar El Tor N16961). Similarly, the HP1275 protein also showed a certain degree of homology with PMM and PGM in other Gramnegative bacteria (28.2% identity of PMM/PGM for Xanthomonas campestris pv. Vesicatoria, 38.9% identity of PMM/PGM for Pseudomonas aeruginosa PAO1, 27.4% identity of PMM for Salmonella typhimurium SL1344, 26.4% identity of PMM for Vibrio cholerae 2017V-1105, 21.6% identity of PGM for V. cholerae and 40.4% identity of PGM for Neisseria gonorrhoeae). This result indicated that the HP0044 protein could be a putative GMD that catalyzes the conversion of GDP-mannose into GDP-4-dehydro-6-deoxy-D-mannose, and the HP1275 protein could be a putative PMM/PGM that catalyzes the interconversion between α-Dmannose-6-phosphate and α-D-mannose-1-phosphate as well as α-D-glucose-6-phosphate and α-D-glucose-1-phosphate. The two enzymes could be involved in the biosynthesis of fucose and LPS in H. pylori ( Figure 1).

The HP0044 and HP1275 Proteins Are Critical for LPS Expression
To understand the role of the HP0044 and HP1275 proteins in H. pylori, the HP0044 and HP1275 knockout mutants and their corresponding knockout complementary mutants were constructed. The HP0044 and HP1275 genes in the H. pylori genomic DNA were disrupted by the insertion of a chloramphenicol resistance cassette (Cm r ) to generate the knockout mutants, and the resulting mutants were complemented by the insertion of a DNA fragment containing the promoter region of the HP1563 gene and the ORFs of the HP0044 or HP1275 gene into the rdxA gene region to generate the knockout complementary mutants. The successful generation of these mutants was verified by PCR (data not shown).
The recently redefined LPS structure in H. pylori is relatively linear, and the fucose residues appear not only in the O-antigen but also in the "Trio" region, which connects the core oligosaccharide to the LPS O-antigen. This suggests that a lack of fucose residues may lead to a truncated LPS structure shorter than that being considered previously. To explore the involvement of fucose in the construction of LPS, the effect of the HP0044and HP1275-disrupted mutations on LPS expression was evaluated by silver staining and immunoblotting analysis to examine the LPS profile and Lewis antigen moieties, respectively. As shown in Figure 3a, the complete loss of the O-antigen and an aberrant core region were observed in both HP0044 and HP1275 knockout mutants using tricine gel with silver staining. Furthermore, while testing with immunoblotting analysis to investigate the presence of Lewis antigens, the signals of Lewis X and Lewis Y antigens completely disappeared in both HP0044 and HP1275 knockout mutants (Figure 3b). The LPS truncation was fully restored in the HP1275 knockout complementary mutant while the HP0044 knockout complementary mutant showed slight abnormalities in silver staining and immunoblotting analysis, suggesting that the complementation of the HP0044 gene is not complete, possibly due to the lower level of expression of the complemented HP0044 protein. Interestingly, a slight difference between the LPS core region of the HP1275 knockout mutant and that of the HP0044 knockout mutant was also observed in Figure 3a. The reason for the difference between the LPS profiles of these two knockout mutants could result from another activity aquired by the HP1275 protein, which may potentially preform the enzymatic function of PGM, associating with the biosynthesis of glucose, which is also present in the side chain of the LPS core region in H. pylori (Figure 1). In summary, our results support the newly proposed LPS structure, that the fucose residue is involved in the "Trio" region of the LPS O-antigen in H. pylori wild-type strain 26695.
Biomedicines 2022, 9, x FOR PEER REVIEW 11 of 21 mutants could result from another activity aquired by the HP1275 protein, which may potentially preform the enzymatic function of PGM, associating with the biosynthesis of glucose, which is also present in the side chain of the LPS core region in H. pylori ( Figure 1). In summary, our results support the newly proposed LPS structure, that the fucose residue is involved in the "Trio" region of the LPS O-antigen in H. pylori wild-type strain 26695.

HP00044/HP1275 Gene Disruptions Affect Bacterial Growth, Sensitivity to the Detergent SDS and Resistance to the Antibiotic Novobiocin
To evaluate the impact of the HP0044-and HP1275-disrupted mutations on the growth of H. pylori, wild-type strain 26695, the HP0044 and HP1275 knockout mutants and the corresponding knockout complementary mutants were subjected to a growth assay. The growth curves were generated by recording the OD600 of bacteria, starting at an OD600 of 0.05 in Brucella broth supplemented with 10% FBS under microaerophilic conditions at 37 °C for 84 h. The data were recorded by a spectrophotometer for every 12 hr within this To evaluate the impact of the HP0044and HP1275-disrupted mutations on the growth of H. pylori, wild-type strain 26695, the HP0044 and HP1275 knockout mutants and the corresponding knockout complementary mutants were subjected to a growth assay. The growth curves were generated by recording the OD 600 of bacteria, starting at an OD 600 of 0.05 in Brucella broth supplemented with 10% FBS under microaerophilic conditions at 37 • C for 84 h. The data were recorded by a spectrophotometer for every 12 h within this period. The results showed that both HP0044 and HP1275 knockout mutants had a slight delay in the lag phase of bacterial growth and grew slower than wild-type strain 26695 and the corresponding knockout complementary mutants (Figure 4a). However, all of the tested strains reached their maximal culture densities after 48 h of growth and showed similar durations in each phase of bacterial growth. This observation suggested that the loss of either the HP0044 or HP1275 proteins can hamper H. pylori growth.

HP00044/HP1275 Gene Disruptions Affect Bacterial Growth, Sensitivity to the Detergent SDS and Resistance to the Antibiotic Novobiocin
To evaluate the impact of the HP0044-and HP1275-disrupted mutations on the growth of H. pylori, wild-type strain 26695, the HP0044 and HP1275 knockout mutants and the corresponding knockout complementary mutants were subjected to a growth assay. The growth curves were generated by recording the OD600 of bacteria, starting at an OD600 of 0.05 in Brucella broth supplemented with 10% FBS under microaerophilic conditions at 37 °C for 84 h. The data were recorded by a spectrophotometer for every 12 hr within this period. The results showed that both HP0044 and HP1275 knockout mutants had a slight delay in the lag phase of bacterial growth and grew slower than wild-type strain 26695 and the corresponding knockout complementary mutants (Figure 4a). However, all of the tested strains reached their maximal culture densities after 48 hr of growth and showed similar durations in each phase of bacterial growth. This observation suggested that the loss of either the HP0044 or HP1275 proteins can hamper H. pylori growth. Detergent sensitivity and antibiotic resistance are important determinants for the survival of bacteria in harsh environments. LPS on the OM is pivotal for Gram-negative bacteria to protect themselves from the damage of xenobiotics. Therefore, defective LPS may make bacteria lose their ability to withstand the stress exerted by toxic molecules, and thus die. Due to the importance of LPS in bacterial protection, SDS sensitiviy and novobiocin resistance assays were conducted to evaluate the impacts of the HP0044and HP1275-disrupted mutations on the susceptibility of H. pylori to detergents and hydrophobic antibiotics. As shown in Figure 4b,c, the survival rates of the HP0044 and HP1275 knockout mutants to SDS and novobiocin were both decreased more significantly than those of wild-type strain 26695 at various tested concentrations, suggesting that the tolerance of the knockout mutants to these two xenobiotics was compromised. As for the corresponding LPS-related knockout complementary mutants, the survival rates were restored to levels similar to those of wild-type strain 26695. The results indicated that the HP0044 and HP1275 proteins are required for the biosynthesis of LPS and thus the maintainence of the membrane integrity which is critical for bacterial protection against toxic molecules in harsh environmental conditions.

The HP0044-and HP1275-Disrupted Mutations Significantly Affect H. pylori Membrane Properties
The effects of the HP0044and HP1275-disrupted mutations on the bacterial membrane properties were examined by treating the tested H. pylori strains with an organic solvent and PBS for the surface hydrophobicity analysis and autoaggregation assay, respectively. The results showed that the surface hydrophobicity of both the HP0044 and HP1275 knockout mutants significantly increased at all of the time points tested (6 h, 8 h and 24 h) compared to that of wild-type strain 26695 (Figure 5a). A similar trend was also observed when the bacterial autoaggregation was measured (Figure 5b). The HP1275 knockout complementary mutant displayed a similar degree of surface hydrophobicity and bacterial autoaggregation compared to that of wild-type strain 26695, while the HP0044 knockout complementary mutant only showed a partial level of restoration. This observation agrees very well with the results obtained in the LPS profile and Lewis antigen analyses, indicating that the incomplete complementation of the HP0044 gene in the HP0044 knockout complementary mutant currently generated cannot fully recover all of the functions or structures that the fucose residues are involved in.

The HP0044-and HP1275-Disrupted Mutations Have Profound Influences on Bacterial Virulence
Studies have shown that the morphology of human gastric adenocarcinoma (AGS) cells would alter after H. pylori infection, and that the infected cells display a typical elongated and spreading appearance called the "hummingbird phenotype". To elucidate on whether the HP0044-and HP1275-disrupted mutations would affect the outcomes of bacterial infection, a cell infection assay was conducted. When the AGS cells were infected by wild-type strain 26695, the classical hummingbird phenotype indeed appeared (Figure 6a). Although the hummingbird phenotype of the AGS cells could also be observed when the AGS cells were infected with the HP0044 and HP1275 knockout mutants, the percentage of the AGS cells showing the hummingbird phenotype was significantly lower than that of the cells infected by wild-type strain 26695 (20% compared to 38.5%, Figure 6b). The appearance of the hummingbird phenotype of the infected AGS cells was restored when the cells were infected by the HP1275 knockout complementary mutant, while the restoration was less significant by the infection of the HP0044 knockout complementary mutant. These results suggested that the loss of the HP0044 and HP1275 proteins could lead to a reduction in the infection ability of H. pylori, thus indicating that the intactness of H. pylori LPS is highly correlated with the virulence of H. pylori.
Previous studies have reported that LPS is associated with the adhesion ability of many Gram-negative bacteria [33][34][35]41]. However, the involvement of LPS in H. pylori colonization is under debate. Therefore, bacterial adhesion and internalization assays were conducted to investigate the role of LPS in H. pylori colonization. As shown in Figure 6c, the disruption of the HP0044 and HP1275 genes significantly suppressed the adhesion ability of H. pylori. The HP0044 knockout complementary mutant showed partial restoration of the bacterial adhesion ability and the HP1275 knockout complementary mutant displayed full restoration of the bacterial adhesion ability compared to wild-type strain 26695, which

The HP0044-and HP1275-Disrupted Mutations Have Profound Influences on Bacterial Virulence
Studies have shown that the morphology of human gastric adenocarcinoma (AGS) cells would alter after H. pylori infection, and that the infected cells display a typical elongated and spreading appearance called the "hummingbird phenotype". To elucidate on whether the HP0044and HP1275-disrupted mutations would affect the outcomes of bacterial infection, a cell infection assay was conducted. When the AGS cells were infected by wild-type strain 26695, the classical hummingbird phenotype indeed appeared (Figure 6a). Although the hummingbird phenotype of the AGS cells could also be observed when the AGS cells were infected with the HP0044 and HP1275 knockout mutants, the percentage of the AGS cells showing the hummingbird phenotype was significantly lower than that of the cells infected by wild-type strain 26695 (20% compared to 38.5%, Figure 6b). The appearance of the hummingbird phenotype of the infected AGS cells was restored when the cells were infected by the HP1275 knockout complementary mutant, while the restoration was less significant by the infection of the HP0044 knockout complementary mutant. These results suggested that the loss of the HP0044 and HP1275 proteins could lead to a reduction in the infection ability of H. pylori, thus indicating that the intactness of H. pylori LPS is highly correlated with the virulence of H. pylori. ternalization ability, but the HP1275 knockout complementary mutant had similar internalization ability to that of the wild-type strain. The results suggested that a defect in the biosynthesis of fucose and thus the LPS structure would have a profound influence on H. pylori colonization. Various H. pylori strains were co-incubated with the AGS cells at an MOI (multiplicity of infection) of 100. The adhesion of different mutants measured by the viable plate counting method was compared to that of wild-type strain 26695 to obtain the relative percentage of bacterial adhesion. The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; * p < 0.05 and ** p < 0.01). (d) Internalization of the AGS cells by different strains of H. pylori infection. Various H. pylori strains were co-incubated with the AGS cells at an MOI of 100. The internalization of various knockout mutants measured by the viable plate counting method was statistically compared to that of wild-type strain 26695 to obtain the relative percentage of bacterial internalization. The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; ** p < 0.01). WT, wild-type strain; KO, knockout mutant; and Com, knockout complementary mutant.

HP0044 and HP1275 Gene Disruptions Affect Protein Sorting into H. pylori OMVs
Previous studies have shown that the alteration of protein sorting into OMVs is associated with the biosynthesis of LPS in some bacteria, including H. pylori [35,42]. To test whether the biosynthesis of fucose is also involved in the control of the protein-sorting mechanism in H. plori, proteins extracted from both bacteria and OMVs were collected from the tested H. pylori strains and subjected to subsequent Commassie Blue staining and (b) Quantitation of the AGS cells with the elongated phenotype after H. pylori infection. The percentage of elongated cells was quantified from 10 different 0.25 mm 2 fields. The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; * p < 0.05 and ** p < 0.01). (c) Adhesion of the AGS cells by different strains of H. pylori infection. Various H. pylori strains were co-incubated with the AGS cells at an MOI (multiplicity of infection) of 100. The adhesion of different mutants measured by the viable plate counting method was compared to that of wild-type strain 26695 to obtain the relative percentage of bacterial adhesion. The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; * p < 0.05 and ** p < 0.01). (d) Internalization of the AGS cells by different strains of H. pylori infection. Various H. pylori strains were co-incubated with the AGS cells at an MOI of 100. The internalization of various knockout mutants measured by the viable plate counting method was statistically compared to that of wild-type strain 26695 to obtain the relative percentage of bacterial internalization. The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; ** p < 0.01). WT, wild-type strain; KO, knockout mutant; and Com, knockout complementary mutant.
Previous studies have reported that LPS is associated with the adhesion ability of many Gram-negative bacteria [33][34][35]41]. However, the involvement of LPS in H. pylori colonization is under debate. Therefore, bacterial adhesion and internalization assays were conducted to investigate the role of LPS in H. pylori colonization. As shown in Figure 6c, the disruption of the HP0044 and HP1275 genes significantly suppressed the adhesion ability of H. pylori. The HP0044 knockout complementary mutant showed partial restoration of the bacterial adhesion ability and the HP1275 knockout complementary mutant displayed full restoration of the bacterial adhesion ability compared to wild-type strain 26695, which was in line with most of results of the functional and characteristic analyses conducted in this study. Consistent with the result of the adhesion assay, the relative percentage of the internalized HP0044 and HP1275 knockout mutants in the AGS cells was also reduced significantly compared to that of wild-type strain 26695 (Figure 6d). In terms of the complementary strains, the HP0044 knockout complementary mutant still possessed less internalization ability, but the HP1275 knockout complementary mutant had similar internalization ability to that of the wild-type strain. The results suggested that a defect in the biosynthesis of fucose and thus the LPS structure would have a profound influence on H. pylori colonization.

HP0044 and HP1275 Gene Disruptions Affect Protein Sorting into H. pylori OMVs
Previous studies have shown that the alteration of protein sorting into OMVs is associated with the biosynthesis of LPS in some bacteria, including H. pylori [35,42]. To test whether the biosynthesis of fucose is also involved in the control of the protein-sorting mechanism in H. plori, proteins extracted from both bacteria and OMVs were collected from the tested H. pylori strains and subjected to subsequent Commassie Blue staining and immunoblotting analyses. Figure 7 shows a dramatic difference between the protein contents from the whole-cell lysates and OMVs, indicating that only a portion of specific proteins can be sorted into or even enriched in OMVs. The protein profiles of the whole-cell lysates were similar among wild-type strain 26695, the HP0044 and HP1275 knockout mutants and the corresponding knockout complementary mutants, while a significant difference could be seen in the protein profiles of OMVs between wild-type strain 26695 and the knockout mutants. Many proteins enriched in the OMVs of wild-type strain 26695 were reduced or even completely disappeared in those obtained from the knockout mutants. Specifically, virulence factors UreA, CagA and VacA were chosen for the immunoblotting analysis. The results showed that the amount of CagA and VacA was significantly reduced in the OMVs obtained from the HP0044 and HP1275 knockout mutants, while the amount of UreA remained similar among the tested strains. The protein profiles of OMVs and the amount of CagA and VacA were not significantly recovered in the HP0044 knockout complementary mutant compared to those of the HP0044 knockout mutant, which is in accordance with the previous observation that the complementary ability of the HP0044 knockout complementary mutant is incomplete. These results suggested that certain proteins in H. plori can be specifically sorted into the OMVs, and that the HP0044 and HP1275 proteins involved in the biosynthesis of fucose of LPS play an important role in affecting the protein-sorting mechanism of H. pylori. recovered in the HP0044 knockout complementary mutant compared to those of the HP0044 knockout mutant, which is in accordance with the previous observation that the complementary ability of the HP0044 knockout complementary mutant is incomplete. These results suggested that certain proteins in H. plori can be specifically sorted into the OMVs, and that the HP0044 and HP1275 proteins involved in the biosynthesis of fucose of LPS play an important role in affecting the protein-sorting mechanism of H. pylori. The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; * p < 0.05, ** p < 0.01 and *** p < 0.001). Proteins from the whole-cell lysates and OMVs were analyzed by 15% SDS-PAGE with Coomassie blue staining and immunoblotted using anti-CagA, anti-VacA and anti-UreA antibodies. WT, wild-type strain; KO, knockout mutant; and Com, knockout complementary mutant.

Discussion
LPS plays an important role in Gram-negative bacteria, maintaining the membrane integrity, forming an impermeable barrier, influencing the host immune responses and promoting bacterial adhesion. It also serves as a key virulence factor for the long-term colonization of H. pylori in the harsh gastric environment [16,43]. The terminus of the Oantigen in H. pylori is usually decorated by fucose and forms the structure of Lewis X and The data presented were derived from triplicate trials with statistical analysis (unpaired, two-tailed Student's t-test; * p < 0.05, ** p < 0.01 and *** p < 0.001). Proteins from the whole-cell lysates and OMVs were analyzed by 15% SDS-PAGE with Coomassie blue staining and immunoblotted using anti-CagA, anti-VacA and anti-UreA antibodies. WT, wild-type strain; KO, knockout mutant; and Com, knockout complementary mutant.

Discussion
LPS plays an important role in Gram-negative bacteria, maintaining the membrane integrity, forming an impermeable barrier, influencing the host immune responses and promoting bacterial adhesion. It also serves as a key virulence factor for the long-term colonization of H. pylori in the harsh gastric environment [16,43]. The terminus of the O-antigen in H. pylori is usually decorated by fucose and forms the structure of Lewis X and Lewis Y antigens. The Lewis antigens on the H. pylori O-antigen serve as a form of molecular mimicry because similar or even identical glycan structures, including Lewis X/Y and Lewis a/b antigens, can be found on the epithelial cells in the human stomach. This similarity provides an advantageous strategy for H. pylori to evade host immune surveillance and enables this bacterium to persist in the stomach [19].
In the past, the structure of LPS in H. pylori was considered to be a branched form, and the O-antigen was attached to the inner core of LPS as a side chain (Figure 1a). According to this branched LPS structure, fucose residues only exist in the terminus of the LPS O-antigen. Recently, the LPS structure in H. pylori wild-type strains G27 and 26695 was re-examined [13]. The results showed that the core oligosaccharide is shorter than the one previously recognized, and that the O-antigen is connected to the core oligosaccharide in a linear form (Figure 1a). The most significant change in this newly proposed LPS structure is the presence of a fucose residue in the "Trio" region that directly connects to the core region of LPS. The existence of fucose in the LPS backbone potentially increases the importance of fucose in the structure and function of H. pylori LPS. Previous reports have demonstrated that GMD and PMM are necessary for the synthesis of a complete LPS structure in Pseudomonas species and Shigella flexneri [32,44,45]. In the current study, two genes, HP0044 and HP1275, which are homologous to GMD and PMM/PGM, respectively, and predicted to be associated with the biosynthesis of fucose in wild-type strain 26695, were knocked out to generate the corresponding mutants. The LPS profiles and physiological traits of these knockout mutants were also characterized, showing that a defect in the biosynthesis of fucose in the HP0044 and HP1275 knockout mutants not only causes the disappearance of the Lewis antigens on the LPS terminus but also leads to a significant loss in the entire O-antigen polysaccharide (Figure 3). In addition, the amount and position of the core oligosaccharide in the knockout mutants also show significant alterations, thus suggesting that a deficiency in the biosynthesis of fucose may impose more complex influences on the biosynthesis of LPS in H. pylori originally thought. According to the data presented, the fucose is indeed present in the backbone of the LPS O-antigen of H. pylori wild-type strain 26695, which supports the H. pylori LPS structure newly proposed by Li et al. [13].
Our group has previously demonstrated that HP0857, HP0859 and HP0860 are all essential for the biosynthesis of heptose, which is involved in the LPS core region of H. pylori [33][34][35]. Disruptions of the HP0857, HP0859 and HP0860 genes not only cause the loss of a major part of the LPS structure in H. pylori but also result in an increase in its susceptibility to the hydrophobic antibiotic novobiocin and the detergent SDS. Furthermore, these knockout mutants all show a dramatic decline in the adhesion ability to AGS cells and thus cause less "hummingbird" phenotypes of AGS cells after bacterial infection. In this study, we further demonstrated that LPS truncation caused by fucose deficiency can cause a significant decrease in bacterial membrane permeability, adherence and internalization ability and virulence, which echoes our previous findings on the biosynthesis of heptose in H. pylori about the importance of LPS in the fitness and virulence of H. pylori. It is worth noting that similar results are also found in other bacteria. Chiku et al. showed that the deficiency of GMD affects the structure of LPS, bacterial motility and cell surface hydrophobicity in Pseudomonas syringae Pathovar glycinea Race 4 [45]. Tang et al. reported that the algC knockout mutation can decrease the virulence of P. aeruginosa [46], and that the ability of the resultant mutant to cause bacteriemia and death in a neonatal mouse model of infection is significantly reduced compared to that of wild-type strain PAO1. Taken together, our findings agree with others in certain bacteria and suggest that the deficiency of GMD and PMM/PGM and thus the loss of an intact LPS structure could lead to profound defects in the fitness and virulence of H. pylori.
Bacterial biofilms are composed of communities of multiple bacterial species with various self-produced extracellular polymeric substances (EPS), including proteins, nucleic acids, carbohydrates and OMVs, which together are attached to biotic or abiotic surfaces. Research has demonstrated that H. pylori is capable of forming biofilms in response to different environmental stresses and for adaptating to the gastric niche as well as host colonization [47]. The formation of H. pylori biofilms can alleviate susceptibility to antimicrobial agents, which may lead to the failure of H. pylori eradication therapy and increase the medical burden [48]. LPS was found to be one of the major macromolecules in the H. pylori biofilm matrix [49]. Some H. pylori genes related to LPS biosynthesis and export have been shown to be associated with the transition from the planktonic to the biofilm phase of bacterial growth [50][51][52]. Both bacterial surface hydrophobicity and autoaggregation are key factors for the development of biofilm [53]. Interestingly, a recent study showed that the formation of biofilm in H. pylori is regulated by a two-component acid response system (ArsRS) [54]. Mutations in the ArsRS system not only increased the level of bacterial autoaggregation but also enhanced the formation of biofilm, thus revealing the correlation between these two phenotypes in H. pylori and the involvement of environmental pH in the modulation of the development of biofilm. In the present study, the surface hydrophobicity of H. pylori is significantly increased and the bacterium is prone to autoaggregation when the biosynthesis of fucose is out of order. Accordingly, it is appealing to presume that the mutations of HP0044 and HP1275 may have a positive effect on promoting the formation of biofilm. The fact that the HP0044 and HP1275 mutations can reduce the bacterial virulence associated with toxicity, but that they may also enhance the development of biofilm by H. pylori, has made a further study aiming at revealing the importance of the biosynthesis of fucose in the formation of biofilm in H. pylori and its clinical impact an interesting topic to explore.
The production of OMVs budding from the outer membrane of most Gram-negative bacteria can also be found in H. pylori [55,56]. Accumulating evidence has suggested that the formation of OMVs is a general strategy for Gram-negative bacteria to export certain kinds of proteins to the environment or the host cells [57]. In addition, OMVs play mutiple roles in promoting bacterial adhesion [58], invasion [59] and the formation of biofilm [60], enhancing antibiotic resistance for the whole community [61], causing damage to the host cells through toxin transport [62] and modulating the host immune responses by regulating the proteins in the host immune system [63]. Research has shown that the composition of proteins in the OMVs is different from that in the cytoplasm, periplasm and OM; however, the detailed mechanism for the process of protein enrichment or exclusion is still not clear [64,65]. Previously, Haurat et al. demonstrated that the protein-sorting process is associated with the biosynthesis of LPS in Porphyromonas gingivalis [42]. The loss of the LPS O-antigen in P. gingivalis resulted in an aberrent protein presentation in OMVs. Alterations in the protein composition in OMVs due to a loss of the O-antigen has been reported in Klebsiella pneumoniae as well [66]. Similar results were also found in our previous study, which showed that a gene-disrupted mutation affecting the biosynthesis of heptose not only caused a severe truncation of LPS but also altered the protein profile of OMVs [35]. In addition, an important virulence factor, CagA, was nearly absent from the OMVs of the derived heptose-deficient mutant [35]. In the current study, we further confirmed the relationship between the biosynthesis of LPS and the mechanism of protein sorting into OMVs. The protein profile of OMVs from the HP0044 and HP1275 knockout mutants, which displayed an O-antigen-deficient phenotype, showed significant alterations compared to that from wild-type strain 26695. Notably, not only CagA but also VacA, another key virulence factor of H. pylori, was substantially reduced in these knockout mutants. The data presented in this study are in agreement with our previous findings and imply that the integrity of LPS is associated with the selective mechanism of sorting proteins into OMVs in H. pylori.
In this study, most of the data presented indicated that the HP0044 knockout complementary mutant does not have its GMD function fully restored. Only a faintish signal of the LPS O-antigen can be observed in the HP0044 knockout complementary mutant from the silver staining of the LPS profile and the Lewis Y antigen staining (Figure 3). Similar phenomena also appear in the data obtained from the membrane property assays ( Figure 5), the bacterial infection assays ( Figure 6) and the assays for protein sorting into OMVs (Figure 7). These abnormal results may be related to the poor efficiency of HP0044 gene complementation. This inefficiency could have resulted from (1) polar effects that occurred during the homologous recombination of the HP0044 gene to the RdxA locus in the genomic DNA of the HP0044 knockout mutant, or (2) an inappropriate choice of the promoter sequence for the expression of the HP0044 gene. In contrast, the HP1275 knockout complementary mutant was shown to be comparable to wild-type strain 26695 in all of the tests. Although the complementary efficiency of the allotopic HP0044 gene was lower than expected, both the HP0044 and HP1275 knockout complementary mutants can still restore the functions of the corresponding knockout mutations to a certain degree, thus confirming that the HP0044 and HP1275 mutations indeed contribute to the alterations observed in this study.
In summary, we demonstrated that disruptions of the HP0044 and HP1275 genes could lead to a deficiency in fucose production, which would further affect the biosynthesis of LPS. Our data also supported the newly proposed LPS structure that, in addition to decorating the terminus of the LPS O-antigen to form Lewis antigen moieties, posited that fucose is also involved in a "Trio" region of the O-antigen, connected to the core oligosaccharide in H. plyori. A defect in the production of fucose caused a loss of a major part of LPS in H. pylori, and thus resulted in a slight retardation in bacterial growth and higher susceptibility to the detergent SDS and the hydrophobic antibiotic novobiocin. In addition, a truncated LPS without the O-antigen increased the surface hydrophobicity of H. pylori and promoted bacterial autoaggregation. Furthermore, fucose deficiency also significantly affected bacterial virulence and protein sorting into OMVs of H. pylori. The findings aquired demonstrated that GMD and PMM/PGM in H. pylori are potentially important for bacterial fitness and pathogenesis. Therefore, these enzymes can be candidates for the development of drugs against H. pylori infection, and further researche on this aspect is worth conducting.

Data Availability Statement:
The data presented in this study are available within the article.