The Lst Sialyltransferase of Neisseria gonorrhoeae Can Transfer Keto-Deoxyoctanoate as the Terminal Sugar of Lipooligosaccharide: a Glyco-Achilles Heel That Provides a New Strategy for Vaccines to Prevent Gonorrhea

Abstract

The lipooligosaccharide (LOS) of Neisseria gonorrhoeae plays key roles in pathogenesis and is composed of multiple possible glycoforms. These glycoforms are generated by the process of phase variation and by differences in the glycosyltransferase gene content of particular strains. LOS glycoforms of N. gonorrhoeae can be terminated with an N-acetylneuraminic acid (Neu5Ac), which imparts resistance to the bactericidal activity of serum. However, N. gonorrhoeae cannot synthesize the CMP-Neu5Ac required for LOS biosynthesis and must acquire it from the host. In contrast, Neisseria meningitidis can synthesize endogenous CMP-Neu5Ac, the donor molecule for Neu5Ac, which is a component of some meningococcal capsule structures. Both species have an almost identical LOS sialyltransferase, Lst, that transfers Neu5Ac from CMP-Neu5Ac to the terminus of LOS. Lst is homologous to the LsgB sialyltransferase of nontypeable Haemophilus influenzae (NTHi). Studies in NTHi have demonstrated that LsgB can transfer keto-deoxyoctanoate (KDO) from CMP-KDO to the terminus of LOS in place of Neu5Ac. Here, we show that Lst can also transfer KDO to LOS in place of Neu5Ac in both N. gonorrhoeae and N. meningitidis Consistent with access to the pool of CMP-KDO in the cytoplasm, we present data indicating that Lst is localized in the cytoplasm. Lst has previously been reported to be localized on the outer membrane. We also demonstrate that KDO is expressed as a terminal LOS structure in vivo in samples from infected women and further show that the anti-KDO monoclonal antibody 6E4 can mediate opsonophagocytic killing of N. gonorrhoeae Taken together, these studies indicate that KDO expressed on gonococcal LOS represents a new antigen for the development of vaccines against gonorrhea.IMPORTANCE The emergence of multidrug-resistant N. gonorrhoeae strains that are resistant to available antimicrobials is a current health emergency, and no vaccine is available to prevent gonococcal infection. Lipooligosaccharide (LOS) is one of the major virulence factors of N. gonorrhoeae The sialic acid N-acetylneuraminic acid (Neu5Ac) is present as the terminal glycan on LOS in N. gonorrhoeae In this study, we made an unexpected discovery that KDO can be incorporated as the terminal glycan on LOS of N. gonorrhoeae by the alpha-2,3-sialyltransferase Lst. We showed that N. gonorrhoeae express KDO on LOS in vivo and that the KDO-specific monoclonal antibody 6E4 can direct opsonophagocytic killing of N. gonorrhoeae These data support further development of KDO-LOS structures as vaccine antigens for the prevention of infection by N. gonorrhoeae.

Keywords: Lst sialyltransferase; N-acetylneuraminic acid (Neu5Ac); Neisseria gonorrhea; gonococcal vaccine; keto-deoxyoctanoate (KDO); lipooligosaccharide (LOS); sialic acid.

FIG 1

(A) Structure of the KDO-containing LOS of NTHi and N. gonorrhoeae. MAb 6E4 only recognizes the terminal KDO glycan. Lst and LsgB are the sialyltransferases that can also incorporate KDO onto the terminal Gal in the absence of Neu5Ac. LgtB is the galactosyltransferase that transfers the KDO acceptor galactose (Gal) to N-acetylglucosamine (GlcNAc). (B) Silver staining and 6E4 Western blot of purified LOS samples from MC58, 1291 wild-type, and glycotransferase mutant strains.

FIG 2

Whole-cell ELISA of 6E4. (A) ELISA of whole-cell N. gonorrhoeae WT and lst mutant strains using 6E4 MAb. Lst transfers the terminal sialic acid to the terminal Gal of LOS in N. meningitidis. Here, the lst mutant showed Lst also transfers KDO to the terminal Gal in N. gonorrhoeae. (B) Whole-cell ELISA of 6E4 was performed to test 20 clinical N. gonorrhoeae isolates. 1291 and 1291Lst were the positive and negative controls in this ELISA. (C) ELISA of whole-cell Neisseria meningitidis MC58 WT, SiaB, and lst mutant strains using 6E4 MAb. MC58 WT has Neu5Ac as the terminal glycan of LOS. Without SiaB, N. meningitidis cannot make CMP-Neu5Ac; therefore, Lst adds KDO at the terminal of LOS instead of sialic acid. (D) Incorporation of KDO on gonococcal LOS. Whole-cell ELISA of 6E4 showed that the presence of CMP-Neu5Ac in the growing medium could inhibit the incorporation of KDO into the terminal Gal. N. gonorrhoeae 1291 and lst mutant were grown in 0, 30 mM, and 100 μg/ml CMP-Neu5Ac supplemented media.

FIG 3

Cellular localization of Lst. (A) Western blot analysis of anti-Lst polyclonal antisera binding to Lst in whole-cell lysates. Coomassie staining gel showed equal loading of cell lysates. (B) Strain 1291 wild type and lst mutant were blotted on the membrane. Cells were treated with or without 10% SDS. Anti-Lst could detect Lst only when the cell was lysed. (C) Western blot analysis of whole, intact N. gonorrhoeae 1291 and N. meningitidis MC58 cells treated with increasing concentrations of trypsin (0, 10, and 20 μg) and probing with anti-Lst serum. Published neisserial MsrAB protein and the meningococcal PorA protein are the controls of surface protein.

FIG 4

(A) Confocal microscopy study of cervical swab from a patient with a documented N. gonorrhoeae infection stained with both anti-KDO MAb 6E4 (Texas red) and anti-LOS MAb 6B4 (FITC). Infected PMNs can be seen with gonococci stained with both antibodies separately (red or green) and with antibodies colocalized to the same organisms (yellow). This study indicates that KDO termination occurs in vivo as well as in vitro. (B) Opsonophagocytic activity of MAb 6E4 against N. gonorrhoeae strain 1291. The survival of N. gonorrhoeae in the presence of 0 to 10 μg/ml MAb 6E4, primary human PMNs, and 10% normal human serum as a complement source is shown. Data represent the average survival (±1 standard deviation) for triplicate samples, shown as a percentage of bacteria in the absence of antibody (0 μg/ml MAb 6E4, set at 100%, represents 5.07 × 10³ CFU). There was a statistically significant difference between groups, as determined by one-way ANOVA [F(6,14) = 59.20, P < 0.0001]. Statistically significant differences relative to the untreated wild type, using a two-tailed Student's t test, are indicated: *, P < 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. (P values were the following: 1.25 μg/ml, 0.012; 2.5 μg/ml, 0.005; 5 μg/ml, 0.001; and 10 μg/ml, 0.0003.) (C) Serum bactericidal activity. The survival of N. gonorrhoeae in the presence of 0 to 10 μg/ml MAb 6E4 and 10% normal human serum is shown. Data represent the mean survival (±1 standard deviation) for triplicate samples as a percentage of bacteria in the absence of antibody (0 μg MAb 6E4, set at 100%, represents 4.53 × 10³ CFU). There was no statistically significant difference between group means, as determined by one-way ANOVA [F(6,14) = 0.71, P = 0.65]. There was no statistically significant difference for any group relative to the untreated wild type, as determine using a two-tailed Student's t test (P ≥ 0.12).

FIG 5

Schematic diagram of proposed biosynthesis pathway of terminal KDO on LOS in N. gonorrhoeae. (A) Inside cells, CMP-KDO is synthesized by WaaA and transferred to the terminal glycan of LOS by Lst in the absence of CMP-Neu5Ac. (B) In the presence of CMP-Neu5Ac, CMP-Neu5Ac can be taken up in the intracellular environment by an unknown mechanism and competes with CMP-KDO for addition to LOS by Lst inside cells.