Fig 1.
Structures of the CMP-nonulosonate (CMP-NulO) analogs used in this study.
A General chemical structure, and B 2C5 chair chemical structure of CMP-nonulosonates used. All NulO sugars except for pseudaminic acid (Pse5Ac7Ac) have the same stereochemistry (D-glycero-D-galacto configuration), where Pse has an L-glycero-L-manno configuration, differing stereochemically at carbons 5, 7 and 8. For reference, the nine carbon atoms of the NulOs are numbered in A, where the NulO exocyclic moiety is highlighted in red.
Table 1.
Summary of nonulosonate (NulO) incorporation by N. gonorrhoeae lipooligosaccharide and key functional consequences.
Fig 2.
Substrate specificity of gonococcal LOS sialyltransferase (Lst).
A. N. gonorrhoeae (Ng) strain F62 ΔlgtD was grown in gonococcal media alone (‘No sialic acid’) or in media supplemented with CMP salts of the indicated sialic acid (Sia), each at a concentration of 20 μg/ml (30 μM), for 2 h at 37°C. Bacteria were washed, pellets treated with protease K, lysed in 4× LDS buffer and lysates were separated on 12% Bis-Tris gels using MES running buffer. LOS was transferred to a PVDF membrane and probed with monoclonal antibody (mAb) 3F11 that recognizes the lacto-N-neotetraose only in the non-sialylated state; the addition of a Sia residue abrogates mAb 3F11 binding (upper panel). LOS was also visualized by silver staining following electrophoresis on 16.5% tricine gels (lower panel). B. Incorporation of NulO by LNnT LOS was assessed using whole cell ELISA with mAb 3F11, *** P<0.001. C. Wild-type Ng F62 was grown with (shaded) or without (solid line) 30 μM CMP-Neu5Gc for 3 hours and analyzed by flow cytometry using a polyclonal Neu5Gc-specific chicken IgY Ab followed by a FITC conjugated donkey anti-chicken IgY secondary Ab. As a negative control, wild-type Ng F62 was grown in CMP-Neu5Gc as above and incubated with the secondary Ab only (dashed line).
Fig 3.
Select sialic acid (Sia) analogs enhance gonococcal serum resistance.
N. gonorrhoeae (Ng) F62 ΔlgtD was grown in media alone (no CMP-Sia added), or media that contained 20 μg/ml (~30 μM) of each of the indicated CMP-Sias (NulOs). Resistance of bacteria to complement-dependent killing in the presence of 3.3%, 6.7% or 10% normal human serum (NHS) was measured in serum bactericidal assays. The mean (SD) of two independent experiments is shown. A survival greater than 100% indicates growth of bacteria.
Fig 4.
Factor H (FH) binding when sialic acid (Sia) analogs (NulOs) were substituted into N. gonorrhoeae (Ng) lipooligosaccharide (LOS).
A. Ng F62 ΔlgtD was grown for 2 h in media alone, or media supplemented with 20 μg/ml (~30 μM) of each of the indicated CMP-sialic acids (NulOs). Binding of bacteria to FH (10 μg/ml) was measured by flow cytometry using anti-FH mAb 90X. The control (open bar) indicates reaction mixtures that lacked FH. The median fluorescence was recorded for each reaction and the Y-axis shows the mean fluorescence (SD) of FH binding of 2 independent experiments. The mean (SD) of 2 independent repeat experiments is shown. B. FH binding to Ng F62 ΔlgtD grown in CMP-Neu5Ac and CMP-Neu5Gc at concentrations ranging from 0.5 μg/ml to 4 μg/ml. Bacteria were incubated with FH (1 μg/ml) and bound FH was detected as indicated above.
Fig 5.
IgG, IgM, and complement components C3, C4 and FB deposition on N. gonorrhoeae (Ng) F62 that have incorporated each indicated sialic acid analog into lipooligosaccharide (LOS).
Ng F62 ΔlgtD was grown in media alone (open bars), or media supplemented with 20 μg/ml (~30 μM) of each of the indicated CMP-Sias (NulOs). Bacteria were incubated with either 3.3% or 10% normal human serum (NHS) at 37°C for 10 min and IgG and IgM binding, and C3, C4 and factor B (FB) deposited on bacteria were measured by whole cell ELISA. Data with 3.3% and 10% NHS is shown using hatched and solid bars, respectively. mAb 2-8C-4-1 [62] that detects the Neisserial lipoprotein H.8 was similar across all groups and confirmed similar capture of bacteria in all wells (bottom right graph). Measurement of gonococcal H.8 lipoprotein antigen was performed to assess similarity of bacterial capture across microtiter wells. The mean (SD) of three observations is shown. Significance of differences in Ig or complement component binding/deposition using each analog vs, Neu5Ac is indicated for results that used 10% NHS only (for simplicity): *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001 (ANOVA). Controls with bacteria alone yielded OD405nm readings <0.1 (negative).
Fig 6.
Survival of N. gonorrhoeae (Ng) first incubated in CMP-Neu5Ac or CMP-Neu5Gc followed by incubation with CMP-NulOs.
Ng F62 ΔlgtD (A and B) and ceftriaxone-resistant (CRO-R) Ng H041 (C) were suspended in media containing 20 μg/ml CMP-Neu5Ac or CMP-Neu5Gc as indicated for 15 min at 37°C. Following incubation, CMP-Neu5Ac9Az or CMP-Leg5Ac7Ac (at concentrations of 20, 2 or 0.2 μg/ml) were added to bacterial suspensions and then incubated for 2 h. Bacteria were serially diluted and serum bactericidal assays were performed using 2000 CFU and 10% normal human serum (NHS). Controls included bacteria in media that lacked any CMP-Sia altogether or bacteria that were incubated in CMP-Neu5Ac or CMP-Neu5Gc only. Y-axis, % survival (mean (SD) of two independent experiments is shown). ****, P<0.0001 compared to all other reactions (ANOVA).
Fig 7.
CMP-Neu5Ac9Az and CMP-Leg5Ac7Ac interfere with inhibition of the classical and alternative pathways of complement mediated by CMP-Neu5Ac.
N. gonorrhoeae (Ng) F62 ΔlgtD was incubated with CMP-Neu5Ac for 15 min followed by addition of CMP-Neu5Ac9Az or CMP-Leg5Ac7Ac (at concentrations of 20, 2 or 0.2 μg/ml) for 2 h as described in Fig 6. Bacteria were incubated in 10% normal human serum (NHS) and IgG binding and complement C3 and C4 fragment deposition was measured by ELISA as described in Fig 5. Mean (±SD) of two independent experiments is shown. Controls with heat-inactivated serum were carried out as described in Fig 5. MAb 2-8C-4-1 that detects the Neisserial lipoprotein H.8 was similar across all groups and confirmed similar capture of bacteria in all wells (OD405nm values ranged from 0.964 to 1.039). ***, P<0.001; ****, P<0.0001 compared to other bars in the same graph (ANOVA).
Fig 8.
CMP-Neu5Ac9Az and CMP-Leg5Ac7Ac block CMP-Neu5Ac-mediated alternative pathway inhibition despite continued factor H (FH) binding.
N. gonorrhoeae (Ng) F62 ΔlgtD was incubated in media that contained CMP-Neu5Ac (20 μg/ml), followed by addition of either CMP-Neu5Ac9Az or CMP-Leg5Ac7Ac (20, 2 or 0.2 μg/ml). Bacteria were grown for 2 h and factor B (FB) binding (A) to bacteria following incubation with 10% normal human serum (NHS) was measured by whole cell ELISA as described in Fig 7. Binding to purified FH (B) was measured by flow cytometry using polyclonal goat anti-human FH. The Y-axis represents OD405nm for FB binding and median fluorescence of FH binding. Mean (SD) of two independent observations is shown for each measurement. ****, P<0.0001 (ANOVA) compared to other reactions. C. Alternative pathway activation is required for maximal killing of Ng F62 ΔlgtD grown in CMP-Leg5Ac7Ac-containing media. Bacteria were grown in media alone, or media containing CMP-Neu5Ac (20 μg/ml) or CMP-Leg5Ac7Ac (20 μg/ml) or both CMP-NulOs together (each at 20 μg/ml) as indicated on the X-axis. Serum bactericidal assays were performed using normal human serum (all complement pathways intact; black bars) or NHS treated with an anti-factor Bb mAb that blocks FB function (AP inactivated, but classical and lectin pathways intact; grey bars). Y-axis, percent survival. The mean (range) of two independent experiments is shown.
Fig 9.
CMP-Leg5Ac7Ac treatment reduces the duration of infection and the burden of N. gonorrhoeae (Ng) F62 in the BALB/c mouse vaginal colonization model.
Three groups of 17β-estradiol treated BALB/c mice (n = 10/group) were inoculated and treated as follows: i) wild-type F62; saline (vehicle control), ii) wild-type Ng F62; treated with CMP-Leg5Ac7Ac 10 μg/ml intravaginally daily and iii) Ng F62 Δlst; saline. Vaginal swabs were obtained daily to quantify Ng CFUs. A. Kaplan Meier analysis of time to clearance. B. Colonization of bacteria (log10 CFU) measured daily. C. Bacterial burdens consolidated over time (Area Under the Curve [log 10 CFU] analysis) for the three groups. There were no statistically significant differences noted between the group treated with CMP-Leg5Ac7Ac and the Ng F62 Δlst group. D. Complement resistance of Ng recovered directly from the genital tract of mice simulates findings in vitro. Equal volumes of vaginal secretions from all mice within a group on each day were pooled, serially diluted and incubated with 10% NHS. Survival of bacteria at 30 min relative to 0 min was measured and the percentage survival expressed on the Y-axis.
Fig 10.
CMP-Leg5Ac7Ac treatment reduces the duration and burden of ceftriaxone-resistant (CRO-R) N. gonorrhoeae (Ng) H041 in the BALB/c mouse vaginal colonization model.
Two groups of 17β-estradiol treated BALB/c mice (n = 10/group) were challenged with 9 x 105 CFU Ng H041. One group received CMP-Leg5Ac7Ac 10 μg/ml intravaginally daily; the other group received saline (vehicle control). Vaginal swabs were obtained daily to quantify Ng CFUs. A. Kaplan Meier analysis of time to clearance. B. Colonization of bacteria (log10 CFU) measured daily. C. Bacterial burdens consolidated over time (Area Under the Curve [log 10 CFU] analysis) for the two groups.