NF-κB reporter assays.

HEK-293 cells overexpressing the human versions of NOD1, NOD2, TLR4, and TLR9 and carrying a secreted embryonic alkaline phosphatase (SEAP) reporter (HEK-Blue cells; InvivoGen) were used to measure NF-κB and AP-1 activation essentially as described previously for hNOD1 (58, 61). In accordance with the manufacturer’s instructions, 180-μl volumes of hNOD1 cells were seeded for assays in 96-well plates at 2.8 × 10⁵ cells/ml, hNOD2 and hTLR4 cells were seeded at 1.4 × 10⁵ cells/ml, and TLR9 cells were seeded at 4.5 × 10⁵ cells/ml. Cells were treated in triplicate wells with either 20-μl volumes of bacterial supernatant or appropriate controls as indicated, including 10 μg/ml TriDAP (hNOD1 agonist), 10 μg/ml muramyl dipeptide (MDP; hNOD2 agonist), 10 μg/ml LPS-EK (lipopolysaccharide from E. coli K-12; hTLR4 agonist), 50 μg/ml ODN2006 (hTRL9 agonist), or cGCBL media as a blank. Treatments of HEK-Blue hNOD1 and hTLR9 cells were done in parallel with the parental line HEK-Blue Null1-k cells, and treatments of HEK-Blue hNOD2 and hTLR4 cells were done in parallel with the parental line HEK-Blue Null2 cells. Parental cell lines contained the SEAP reporter but did not overexpress pathogen-associated molecular pattern (PAMP) receptors in order to allow subtraction of background NF-κB activation. For treatment of HEK-Blue cells, three independent cultures of nonpiliated N. gonorrhoeae MS11 (WT), ΔldcA (JL539), ΔldcA + ldcA (JL540), ldcA_S165A (JL507), ldcA_S165A + ldcA (JL508), ldcA_Δ4−75 (KH619), and ldcA_Δ4−75 + ldcA (KH620) strains were grown in liquid culture (cGCBL) on three separate days, with 1 mM IPTG added to cultures where appropriate. For each growth step, culture supernatants were prepared by removing whole cells first by low-speed centrifugation (3,000 × g, 5 min) and then by passing supernatants through a 0.22-μm-pore-size syringe filter. Growth of each strain was determined by total protein accumulation in cells, measured by bicinchoninic acid (BCA) assay (Pierce), and supernatants were normalized to cellular growth using blank growth media (cGCBL) prior to treatment. Each of the three independent growths of the strains described above was tested on three independent passages of the HEK-Blue cells, and averages (± standard deviations) of data from each experiment were combined to produce an overall average (± standard error).

HCT116-Dual human colorectal carcinoma cells (InvivoGen) were also used to measure NF-κB and AP-1 activation via the use of a SEAP reporter identical to that used with the HEK-Blue cells. HCT116-Dual cells express multiple pattern recognition receptors, including NOD1, NOD2, TLR3, TLR5, and RIG-I but not TLR2 or TLR4. In accordance with the manufacturer’s instructions, HCT116-Dual cells were washed with PBS, detached with trypsin, suspended in fresh media, and adjusted to 2.8 × 10⁵ cells/ml, prior to seeding 180-μl volumes in wells of a 96-well plate. N. gonorrhoeae strains MS11, JL539, JL540, JL507, and JL508 were grown in cGCBL medium–1 mM IPTG for 3 h. One OD₆₀₀ equivalent of bacteria was then harvested from each culture, pelleted for 30 s at 10,000 × g, washed once with tissue culture media, pelleted, and suspended again in 1 ml of tissue culture media. Triplicate wells of HCT116 cells were infected with 20 μl of prepared cell suspension from each strain or with 20 μl of a 1:10 or 1:100 dilution of each inoculum. Triplicate control wells were treated with 100 μg/ml Tri-DAP, 100 μg/ml MDP, 100 μg/ml LPS-EK, 100 ng/ml tumor necrosis alpha (TNF-α), 100 ng/ml interleukin-1β (IL-1β), or cGCBL media (blank). Inocula were serially diluted and plated for CFU determination on GCB agar.

All cells were maintained at 37°C in a 5% CO₂ atmosphere in Dulbecco’s modified Eagle’s medium (DMEM) with 4.5 g/liter glucose and 2 mM l-glutamine, supplemented with 10% fetal bovine serum (FBS) and 100 μg/ml Normocin. When appropriate for selection (according to manufacturer’s instructions), the medium was supplemented with 100 μg/ml zeocin (all cells) and with 30 μg/ml (hNOD1, hNOD2, hTLR4, and hTLR9) or 10 μg/ml (HCT116-Dual) blasticidin. At the start of each assay, the medium described above was exchanged for DMEM with 4.5 g/liter glucose and 2 mM l-glutamine containing 10% heat-inactivated fetal bovine serum (HI-FBS) and lacking the antibiotics described above (test medium). For all of the above-described experiments, the treated/infected cells were incubated for the times suggested by the manufacturer (6 to 24 h, depending on the cell line), after which 20 μl (30 µl for HCT116-Dual) was removed from each well to a new 96-well assay plate and mixed with 180 μl (170 µl for HCT116-Dual) of QUANTI-Blue medium (InvivoGen) for the detection of alkaline phosphatase. After a 1-h incubation, absorbance was read at 650 nm in a plate reader. For HEK-Blue cells, the A₆₅₀ value was subtracted from the background measurement of the matching Null1 cell line. With the resulting A₆₅₀ value, data (± standard deviations) from triplicate technical replicates of control wells were averaged and the technical replicate averages (± standard errors of the means) of the data from the three biological replicates (for the supernatant from each strain) were averaged. Data were graphed, and statistical analysis was performed in GraphPad Prism 4.0c.

LdcA active site mutations and signal sequence truncations eliminate release of human NOD1 agonist by GC. (A to D) HEK-293 cells expressing hNOD1 (A), hNOD2 (B), TLR4 (C), or TLR9 (D) and transfected with a secreted alkaline phosphatase (SEAP) reporter for NF-κB activity (HEK-Blue cells) were exposed to cell-free supernatant from growth of strains with ldcA_S165A or ldcA_Δ4−75 or complemented mutants (ldcA_S165A + ldcA⁺, ldcA_Δ4−75 + ldcA⁺). Samples were assayed simultaneously with the samples and controls represented in Fig. 6, using the same measurement protocol. Values represent means ± standard errors of averages of results from three independent experiments. (E) HCT116 cells transfected with a secreted alkaline phosphatase (SEAP) reporter for NF-κB activity (HCT116-Dual cells) were infected with ~1 × 10⁷ CFU/ml of pilus⁺ Opa⁻ versions of the following strains: the wild-type strain (MS11), mutant ΔldcA, mutant ΔldcA + ldcA⁺, mutant ldcA_S165A, and mutant ldcA_S165A + ldcA⁺ (representing a 10-fold increase over the level shown in Fig. 6E). Samples were assayed simultaneously with samples and controls in Fig. 6E, using the same measurement protocol. Values displayed are the means ± standard deviations of results from triplicate wells determined in one experiment representative of three independent experiments. (A to E) Significance was determined for bracketed comparisons by unpaired t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; n.s., not significant. Download FIG S5, EPS file, 1.1 MB.