Peptidoglycan fragment release and NOD activation by commensal Neisseria species from humans and other animals

Abstract

Neisseria gonorrhoeae, a human restricted pathogen, releases inflammatory peptidoglycan (PG) fragments that contribute to the pathophysiology of pelvic inflammatory disease. The genus Neisseria is also home to multiple species of human- or animal-associated Neisseria that form part of the normal microbiota. Here we characterized PG release from the human-associated nonpathogenic species Neisseria lactamica and Neisseria mucosa and animal-associated Neisseria from macaques and wild mice. An N. mucosa strain and an N. lactamica strain were found to release limited amounts of the proinflammatory monomeric PG fragments. However, a single amino acid difference in the PG fragment permease AmpG resulted in increased PG fragment release in a second N. lactamica strain examined. Neisseria isolated from macaques also showed substantial release of PG monomers. The mouse colonizer Neisseria musculi exhibited PG fragment release similar to that seen in N. gonorrhoeae with PG monomers being the predominant fragments released. All the human-associated species were able to stimulate NOD1 and NOD2 responses. N. musculi was a poor inducer of mouse NOD1, but ldcA mutation increased this response. The ability to genetically manipulate N. musculi and examine effects of different PG fragments or differing amounts of PG fragments during mouse colonization will lead to a better understanding of the roles of PG in Neisseria infections. Overall, we found that only some nonpathogenic Neisseria have diminished release of proinflammatory PG fragments, and there are differences even within a species as to types and amounts of PG fragments released.

Keywords: NOD1; NOD2; Neisseria; ampG; immune response; inflammation; ldcA; peptidoglycan.

Fig 1

(A) Peptidoglycan (PG) fragments released into the milieu by Neisseria spp. (B) Peptidoglycan fragments that stimulate the human or mouse pattern recognition receptor NOD1 or that stimulate NOD2 in both species. Cartoon depictions of the PG fragments found in gonococcal supernatants use symbols from Jacobs et al. (44).

Fig 2

PG fragment release from nonpathogenic, human-associated species Neisseria mucosa strain ATCC 25996. The PG fragments released were separated by size-exclusion chromatography. (A) PG fragments released during the chase period following metabolic labeling of PG with [³H]-glucosamine (glcNH₂). (B) PG fragments released from N. mucosa following [³H]-DAP labeling compared quantitatively to those released by N. gonorrhoeae.

Fig 3

PG fragments released by Neisseria lactamica strain ATCC 23970 (A) or N. lactamica strain ATCC 49142 (B) following metabolic labeling with [³H]-DAP. ATCC 23970 breaks down more of its PG fragments before release, resulting in the most abundant fragments being tetrapeptides and tripeptides, similar to the results seen with N. mucosa. ATCC 49142 releases substantial amounts of PG monomers as well as free tetrapeptides and tripeptides.

Fig 4

Function of N. lactamica ampG alleles assessed by metabolic labeling of PG with [6-³H]-glucosamine and size-exclusion chromatography separation of PG fragments released into the medium. Gonococcal ampG⁻ strains expressing ampG from two different strains of N. lactamica released different amounts of PG monomer. The ampG gene from N. lactamica ATCC 49142 or N. lactamica ATCC 23970 was expressed in N. gonorrhoeae in lieu of the native gonococcal ampG gene. Expression of ampG_NL23970 but not ampG_NL49142 in N. gonorrhoeae reduced the amount of PG monomer released.

Fig 5

PG fragment release from N. gonorrhoeae expressing mutant versions of N. lactamica ampG alleles. Polymorphism of N. lactamica AmpG contributes to the differences in the amounts of PG monomer released. (A) N. gonorrhoeae lacking gonococcal ampG and instead expressing ampG_NL49142 with a single substitution at AmpG_NL49142 residue 272 from a threonine to an alanine shows reduced PG monomer release by approximately half. (B) N. gonorrhoeae expressing ampG_NL23970 with a single substitution at AmpG_NL23970 residue 272 from an alanine to a threonine released N. gonorrhoeae WT-like levels of PG monomer.

Fig 6

PG fragment release from Neisseria spp. that colonize macaques. Macaque isolates AP312 (A) and AP678 (B) released different types and amounts of PG fragments as demonstrated by size-exclusion chromatography of [6-³H]-glucosamine labeled molecules released into the medium.

Fig 7

N. musculi releases a variety of PG fragments as determined from experiments using metabolic labeling with [³H]-glucosamine (A) or [³H]-DAP (B). The types of PG fragments released and their relative abundance are similar to that seen with N. gonorrhoeae in that the PG monomers are the most abundant fragments, and PG dimers, free disaccharide, and free peptides are all observed.

Fig 8

Mutation of ampG impaired PG fragment recycling in N. musculi and macaque symbiont AP312. Mutation of ampG in the macaque isolate AP312 (A) and N. musculi (B) resulted in large increases in the amount of PG monomer released, without affecting PG dimer release.

Fig 9

hNOD1 and hNOD2 activation by supernatants from different Neisseria spp. Supernatants from N. mucosa, N. lactamica (A and B), and ampG mutants of N. mucosa (C and D) induced hNOD1 (A and C) and hNOD2 (B and D) activation in HEK293 reporter cells overexpressing NOD1 or NOD2. Supernatant from an ampG mutant of N. mucosa induced a larger hNOD1 response compared to wild type. Supernatant from N. lactamica ATCC23970 induced the largest hNOD2 response of all supernatant treatment samples. Statistical significance was determined using Student’s t-test. *P < 0.05.

Fig 10

mNOD1 activation by supernatants from N. musculi used to treat HEK-293 cells overexpressing mNOD1. Supernatants from N. musculi WT, ampG or ldcA mutants, or a strain overexpressing ldcA were used. Statistical significance was determined using Student’s t-test. *P < 0.05.