Microbial Mannan Inhibits Bacterial Killing by Macrophages: A Possible Pathogenic Mechanism for Crohn’s Disease

Inhibitory effect of yeast mannan on bacterial killing by normal adherent human monocytes. (A) Dose response of mannan inhibition of killing of Crohn’s disease mucosa-associated E coli isolate HM605 by adherent monocytes at 2 hours (n = 9; P = .0002, Cuzick test for linear trend across the dose response). (B) Mannan, at 1 mg/mL, inhibits killing of control E coli ATCC 25922 and Crohn’s disease mucosa-associated E coli HM605 (n = 10) by normal human adherent monocytes. E coli HM605 is shown to be more resistant to killing than E coli ATCC 25922 (P = .03). Mannan had a similar effect on the killing of S aureus by adherent monocytes. Data are presented as mean ± SEM. (C) Transmission electron micrographs showing internalization and apparent replication of Crohn’s disease mucosal E coli isolate HM605 within vacuoles in J774 murine macrophages.

Inhibition by mannan (1 mg/mL) of the killing of Crohn’s disease E coli isolate HM605 (n = 6; ***P < .001, unpaired t test) but not S aureus (n = 6; P = .18) by human monocyte-derived macrophages (MDM) at 2 hours. Data are presented as mean ± SEM.

Lack of effect of mannan (1 mg/mL) on bacterial survival in suspended mononuclear cells by 2 hours. (n = 8; S aureus, P = .358; and E coli HM605, P = .977, unpaired t test). Data are presented as mean ± SEM.

Effect of mannan (1 mg/mL) on killing of E coli in bone marrow-derived MyD88^−/−and TLR4^−/− murine macrophages. Mannan is shown to inhibit the killing of both E coli ATCC 25922 (P = .0005, ANOVA) and E coli HM605 (P = .0006) by wild-type (C57B1/6) macrophages but to have no effect on the already impaired killing of both E coli strains by TLR4^−/− and MyD88^−/− macrophages. Data are presented as mean ± SEM (n = 3).

(A) Mannan dose dependently inhibits the peak respiratory burst in PMA prestimulated monocytes (n = 10; amplified by HRP and detected with isoluminol, P = .0023, Cuzick’s test for linear trend across the dose response (B) Time course of inhibition of peak respiratory burst in PMA prestimulated monocytes by mannan (1 mg/mL), amplified by HRP and detected with isoluminol. Representative example performed with 10 replicates at each sample time point. (C) Mannan inhibits peak respiratory burst in formyl-Met-Leu-Phe prestimulated neutrophils in a dose-dependent manner (n = 4 or 5; P = .0002, Cuzick’s test for linear trend across the dose response).

Slot blotting probed with biotin-conjugated Snowdrop (Galanthus nivalis) lectin (GNA) recognizes terminal Manα1–3Man-expressing glycoconjugates on M paratuberculosis, M avium, M kansasii, M bovis, S cerevisiae, and C albicans. Crohn’s disease mucosa-associated E coli HM427, M tuberculosis, B fragilis, E faecalis, S bovis, and L monocytogenes were all negative. Optical density (OD; 600 nm) was used to measure bacterial cell number prior to loading of 5 × 10⁶ bacteria or 1 × 10⁶ yeasts per slot. S cerevisiae mannan (200 μg) was used as a positive control and sterile PBS as a blank (background) control. Blots shown are representative of 3 replicates.

(A) Effect of supernatant from Mycobacterium paratuberculosis on survival of Crohn’s disease mucosal E coli isolate HM605 within adherent monocytes derived from normal donor peripheral blood mononuclear cells. (B) Inhibitory effect of Mycobacterium paratuberculosis culture medium on E coli killing by J774-A1 macrophages disappears after GNA lectin affinity removal of Manα1–3Man containing-glycoconjugates.