Aeromonas salmonicida binds α2-6 linked sialic acid, which is absent among the glycosphingolipid repertoires from skin, gill, stomach, pyloric caecum, and intestine

Abstract

Carbohydrates can both protect against infection and act as targets promoting infection. Mucins are major components of the slimy mucus layer covering the fish epithelia. Mucins can act as decoys for intimate pathogen interaction with the host afforded by binding to glycosphingolipids in the host cell membrane. We isolated and characterized glycosphingolipids from Atlantic salmon skin, gill, stomach, pyloric caeca, and intestine. We characterized the glycosphingolipids using liquid chromatography - mass spectrometry and tandem mass spectrometry and the glycan repertoire was compared with the glycan repertoire of mucins from the same epithelia. We also investigated Aeromonas salmonicida binding using chromatogram and microtiter well based binding assays. We identified 29 glycosphingolipids. All detected acid glycans were of the ganglio-series (unless shorter) and showed a high degree of polysialylation. The non-acid glycans were mostly composed of the neolacto, globo, and ganglio core structures. The glycosphingolipid repertoire differed between epithelia and the proportion of the terminal moieties of the glycosphingolipids did not reflect the terminal moieties on the mucins from the same epithelia. A. salmonicida did not bind the Atlantic salmon glycosphingolipids. Instead, we identified that A. salmonicida binding to sialic acid occurred to α2-6 Neu5Ac but not to α2-3 Neu5Ac. α2-6 Neu5Ac was present on mucins whereas mainly α2-3 Neu5Ac was found on the glycosphingolipids, explaining the difference in A. salmonicida binding ability between these host glycoconjugates. A. salmonicida´s ability to bind to Atlantic salmon mucins, but not the glycosphingolipids, is likely part of the host defence against this pathogen.

Keywords: Aeromonas salmonicida; Atlantic salmon; Glycosphingolipid; bacteria; epithelial surface; glycan; host–pathogen interaction; mucin; mucosa.

Figure 1.

Summary of neutral GSL found in Atlantic salmon epithelia. Symbol explanation: yellow square=N-acetylgalactosamine (GalNac), yellow circle=galactose (Gal), blue square=N-acetylglucosamine (GlcNac), blue circle=glucose (Glc), white square=N-acetylhexosamine (HexNac) and red triangle=fucose (Fuc).

Figure 2.

MS/MS interpretation of two fucosylated structures from Atlantic salmon gills. A) m/z 852 suggest a fucosylated globo structure. B) m/z 1217 suggest an elongated fucosylated globo structure with a 1–4 linkage between the terminal end hexose and middle HexNac giving the ^0,2A₄ cross ring fragment. Symbol explanation: yellow square=N-acetylgalactosamine (GalNac), yellow circle=galactose (Gal), blue circle=glucose (Glc), blue square=N-acetylglucosamine (GlcNac), white square=N-acetylhexosamine (HexNac) and red triangle=fucose (Fuc).

Figure 3.

Summary of acid GSL found in Atlantic salmon epithelia. Symbol explanations: yellow square=N-acetylgalactosamine (GalNac), yellow circle=galactose (Gal), blue square=n-acetylglucosamine (GlcNac), blue circle=glucose (Glc), white square=N-acetylhexosamine (HexNac), red triangle=fucose (Fuc), purple diamond= N-acetylneuraminic acid (Neu5ac) and S=sulphate group.

Figure 4.

MS/MS of two native acid GSL from Atlantic salmon intestine. A) MS/MS of m/z 995.8 B) MS/MS of m/z 926.8. Symbol explanations: yellow square=N-acetylgalactosamine (GalNac), yellow circle=galactose (Gal), blue circle=glucose (Glc), red triangle=fucose (Fuc) and purple diamond= N-acetylneuraminic acid (Neu5ac).

Figure 5.

Distribution and A. salmonicida binding to α2–3 and α2–6 linked Neu5Ac. A) the relative abundance of α2–3 and α2–6 linked Neu5Ac on mucins and GSLs. Note that some glycan structures contain both α2–3 and α2–6 linked sialic acid. B) Binding of A. salmonicida to α2–3 and α2–6 sialyl-lacto HSA conjugates (8 µg/ml). No binding to α2–3 linked sialic acid was detected when compared to control (PBS), while the α2–6 linked sialic acids showed significant binding (n = 8). Data are shown after subtracting the average background control signal. Significance was calculated using Student’s-t test (****=p < 0.0001). Outliers in the background control (blank) were removed by ROUT (Q = 1%).

Figure 6.

Comparison of terminal epitopes and size between mucins and GSLs. A, B, D, E, G, H, J and K) Comparison of terminal epitopes from GSL and mucin O-glycans. The acid and neutral GSL data were proportionally combined based on isolation data from table 1. The data was calculated based on the weight of the GSL (neutral/acid GSL) and number of terminal moieties. C, F, I and L) Comparison of number of monosaccharides in the GSL or O-glycans from the same organ tissue, based on previously published data [8–11]. *neutral GSL containing one or two monosaccharides could not be included due to having too small mass for MS detection. ^# Mucin glycans containing a single monosaccharide can not be detected either.

Figure 7.

Location of sulfatide in Atlantic salmon pyloric caeca, skin and gills. (A, G and S) sulfatide tissue localization was visualized with the O4 antibody (green). (B, H, N and T) the tissue was outlined by CellMask (red). (C, I, O and U) DAPI (blue) stained the nuclei. (D, J, P and V) show the channels merged at 20x and (E, K, Q and X) at 40x. (F, L and R) Show close up images of the regions enclosed by white boxes in (E, K, Q and X). (M-R) show skin tissue without the O4 antibody, used as a background control.