Figure 1.
Isolation and differential expression of C. cinerea CCL1 and CCL2.
(A) Specific binding of CCL2 to horseradish peroxidase (HRP). Coomassie-stained SDS-PAGE showing Input, Flow through and Bound (Beads) fractions of a soluble protein extract from C. cinerea fruiting bodies upon affinity-chromatography using immobilized HRP. The Bound fraction was released by boiling the HRP-sepharose beads in Lämmli sample buffer. The loaded protein amount of the Bound fraction (Beads) corresponds to two equivalents of Input and Flow through fractions. Sizes of the marker proteins are indicated. (B) Immunoblot comparing expression levels of CCL2 between vegetative mycelium and fruiting bodies of C. cinerea. Equal amounts of total protein were loaded in each lane. A polyclonal antiserum raised in rabbits against purified CCL2 was used for detection. (C) Comparisons of relative expression ratio (or fold up-regulation) of the genes encoding CCL1 and CCL2 by qRT-PCR in fruiting bodies relative to vegetative mycelium. Error bars represent standard deviation of the mean.
Figure 2.
Carbohydrate-binding specificity of CCL2.
Fluorescently labeled CCL2 was analyzed for binding to the mammalian glycan array (V3.1) of the Consortium for Functional Glycomics (CFG). Results shown are averages of triplicate measurements of fluorescence intensity at a lectin concentration of 200 µg/ml. Error bars indicate the standard deviations of the mean. Glycan structures are depicted for those epitopes with highest relative fluorescence. The raw data and the entire list of glycans with the respective spacers can be found on the CFG homepage [http://functionalglycomics.org/] or in Tables S2 and S3. Binding of 6'sulfo-sialyllactose (glycan #45) is likely to be an artifact since it is also bound by fucose-binding lectin AAL [http://functionalglycomics.org/].
Figure 3.
Refining the specificity of the CCL2 lectin.
(A) The chemical and schematic structure of the fucosylated chitobiose (GlcNAcβ1,4[Fucα1,3]GlcNAc-spacer) that was used as ligand for binding studies and structure determination. Indicated is also the B face that is defined as the face on which the carbons are numbered in an anticlockwise order [69]. (B) Chemical shift deviations upon complex formation at a protein concentration of 0.4 mM at pH 5.7. Overlay of 15N-HSQC spectra of free CCL2 (blue) and CCL2 bound to one equivalent of fucosylated chitobiose (red). (C) Titration of the amide signal of T111 in CCL2 with fucosylated chitobiose using 15N-HSQC spectra. The protein∶ligand ratio is displayed on the left. (D) Plot of the chemical shift differences between free and bound CCL2 ( δ = [ δHN2+(δN/Rscale)2 ]1/2, Rscale = 5).
Table 1.
Binding of CCL2 wild-type to different carbohydrates and CCL2 variants to GlcNAcβ1,4[Fucα1,3]GlcNAcβ1-sp (sp: spacer O-[CH2]5COOH) measured with isothermal titration calorimetry and NMR spectroscopy at 299K.
Figure 4.
Solution structure of the CCL2 lectin in the absence of a ligand determined by NMR spectroscopy.
The side (A) and top (B) view of the most representative structure out of 20 structures is shown. The three pseudo symmetric sections of the β-trefoil fold corresponding to residues S9–N60, S61–S100 and G101–V142 are colored green, yellow and orange, respectively. Characteristic regions are labeled according to Renko et al. for better orientation [26]. (C) Chemical shift deviations mapped on the structure of CCL2 in the same orientation as in A. Chemical shifts of residues in red experience a combined NH chemical shift deviation >0.4 ppm, for residues in pink >0.15 ppm. (D) Secondary structure and subdomain borders displayed on the protein sequence. The same color code as in A and B is used. Bold residues are forming the hydrophobic core of the protein.
Figure 5.
NMR solution structure of the CCL2 lectin in complex with fucosylated chitobiose (GlcNAcβ1,4[Fucα1,3]GlcNAc).
(A) Intermolecular NOEs observed in a 3D 13C F1-edited F3-filtered HSQC-NOESY spectrum in a schematic presentation. (B) Structural ensemble of 20 structures of the protein backbone and the carbohydrate in cyan. The subunits α, β and γ are colored green, yellow and orange, respectively. The orientation is identical to Figure 4. (C) Ribbon presentation of the most representative structure. (D) Stereo view of the carbohydrate recognition site. Potential intermolecular hydrogen bonds are shown with dashed magenta lines. (E) Details of the interaction site illustrating how the trisaccharide is recognized by hydrogen bonds. (F) Summary of the interactions between the trisaccharide and CCL2. Potential H-bonds are indicated as dotted lines in magenta and hydrophobic interactions by green lines. (G) Crystal structure of the β-trefoil domain of the fungal lectin MOA in complex with the trisaccharide Galα1,3[Fucα1,2]Gal [19] showing all three occupied canonical binding sites (pdb∶3EF2). For better comparison, the same orientation and colors as in panel B and C were used.
Table 2.
NMR structure determination statistics of CCL2 in the free form and in complex with the fucosylated chitobiose (GlcNAcβ1,4[Fucα1,3]GlcNAc-spacer, the spacer [CH2]5COOH was truncated in the structure calculations to a methyl group.).
Table 3.
Potential intermolecular protein–carbohydrate hydrogen bonds based on the orientations and positions of the carbohydrate in the complex structure.
Figure 6.
Sequence conservation among CCL2-like proteins and comparison to two typical representatives of fungi and plants.
Sequence alignment of several fungal and plant R-type lectins. CCL2_A: CCL2 of C. cinerea strain AmutBmut; CCL2_O: CCL2 of C. cinerea strain Okayama7; CCL1_A: CCL1 of C. cinerea strain AmutBmut; CCL1_O: CCL1 of C. cinerea strain Okayama7; PP_L1: Postia placenta lectin 1 (Pospl1_130016); PP_L2: Postia placenta lectin 2 (Pospl1_121916); SL_L1: Serpula lacrymans lectin 1 (SerlaS7_144703); CP_L1: Coniophora puteana lectin 1 (Conpu1_119225); PO_L1: Pleurotus ostreatus lectin 1 (PleosPC9_89828); PO_L2: Pleurotus ostreatus lectin 2 (PleosPC15_1043947); PO_L3: Pleurotus ostreatus lectin 3 (PleosPC9_64199); PO_L4: Pleurotus ostreatus lectin 4 (PleosPC15_1065820); DS_L1: Dicomitus squalis lectin 1 (Dicsq1); AO_L1: Arthrobotrys oligospora lectin 1 (s00075g2); LB_L1: Laccaria bicolor lectin 1 (Lbic_330799); LB_L2: Laccaria bicolor lectin 2 (Lbic_327918); MOA: Marasmius oreades agglutinin; SNA-II: Sambucus nigra agglutinin/ribosome inactivating protein type II. The distantly related canonical R-type lectins MOA (fungal, 14% sequence identity) and SNA-II (plant, 13% sequence identity) were included in the alignment based on comparison of their 3D structures [70], [71]. The Clustal X color scheme was used. Residues involved in the carbohydrate recognition are indicated at the bottom for CCL2, MOA and SNA-II. The secondary structure of CCL2 and the conservation is indicated as well. The alignment was generated with Jalview [72].
Figure 7.
Thermodynamic binding parameters.
(A) ITC experiment of wild type CCL2 binding to GlcNAcβ1,4[Fucα1,3]GlcNAc-spacer. The raw calorimetric output is shown on top, the fitted binding isotherm at the bottom. The protein concentration in the cell was 70 µM, carbohydrate concentration in the syringe was 2.4 mM. (B) Thermodynamic binding parameters of CCL2 (in red) in comparison to other lectins with a focus on high affinity binding. Anti LeX Fab: Fab fragment of the monoclonal antibody 291-2G3-4; ConA: concavalin A from jack bean seeds (Canavalia ensiformis); CTB: cholera-toxin B subdomain; GS4: Griffonia simplicifolia lectin 4; MOA: Marasmius oreades agglutinin; RSL: Ralstonia solanacearum fucose-binding lectin; TeNT: tetanus neurotoxin; WBA II: winged bean (Psophocarpus tetragonolobus) acidic agglutinin. For simplicity, lectins that use Ca2+ for carbohydrate recognition are not displayed. Details for each correlation are found in Table S6. Data points in blue are discussed in the text.
Figure 8.
Carbohydrate-binding dependent biotoxicity of CCL2.
(A) Schematic representation of N-glycan structures in plants, insects and nematodes. Upper panel, left: Typical paucimannosidic plant N-glycan, highly abundant in HRP. Upper panel right: Fucosylated paucimannosidic N-glycan present in D. melanogaster. Lower panel: Fucose biosynthesis and N-glycan structure in C. elegans. Genes coding for enzymes involved in the fucose biosynthesis (lower panel, left) and fucose transfer to the core of N-glycans in C. elegans (lower panel, right) are indicated in dashed boxes. (B) Toxicity of recombinant E. coli expressing CCL2 (black bars) towards C. elegans wildtype (N2) and various fucosylation mutants. Error bars indicate standard errors of the mean. Asterisks (*) show cases where all data were 0. Significant differences were observed between the vector control and CCL2 for N2 (n = 10, p = 0.013), fut-1(ok892) (n = 10, p = 0.013) and fut-6(ok475) (n = 10, p = 0.013) worms, but not for bre-1(ye4) (n = 10, p = 0.329) or fut-6(ok475)fut-1(ok892) (n = 10, p = 0.329). (C) Fluorescence microscopy of C. elegans feeding on E. coli expressing a dTomato-CCL2 fusion protein, showing the grinder and anterior part of the intestine. (D) Toxicity of purified CCL2 towards D. melanogaster quantified as number of developed pupae (gray bars) or flies (black bars). BSA was included as control. Error bars indicate standard errors of the mean. Development of pupae and flies treated with CCL2 were significantly different from the control (pupae: n = 10, p = 0.013; flies: n = 10, p = 0.013). (E) Toxicity of E. coli expressing different CCL2 variants with mutations in residues involved in carbohydrate binding towards C. elegans wildtype (N2). Vector control and CCL2 wildtype (WT) were included as controls. Asterisks (*) show cases where all data were 0. Error bars indicate standard error of the mean. W78A, Y92A and W94A were significantly different from WT control (n = 10, p = 0.013), whereas L87A, N91A, V93A were not (n = 10, p = 1.0).