A novel lectin from Agrocybe aegerita shows high binding selectivity for terminal N-acetylglucosamine

Changes in glycosylation patterns play important roles in embryonic development, metabolic homoeostasis, pathogenic infections, immune surveillance and various diseases, including oncogenesis, diabetes and autoimmune disorders [1]. Therefore glycan analysis is of great importance for understanding the significance of glycan modifications. Lectins, with their diverse carbohydrate-binding properties, are potent tools for the study of glycan conjugates in various biological materials without glycan liberation. For example, lectins have been applied in the separation of cells at different stages, mitogenic activation of immune cells and identification of blood groups, as well as characterization of glycan alterations on the surface between normal and neoplastic cells [2–4].

The newly developed lectin array has been proved to be a powerful tool for the comprehensive analysis of glycans or glycoconjugates [5]. Much significant information has been obtained from use of lectin arrays. α1-2-Fucose-specific lectin (rBC2LCN) has been applied in the detection of undifferentiated induced pluripotent stem cells/embryonic stem cells, but not differentiated somatic cells [6], TKA (Trichosanthes kirilowii agglutinin) and PNA (peanut agglutinin) were used to distinguish the stem-like glioblastoma neurosphere cultured from a traditional adherent glioblastoma cell line [7], GNA (Galanthus nivalis agglutinin), NPA (Narcissus pseudonarcissus agglutinin), PSA (Pisum sativum agglutinin), LcH (Lens culinaris lectin) and Con A (concanavalin A) have been used for the detection of high-mannose N-linked oligosaccharides on differentiated neutrophils in comparison with the promyelocytic leukaemia cell line HL-60 [8], DSL (Datura stramonium lectin), SLL [Solanum lycopersicum lectin; formerly LEL (Lycopersicon esculentum lectin)] and MAL (Maackia amurensis lectin) showed a binding preference for mouse laminin, whereas SNA (Sambucus nigra agglutinin), SSA (Salvia sclarea agglutinin) and TJA-I (Trichosanthes japonica agglutinin-I) showed strong binding to bovine transferrin [9]. More lectins with unambiguous and/or distinctive glycan-binding selectivity are needed to be studied and exploited for probes of glycan structures. Thus there is an urgent need to find powerful lectin candidates for glycan analysis.

Lectins are useful tools for tumour diagnosis, antivirus research and drug-delivery studies [10]. For example, MAL has been used in prostate cancer diagnosis because of its preferential binding to prostate-specific antigen [11], HPA (Helix pomatia agglutinin) recognizes the glycosylation changes of metastatic breast cancer [12], BCA (Boodlea coacta lectin) can potently provide entry inhibition of HIV-1 and influenza viruses [13], microvirin has anti-HIV-1 activity with a high safety profile and low toxicity [14], and odorranalectin has been reported to be the smallest lectin so far and with potential for drug delivery and targeting [15]. An increasing number of lectins from plants and animals have been purified and characterized; however, the information on lectins isolated from fungal sources remains limited [16]. Fungal lectins are attractive because of their wide distribution, high content, varied carbohydrate-binding specificities and especially anti-tumour activities. Agaricus bisporus lectin was shown to possess anti-tumour activity against human colon cancer HT29 and breast cancer cell lines MCF-7 [17], Russula lepida lectin exhibited anti-proliferative activity in hepatoma HepG2 cells and human breast cancer MCF-7 cells [18], Tricholoma mongolicum lectin-1 and lectin-2 could inhibit the growth of sarcoma 180 cells [19], Pleurotus citrinopileatus lectin exerted potent anti-tumour activity in mice bearing sarcoma 180 [20], and Grifola frondosa lectin was shown to be cytotoxic to HeLa cells [21].

In the present paper, we report a novel fungal lectin AAL-2 (Agrocybe aegerita lectin 2) from the fruiting body of A. aegerita. A glycan array analysis showed that AAL-2 bound to non-reducing GlcNAc (N-acetylglucosamine) termini with high selectivity, even higher than the previously well-known GlcNAc-binding lectins WGA (wheatgerm agglutinin) and GSL-II (Griffonia simplicifolia lectin-II), which have been widely used in biochemical and biomedical research. Moreover, we showed that AAL-2 had anti-tumour activity not only for the induction of hepatoma cells apoptosis in vitro, but also for the inhibition of H22 hepatoma growth and prolongation of survival times of tumour-bearing mice in vivo. The cloning and prokaryotic/bacterial protein expression of AAL-2 provided clues to the deep investigation of the glycan-binding mechanism of AAL-2 and further application of AAL-2 in probing GlcNAc-related glycans or glycoconjugates.

The edible mushroom A. aegerita was collected from the Sanming Institute of Fungi (Sanming, Fujian, P.R. China). GlcNAc was purchased from Sangon Biotech. Epoxy-activated Sepharose 6B was purchased from GE Healthcare.

A murine hepatoma cell line H22, and a human hepatoma cell line Huh7, were provided by the CCTCC (China Center for Type Culture Collection, Wuhan University). Male BALB/c mice (6–8-week-old) were purchased from the Hubei Experimental Animal Laboratory (Hubei, China) and maintained in a pathogen-free facility. Procedures were performed according to approved protocols and in accordance with recommendations for the proper care and use of laboratory animals.

GlcNAc-coupled Sepharose 6B was prepared according to the manufacturer's instructions (GE Healthcare). In brief, 2 g of epoxy-activated Sepharose 6B was washed twice with 200 ml of distilled water, mixed with GlcNAc at a final concentration of 200 μmol/ml in distilled water (pH 13) at a temperature of 30°C for 15 h. Free remaining groups were blocked by incubating in 1 M ethanolamine, and the mixture was gently agitated at 45°C for 4 h. The Sepharose 6B-coupled GlcNAc matrix was washed extensively with TBS (Tris-buffered saline) (pH 7.2) and packed into an empty 100-mm-long glass column with a 10 mm inner diameter. The column was stored at 4°C until use.

Dried A. aegerita fruiting bodies were crushed into a powder, and 50 g of the dried mass was dissolved and extracted three times with 1 litre of distilled water at 4°C. The extracts were pooled and centrifuged at 10000 g for 15 min. Solid ammonium sulfate was added to the supernatant to 40% saturation and centrifuged again at 10000 g for 15 min. Solid ammonium sulfate was then added to the supernatant to obtain 80% saturation, and the 40–80% precipitate was collected by centrifugation at 10000 g for 15 min. The precipitated protein was dissolved in 200 ml of distilled water and dialysed extensively against 50 mM TBS (pH 7.2) by changing the dialysis buffer three times at 4°C overnight. The sample was loaded on to an affinity column (GlcNAc-coupled Sepharose 6B matrix) pre-equilibrated with TBS (pH 7.2), and eluted with 0.2 M GlcNAc in 50 mM TBS (pH 7.2) after extensive washing with TBS. The enriched AAL-2 fraction was stored at −20°C.

SDS/PAGE was performed with a 15% acrylamide running gel and 5% stacking gel. Protein bands were identified by Coomassie Brilliant Blue R-250 staining.

The band corresponding to AAL-2 was excised from a Coomassie Brilliant Blue-stained SDS/polyacrylamide gel and tryptically digested according to standard protocols. The peptides were derivatized for the N-terminus with SPITC (4-sulfophenyl isothiocyanate) and desalted with ZipTip™. MALDI (matrix-assisted laser-desorption ionization)–MS was performed on a 4700-Proteomics Analyzer (Applied Biosystems) and PSD (power spectral density) spectra were collected with 3000 laser shots at 1 keV collision energy without CID (collision-induced dissociation). MS data were analysed by PMF (peptide mass fingerprinting) using MASCOT software.

MALDI–TOF (time-of-flight) mass spectra were obtained using a Voyager-RP mass spectrometer (PerSeptive Biosystems) [22].

Total RNA was prepared from the fresh fruiting bodies of A. aegerita by grinding in liquid nitrogen followed by extraction with Plant RNA Isolation Reagent (BioTeke). Full-length cDNAs were synthesized from 1 μg of total RNA using MMLV (Moloney murine leukaemia virus) reverse transcriptase (Promega) after RQ1 RNase-free DNase digestion (Promega) according to the manufacturer's instructions. Nested PCR was applied to amplify the AAL-2 coding sequence, and two pairs of primers were designed as follows: forward external primer aal-2_5′_F1, 5′-ATGAACGTGAACGTGGGAGG-3′; reverse external primer aal-2_3′_R1, 5′-CTAGTAATCCGTGCGAGTCG-3′; forward ORF (open reading frame) primer aal-2_5′_F2, 5′-ATGACCAGCAACGTCATCAC-3′; reverse ORF primer aal-2_3′_R2, 5′-ATGCCTGCTGCTGTCCAAGT-3′. PCR was performed using Taq DNA polymerase (Promega) with the following temperature profile: 5 min at 95°C, 30 s at 94°C and 3 min at 72°C, for 35 cycles. The amplified product obtained with external primers was used as the template for a second round of amplification with the ORF primers. The purified PCR fragment was cloned into the pET-30a vector (Novagen) according to the manufacturer's instructions. The resulting plasmid containing rAAL-2 coding sequence was used for transformation into E. coli BL21 cells (Novagen). Induction of protein expression was performed with 1 mM IPTG (isopropyl β-D-thiogalactopyranoside) at mid-exponential phase (D₆₀₀ of 0.4–0.6), and the bacteria were grown for an additional 5 h at 37°C with shaking at 220 rev./min. The bacteria were then harvested and lysed. The supernatant was loaded on to a GlcNAc–Sepharose 6B column, and rAAL-2 was purified by GlcNAc-coupled Sepharose 6B affinity chromatography as described above.

To biotinylate AAL-2, purified AAL-2 (200 μg in 250 μl of PBS) was mixed with 400 μl of biotin ester solution (100 μg of biotin ester/200 μg of lectin) and left for 30 min at room temperature (24°C). The biotin-labelled AAL-2 was collected using Sephadex G-250 and stored at −20°C until use.

To determine the carbohydrate-binding specificity of AAL-2, glycan array analysis was performed by the Consortium for Functional Glycomics (Core H; http://www.functionalglycomics.org/static/consortium/resources/resourcecoreh.shtml). The printed glycan array (Mammalian Printed Array version 4.1) [23] was composed of 465 different natural and synthetic mammalian glycans. In the binding assay, array slides, containing six replicates per glycan or glycoconjugate, were incubated with different concentrations of biotinylated AAL-2 (0.1, 1, 50 and 200 μg/ml). The slides were then washed, and bound lectin was detected using fluorescently labelled streptavidin. Array slides were subjected to imaging, and the fluorescence was measured as described previously [24]. Briefly, fluorescence intensity was detected using a ScanArray 5000 confocal scanner (PerkinElmer). ImaGene® image analysis software (BioDiscovery) was used to analyse the image. The relative binding for each glycan was expressed as mean RFU (relative fluorescence units) of four of the six replicates.

ITC experiments on the interaction of GlcNAc with AAL-2 were performed at 25°C with a VP-ITC isothermal titration calorimeter (Microcal). Freshly purified AAL-2 was dialysed overnight in PBS (pH 7.2) at 4°C. The protein concentration in the microcalorimeter cell (1.4478 ml volume) was 50 μM. GlcNAc solution was made up to 7 mM in PBS and placed in the syringe. The first injection (4 μl) was followed by 27 injections of 10 μl, and the stirring rate was set at 300 rev./min. The dilution heat of the GlcNAc was measured by injecting GlcNAc solution into buffer alone and was subtracted from the experimental curves before data analysis. The experimental data were fitted to a theoretical titration curve using Microcal Origin software supplied with the instrument, and the standard molar enthalpy change for the binding, Δ_bH_m⁰, and the dissociation constant, K_d, were derived. The standard molar free energy change, Δ_bG_m⁰, and the standard molar entropy change, Δ_bS_m⁰, for the binding reaction were calculated by using the following thermodynamic equations:

The haemagglutination activity of purified lectin was determined as described previously [25]. Briefly, the haemagglutination activity assay was performed on a microtitre plate, and the lectin sample (50 μl, 1 mg/ml) was serially diluted with PBS and mixed with 50 μl of a 2% suspension of chicken red blood cells at room temperature. Measurements were taken after 30 min when the control (no lectin added) was fully precipitated.

The effects of metal cations on the haemagglutination activity of AAL-2 were determined in the absence and presence of 10 mM Ca²⁺, Zn²⁺, Mn²⁺ or Mg²⁺ after dialysis against 20 mM EDTA. The effect of temperature on lectin stability was monitored over a temperature range by incubation with AAL-2 for 30 min at each temperature. The samples were then cooled rapidly and assayed for haemagglutination activity. The effect of pH on lectin stability was determined by incubating with AAL-2 with buffers of different pHs for 30 min, and haemagglutination activity was determined after neutralization.

A rabbit polyclonal antibody against AAL-2 was prepared as reported previously [26]. Briefly, 500 μg of nAAL-2 (native AAL-2) in 500 μl of PBS was emulsified with 500 μl of complete Freund's adjuvant (Sigma). The emulsion was injected into 6-month-old New Zealand White rabbits by multipoint subcutaneous implantation. The second and third inoculations were performed on days 24 and 36 with Freund's incomplete adjuvant (Sigma), and the fourth inoculation was performed on day 41 without any adjuvant. At 1 week after the final immunization, sera were separated from the collected blood samples and antibody against nAAL-2 was detected by ELISA.

H22 cells and Huh7 cells were cultured in RPMI 1640 medium and DMEM (Dulbecco's modified Eagle's medium) respectively. Cell culture media were supplemented with 10% (v/v) heat-inactivated (56°C, 30 min) FBS (fetal bovine serum), 5 units/ml penicillin and 50 μg/ml streptomycin. All cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂. For all experiments, cells were plated 2 h before treatment to allow for adherence.

In the cell-adhesion assay, H22 and Huh7 cells were incubated with or without AAL-2, and then with polyclonal antibody against AAL-2 and FITC-conjugated goat anti-rabbit IgG. For Annexin V/PI staining, H22 and Huh7 cells were treated with AAL-2 at final concentrations of 0, 0.6, 1.2 or 2.4 μM for 24 and 36 h respectively. The cells were collected, washed twice with ice-cold PBS (pH 7.4), and stained using an Annexin V/PI apoptosis kit (MultiScience Biotech). The samples were mixed gently and incubated in the dark. All samples were analysed by flow cytometry (Beckman, Epics Altra)

H22 cells (10⁷) in 100 μl of PBS were subcutaneously injected into the right flank of BALB/c mice. At days 7, 9, 11, 13, 15, 17 and 19 after tumour inoculation, mice were injected at the tumour site with 5 mg of AAL-2/kg in 100 μl of PBS or diluent control. Tumour volume (V) was determined by caliper measurements (V=L×W²/2), where L is length and W is width.

The two-sample Student's t test was used for comparisons between groups. Statistical analysis was performed with GraphPad Prism 5 software. Results are means±S.E.M., and statistical significance was defined as P<0.05.

Total crude proteins were extracted from dried A. aegerita fruiting bodies by water-extraction, 40–80% ammonium sulfate precipitation and dialysis against TBS. The total crude proteins were loaded on to the GlcNAc-coupled Sepharose 6B column and eluted with 0.2 M GlcNAc after extensive TBS washing (Figure 1A). A total fraction of 6 ml of eluate (48–54 ml) was collected. Representative purification fractions from the affinity chromatographic column were separated by SDS/PAGE in the presence of 2-mercaptoethanol. The eluate fraction yielded a single band (approximately 43 kDa) and was named AAL-2 (Figure 1B). AAL-2 was characterized further using de novo sequencing by MS/MS (tandem MS), and five tryptic peptides were identified (Figure 1C). The five peptide sequences were analysed further by BLASTP at NCBI (http://www.ncbi.nlm.nih.gov/BLAST), and only two peptides (labelled #1 and #3) showed homology with the GlcNAc-binding lectin PVL (Psathyrella velutina lectin) [27] (results not shown).

To determine the coding sequence of AAL-2, the A. aegerita transcriptome was sequenced. Raw Illumina sequencing data of the A. aegerita transcriptome was deposited in the Sequence Read Archive at NCBI (accession number SRA026731.1). The five tryptic AAL-2 peptides were searched for similarity against the A. aegerita transcriptome by BLASTX. The ORF of the complementary strand of UniGene 4198 was analysed, and a fragment containing 1224 bp was identified and labelled as the AAL-2 coding sequence (Figures 2A and 2B). Five tryptic peptides (underlined) from AAL-2 showed full homology with the protein translated from the identified AAL-2 coding sequence, with the latter MALDI–TOF/TOF analysis showing 69% coverage (Figure 2B, grey box); moreover, the molecular mass was calculated to be 43175 Da by the ExPASy Proteomics Server (http://www.expasy.org/tools), and was consistent with the AAL-2 mass (43175.72 Da) identified by MALDI–TOF-MS (Figure 2C). The isoelectric point of AAL-2 was calculated to be 8.19 by ExPASy.

The AAL-2 coding sequence was successfully cloned after nested PCR using total cDNA as a template (Figure 3A). The nucleotide sequence of the cloned AAL-2 coding sequence was fully identical with that from the A. aegerita transcriptome. The AAL-2 coding sequence was then cloned into the expression vector pET-30a, and the resulting plasmid was used for the transformation of E. coli BL21 cells. The bacteria were grown to mid-exponential phase (D₆₀₀ of 0.4–0.6) and then induced with 1 mM IPTG. After 5 h of induction, the lysate of the bacteria was separated by SDS/PAGE, and a strong band of approximately 43 kDa appeared compared with non-induced bacteria (Figure 3B). The supernatant fractions from the lysate were loaded on to the GlcNAc-coupled Sepharose 6B column. rAAL-2 was eluted as a single symmetrical peak at the same elution volume as nAAL-2 (results not shown), and produced an electrophoretic mobility by SDS/PAGE identical with that of nAAL-2 at approximately 43 kDa. In addition, polyclonal antibody raised against nAAL-2 was able to detect rAAL-2, and a precipitin band was formed identical with that of nAAL-2 (Figure 3C).

The purification of AAL-2 by affinity chromatography showed that AAL-2 was able to bind to GlcNAc. To characterize further the glycan-binding profile of AAL-2, glycan array analysis was performed by testing the binding affinity of AAL-2 to 465 glycan candidates. The glycan array was performed at different concentrations (0.1, 1, 50 and 200 μg/ml) of AAL-2, and the binding glycans with more than 5000 RFU were listed according to the glycan-binding profile of AAL-2 at 200 μg/ml (Supplementary Table S1 at http://www.BiochemJ.org/bj/443/bj4430369add.htm). The results showed that the binding selectivity changed little with a wide concentration range (Figure 4A, lower panel). Almost all of the top 30 glycans in the binding profiles possessed terminal non-reducing GlcNAc (Figure 4A, upper panel, and Supplementary Figure S1 at http://www.BiochemJ.org/bj/443/bj4430369add.htm). In addition, Gal (galactose)–GlcNAc [LacNAc (N-acetyl-lactosamine)] linked behind would increase the binding affinity to the terminal GlcNAc, and the glycan GlcNAc-(LacNAc)₃ exhibited the highest binding affinity among the 465 glycan candidates (Figure 4A, upper panel). However, (LacNAc)₃, with the absence of the terminal GlcNAc compared with GlcNAc-(LacNAc)₃, showed significantly low binding affinity. In comparison with other GlcNAc-binding lectins (Figure 5), AAL-2 performed better than the well-known GlcNAc-binding lectins WGA and GSL-II, which have been used widely in glycan studies for their high GlcNAc-binding selectivity. Therefore AAL-2 can be defined as a lectin with high binding selectivity to glycans with non-reducing GlcNAc termini and has potential in GlcNAc-related glycan or glycoconjugate analysis.

ITC was also performed in order to thermodynamically characterize the GlcNAc and AAL-2 interactions. Figure 4(B) (upper panel) shows the raw ITC curve resulting from the injection of GlcNAc into a solution of AAL-2. The titration curve showed that GlcNAc binding to AAL-2 was exothermic, resulting in negative peaks in the plots of power against time. Figure 4(B) (lower panel) shows the plot of the heat evolved per mol of GlcNAc added, with the heat of GlcNAc dilution subtracted. Our calorimetric data were fitted to a sequential binding model, and the thermodynamic parameters for the binding of GlcNAc to AAL-2 are summarized in Table 1. The values of K_d1 and K_d2 are 195 and 459 μM respectively, which indicates that GlcNAc bound to AAL-2 with moderate binding affinity.

AAL-2 could bind to H22 and Huh7 hepatoma cells (Figures 6A and 6B) and induce apoptosis of H22 cells and Huh7 cells in vitro. The amount of apoptosis of H22 cells was 1.8% after 24 h of culture, and increased significantly to 4.4%, 6.3% and 10% at final concentrations of 0.6, 1.2 and 2.4 μM respectively (Figures 6C and 6D). Similarly, the proportion of apoptotic Huh7 cells was 4.7%, and increased to 7.1%, 13.3% and 36.3% when treated with AAL-2 at final concentrations of 0.6, 1.2 and 2.4 μM respectively for 36 h (Figures 6E and 6F). The results indicate that AAL-2 induced cell apoptosis in a concentration-dependent manner.

The anti-hepatoma activity of AAL-2 was investigated further by using a mouse model of H22 hepatocellular carcinoma. Mice treated with AAL-2 showed a significant reduction in tumour volume (Figure 7B), whereas body weight remained unaltered (Figure 7A) compared with control mice treated with PBS. In addition, AAL-2 treatment was able to extend the survival time of tumour-bearing mice (Figure 7C). Overall, we conclude that AAL-2 has anti-hepatoma activity by inhibiting hepatocellular carcinoma growth and can extend survival times of tumour-bearing mice.

A. aegerita is one of the most cultivated mushrooms in China, and has been highly valued as a functional food for its anti-tumour and nutritional benefits [28]. We have previously purified AAL (A. aegerita lectin) from A. aegerita, which showed extensive anti-tumour activity owing to specific binding to the Thomsen–Friedenreich disaccharide tumour antigen and subsequently resulted in tumour cell apoptosis [29,30]. In the present study, a novel fungal lectin AAL-2 was purified from A. aegerita by using one-step affinity chromatography. AAL-2 showed high binding selectivity to glycans with a non-reducing GlcNAc terminus, and the binding affinity increased when one or more LacNAc disaccharides were linked to the terminal GlcNAc via α1-3, α1-4 or β1-3, β1-4 linkages to form a GlcNAc-(LacNAc)_1–3 structure, implying that the addition of a LacNAc could increase the binding affinity to terminal GlcNAc (Supplementary Table S1). The glycan GlcNAc-(LacNAc)₃ represented the highest binding affinity among the 465 candidates; however, the high binding affinity would be lost if the terminal GlcNAc was removed. In addition, the binding selectivity changed little at a wide concentration range of AAL-2, indicating that the glycan-binding activity of AAL-2 was highly selective, low noise and was not affected by the high concentration (200 μg/ml) (Figure 4A, lower panel).

Our glycan array analysis revealed that AAL-2 bound to terminal GlcNAc with high selectivity. We then compared further the glycan-binding specificity of AAL-2 with that of other GlcNAc-binding lectins: PVL, BLL (Boletopsis leucomelas lectin), GSL-II, WGA, HPA and SLL. The GlcNAc-binding specificity of PVL was determined by affinity chromatography, and PVL was shown to have preferential binding to oligosaccharides bearing non-reducing terminal β-GlcNAc linked 1-6 or 1-3, but poor affinity for 1-4 linkages [27], whereas AAL-2 bound preferentially to oligosaccharides bearing a non-reducing terminal β-GlcNAc linked via a 1-4 or 1-3, rather than a 1-6, linkage. The unique structure of carbohydrate-binding domains might contribute to the binding preference of AAL-2 to terminal β-GlcNAc-linked oligosaccharides; our future studies will focus on the analysis of the carbohydrate–lectin interactions via crystallization of the AAL-2 and GlcNAc complex. The GlcNAc-binding specificity of BLL was determined by frontal affinity chromatography, and BLL was shown to have high binding affinity for mono-, bi- and tri-antennary agalactosylated glycans [31], whereas AAL-2 showed high binding affinity for mono-terminal non-reducing GlcNAc residues. Other GlcNAc-binding lectins, i.e. GSL-II, WGA, HPA and SLL, with their top five highest binding glycans from the glycan array data are listed in Figure 5. The comparative analysis showed that AAL-2 and GSL-II had the highest binding selectivity for terminal GlcNAc, whereas AAL-2 showed even higher binding selectivity than did GSL-II, a well known and widely studied lectin for the recognition of terminal GlcNAc residues. Moreover, AAL-2 binds to glycans in a metal-ion-independent manner, whereas Ca²⁺ is required for GSL-II-binding activity (results not shown). The binding activity of AAL-2 was stable under 70°C and within the pH range 6.0–11.0 (Supplementary Figure S2 at http://www.BiochemJ.org/bj/443/bj4430369add.htm). We compared further the amino acid sequences of AAL-2, GSL-II, WGA and HPA. Surprisingly, amino acid sequences of these GlcNAc-binding lectins showed low similarity to each other, although they shared similar glycan-binding properties (Supplementary Figure S3 at http://www.BiochemJ.org/bj/443/bj4430369add.htm). AAL-2, although newly discovered, has the potential to be utilized as a molecular probe to detect terminal GlcNAc sugars with high binding selectivity in glycan analysis.

Polylactosamine has been reported to be present on the immune co-stimulatory molecules CD28 and CD19. In a previous study [32], B-cells, T-cells and macrophages in the absence of polylactosamines were found to be hypersensitive to stimulation, and SLL was used in the detection of the polylactosamines with at least three lactosamine unit repeats [(LacNAc)_n>3] [32]. By using glycan arrays to detect the binding profile of lectins, SLL showed high binding selectivity for Neu5Ac (N-acetylneuraminic acid)/Fuc (fucose)-(LacNAc)₃ and (LacNAc)₃ (Consortium for Functional Glycomics database: primscreen_PA_v2_711_09262005). In the present study, the glycan array data showed that AAL-2 bound to GlcNAc-(LacNAc)₃ with the highest affinity among the 465 glycan candidates, indicating that AAL-2 recognized polylactosamine (LacNAc)₃ with a terminal GlcNAc modification. Overall, the results imply that AAL-2 and SLL would form pairs of complementary probes to monitor changes and modifications of polylactosamine.

GlcNAc has been reported to be aberrantly expressed in several carcinomas, resulting in increased levels and abnormal types of GlcNAc on the cell surface [33,34]. The present study suggests that AAL-2, with high binding selectivity to GlcNAc, could bind to hepatoma cells and induce cell apoptosis. This binding property probably endowed AAL-2 with binding priority to those tumour cells, therefore leading to its potential applications for cancer diagnosis.

AAL-2

Agrocybe aegerita lectin 2

BLL

Boletopsis leucomelas lectin

GlcNAc

N-acetylglucosamine

GSL-II

Griffonia simplicifolia lectin-II

HPA

Helix pomatia agglutinin

IPTG

isopropyl β-D-thiogalactopyranoside

ITC

isothermal titration calorimetry

LacNAc

N-acetyl-lactosamine

MAL

Maackia amurensis lectin

MALDI

matrix-assisted laser-desorption ionization

MS/MS

tandem MS

nAAL-2

native AAL-2

ORF

open reading frame

PI

propidium iodide

PVL

Psathyrella velutina lectin

rAAL-2

recombinant AAL-2

RFU

relative fluorescence units

SLL

Solanum lycopersicum lectin

TBS

Tris-buffered saline

TOF

time-of-flight

WGA

wheatgerm agglutinin

Shuai Jiang, Yi Liang and Hui Sun designed the study. Yijie Chen, Man Wang, Yalin Yin, Yongfu Pan and Bianli Gu carried out the research. Guojun Yu and Yamu Li analysed data. Shuai Jiang, Barry Wong and Hui Sun wrote the paper.

We thank Ronald L. Schnaar (Johns Hopkins School of Medicine, Baltimore, MD, U.S.A.), Richard D. Cummings (Emory University School of Medicine, Atlanta, GA, U.S.A.), Steven Rosen (University of California, San Francisco, San Francisco, CA, U.S.A.) and Xiangdong Fu (University of California, San Diego, San Diego, CA, U.S.A.) for their suggestions on the glycan array and data analysis. We also thank Yingzhu Zhu (Wuhan University) and Bo Liu (Sichuan University, Chengdu, China) for their critical comments, and Jiajia Xie (Wuhan University) for his technical assistance. The glycan array screening resources were provided by the Consortium for Functional Glycomics (No. GM62116) (Emory University School of Medicine).

This work was supported by the National Key Technologies R&D Program [grant numbers 2009ZX09103-676, 2009ZX09301-014-1 and 2008ZX10001-002], the Natural Science Foundation of China [grant number 81102850], the Major State Basic Research Development Program of China, 973 Program [grant number 2011CB811302], the Chinese 111 project [grant number B06018] and Wuhan municipal project [grant number 201160923296].