Prototype galectins are versatile modulators of cell adhesion and growth via their reactivity to certain carbohydrate and protein ligands. These functions and the galectins' marked developmental regulation explain their attractiveness as models to dissect divergent evolution after gene duplication. Only two members have so far been assumed to constitute this group in chicken, namely the embryonic muscle/liver form {C-16 or CLL-I [16 kDa; chicken lactose lectin, later named CG-16 (chicken galectin-16)]} and the embryonic skin/intestine form (CLL-II or C-14; later named CG-14). In the present study, we report on the cloning and expression of a third prototype CG. It has deceptively similar electrophoretic mobility compared with recombinant C-14, the protein first isolated from embryonic skin, and turned out to be identical with the intestinal protein. Hydrodynamic properties unusual for a homodimeric galectin and characteristic traits in the proximal promoter region set it apart from the two already known CGs. Their structural vicinity to galectin-1 prompts their classification as CG-1A (CG-16)/CG-1B (CG-14), whereas sequence similarity to mammalian galectin-2 gives reason to refer to the intestinal protein as CG-2. The expression profiling by immunohistochemistry with specific antibodies discerned non-overlapping expression patterns for the three CGs in several organs of adult animals. Overall, the results reveal a network of three prototype galectins in chicken.

INTRODUCTION

Divergence after gene duplication is the common route towards groups of closely related proteins. They can acquire characteristic properties by sequence alterations. However, the high degree of homology inevitably entails problems to reliably separate members of a gene family on the level of the proteins. It is thus mandatory to pinpoint problematic cases and definitively clarify them to avoid the risk of misclassification. We herein accordingly deal with a special conundrum from the realm of animal lectins, i.e. the galectins. This class of lectins is receiving increasing attention, and for good reason. Galectins are potent regulators of adhesion, cell migration, growth control and tissue invasion [1]. Their large impact on cell sociology is a key factor to explain why galectins are becoming attractive models for the study of divergent evolution. Towards this end, the prototype proteins (non-covalent homodimers or monomers) of chicken are an appealing choice, because their group size is considerably smaller than in mammals [2,3].

On the grounds of electrophoretic mobility or chronological order of detection, the embryonic pectoral or thigh muscle/adult liver form was referred to as C-16 or CLL-I [16 kDa; chicken lactose lectin, later named CG-16 (chicken galectin-16)], and the protein isolated from embryonic skin or adult intestine was referred to as C-14 or CLL-II, then CG-14 [410]. Analysis of the sequences from liver and skin CGs with a focus on phylogenesis suggested that they have likely diverged after, but not far away from, the separation of the lines for birds and mammals [3,1113]. As a consequence of the ensuing sequence evolution, not only organ patterns of expression but also biochemical properties such as the pI values have acquired individual attributes. In detail, the pI value of CG-16 was determined to be between 4.1 and 4.5, one report indicating focusing in the acidic range without satisfactory reproducibility [6,1416]. In contrast, the pI value of CG-14 from embryonic chicken skin was 7.0 [9], while that of CG-14 from intestine was between 6.3 and 6.7 [6,15,16], thus somewhat but not decisively different. Gel filtration and haemagglutination revealed only CG-16 to be completely homodimeric, which likely has a bearing on differential CG activities on cells [6,10,1521]. Fittingly, immunological cross-reactivity has so far not been reported between CG-16 and the CG-14 from intestine, though one study on CG-14 from embryonic skin detected reactivity [8,15,22,23]. Reservations against using cross-reactivity as a strong argument for protein classification notwithstanding, it is noteworthy that antibody preparations against CG-14 from the two tissue types generated disparate immunohistochemical staining of the neural tube and only the antiserum against intestinal CG-14 was free of binding activity against CG-16 [22,24,25].

These results on pI and antibody reactivity raised suspicion that the assumed identity among CG-14 proteins from embryonic skin and adult intestine might not be valid. Actually, this concern was nourished by the recently noted existence of a putative gene for a third prototype galectin in the chicken genome. Separated from CG-14/CG-16, which are closely related to human galectin-1, this newly detected sequence was assigned to the same branch of the phylogenetic tree as mammalian galectin-2 and tentatively named as CG-2 [3]. Whether a so far not detected CG-2 protein is really produced or whether the intestinal CG-14 form, which so far has not been sequenced, may have inadvertently been misclassified are open questions. To answer them, we cloned and expressed CG-2 and then set up a reliable procedure to distinguish CG-2 from the recombinant CG-14 and tested the intestinal protein. CG-2, as it turned out, is identical with the intestinal protein and shares quaternary structure with human galectin-2. Its hydrodynamic shape parameters measured by ultracentrifugation, however, are unusual for a homodimeric galectin. Next, we proceeded to examine genomic sequences of the proximal promoter regions for the three CGs as well as to measure their expression profiles by RT (reverse transcriptase)–PCR/Western-blot analyses and immunohistochemistry. In summary, the present study defines a third prototype CG with characteristic hydrodynamic behaviour and distribution in several types of organs of adult animals. The results presented give reason to revise the current nomenclature.

EXPERIMENTAL

Cloning, protein expression and purification

Total RNA from embryonic chicken kidney (developmental day 15) was isolated using the RNAeasy kit (Qiagen, Hilden, Germany) following the manufacturer's instructions, and 2.5 μg was used for first-strand cDNA synthesis. Reverse transcription was performed using a routine protocol [20], and PCR amplification to obtain full-length cDNA was based on specific primer design according to entry no. CF251077 in the NCBI (National Center For Biotechnology Information; Bethesda, MD, U.S.A.) dbEST database, resulting in the following sequences: sense primer 5′-CCATGGCTAGAATGTTTGAAATGTTCAACCTGG-3′ with an internal NcoI restriction site (underlined) and antisense primer 5′-GGATCCTCACTCCACCTTGAAGGAGGTAAC-3′ with an internal BamHI restriction site (underlined). The encoded full-length protein was expressed in the Escherichia coli strain M15 (pREP4) with the vector pQE60 (Qiagen) at optimal conditions, yielding approx. 11.5 mg of CG-2 from 1.2 litres of TB (Terrific Broth) medium (Roth, Karlsruhe, Germany) using four separate 1 litre flasks, induction with IPTG (isopropyl β-D-thiogalactoside) at a final concentration of 0.4 mM and a shift of temperature from 37 °C during initial growth to 30 °C. Lectin purification from extracts after sonification and centrifugation was performed with affinity chromatography on lactosylated Sepharose 4B, obtained after divinyl sulfone activation and ligand conjugation, as a crucial step [21]. The recombinant forms of CG-14 and CG-16 were similarly produced after cloning of the respective cDNAs, and these proteins as well as the lectin from fresh chicken intestinal tissue were purified to homogeneity as described in detail previously [16,20].

Analytical procedures

Gel electrophoretic analysis was carried out in the one-dimensional setting, using either 4% stacking/15% running gel parts or a 4–15% linear gradient system, as well as in the two-dimensional system, as described previously [26]. The protein spots were visualized by silver staining; the measured pI values of the CGs were compared computationally with the results of theoretical calculations by using the respective program of the EXPASY (Expert Protein Analysis System) proteomics server at the Swiss Institute of Bioinformatics (Basel, Switzerland) (http://www.expasy.org/tools/pi_tools.html) and assigned experimentally using marker proteins and human galectins of known properties. For MS analysis, spots representing recombinant or intestinal galectins respectively were excised from the second-dimension gel and washed; protein was digested in situ in solution without alkylation of SH groups of cysteine residues, using 90 ng of sequencing-grade trypsin (Promega, Heidelberg, Germany) in 40 mM ammonium bicarbonate at 37 °C for 4 h, and further sample processing followed the recently described protocol used for human galectin-1 [27]. Calibration of the tryptic ‘fingerprint’ spectra was performed externally by a two-point linear fit, using autolysis products of angiotensin I and oxidized B chain of bovine insulin. Assignment of the observed peptide signals was done by comparison with the calculated monoisotopic tryptic peptide masses for CG-14 and CG-2. Gel filtration was carried out with a prepacked Superose 12 HR 10/30 column (void volume of 8.3 ml) in 50 mM PBS (pH 7.2) containing 4 mM 2-mercaptoethanol without/with 100 mM lactose at a flow rate of 0.5 ml/min using a set of commercial marker proteins (Bio-Rad, Munich, Germany) and two galectins of known quaternary structure for calibration, as described in [28]. Haemagglutination assays with trypsin-treated, glutaraldehyde-fixed rabbit erythrocytes in 2-fold serial dilutions were carried out in the absence and presence of lactose as inhibitor in microtitre plate wells as described in [29]. Sedimentation-equilibrium experiments were performed by centrifugation of 80 μl samples, adjusted to different protein concentrations, at 12000, 16000 and 20000 rev./min and 20 °C for 12 h in an Optima XL-A analytical ultracentrifuge (Beckman Coulter, Krefeld, Germany) with an AN50-Ti rotor, and sedimentation-velocity experiments were run at 45000 rev./min and 20 °C for 4 h using 400 μl samples as described in [30]. Galectin-dependent staining of human Capan-1 pancreatic carcinoma cells with reconstituted expression of the tumour suppressor p16INK4a, kindly provided by Dr Katharina M. Detjen (Medizinische Klinik mit Schwerpunkt Hepatologie und Gastroenterologie, Charité-Universitätsmedizin, Berlin, Germany), was analysed by flow cytofluorimetry using streptavidin/R-phycoerythrin as a fluorescent marker (1:40; Sigma, Munich, Germany) with biotinylated galectins (10 μg/ml) [27]. Protein labelling by the N-hydroxysuccinimide ester of biotin (Sigma) under activity-preserving conditions, assays ascertaining lectin activity after labelling and determination of the extent of biotinylation followed optimized protocols [21,26,28]. Carbohydrate-dependent lectin binding to cell surfaces was determined in the presence and absence of a mixture of lactose (75 mM) and the glycoprotein asialofetuin (1 mg/ml). Carbohydrate-independent cell surface binding by protein–protein interactions was minimized by the presence of 100 μg/ml sugar-free BSA in Dulbecco's PBS.

Sequence processing

Using the consensus sequence RFAVNLQCGFSVXPGNDIAFHFNPRFDEHNGYXAVVRNTQINGSWGPEERKLPTHMPFSRGQPFEVCILCEXHCFKVAVNGQH, which originated from alignment of sequences of human, rat and mouse galectins, as the template, a TBLAST search through the most recent version of annotation of the chicken genome available from the NCBI (http://www.ncbi.nlm.nih.gov/) was performed. Independently, the information given for CG-2 under accession no. XM_001234399 was used, particularly to locate genomic sequences upstream of the site for initiation of transcription. Corresponding sequence information for genes of CG-14 (based on M11674) and CG-16 (AY553270) was retrieved from the same database, the promoter region of CG-16 still containing a stretch of unknown sequence between the positions −233 and −110. DNA sequences were edited using the EditSeq sequence analysis software version 4.0 (DNAstar, Madison, WI, U.S.A.) as well as the Reformat and Map algorithms included in the GCG (Genetics Computer Group Inc. Sequence Analysis Software Package) programs available from the HUSAR biocomputing service of the German Cancer Research Center (Heidelberg, Germany; http://genius.embnet.dkfz-heidelberg.de/). In order to scour proximal promoter regions for putative transcription-factor-binding sites and compare obtained results systematically, the online versions of the programs Match™ (http://www.generegulation.com/cgi-bin/pub/programs/match/bin/match.cgi) and P-Match™ (http://www.gene-regulation.com/cgi-bin/pub/programs/pmatch/bin/p-match.cgi) were selected, both using the latest update of the TRANSFAC® transcription factor database as the source for weight matrices and consensus sequences. Presettings were deliberately chosen to include ‘low quality’ matrices to avoid excluding important factors [such as Sp1 (stimulating protein-1)] for which only ‘low quality’ weight matrices are listed in the database. Stringency of screening was compensatorily increased by adjusting the cut-off to the most restricted value to limit occurrence of false-positive cases.

Expression profiling by RT–PCR, Western blotting and immunohistochemistry

RT–PCR analysis with cDNA preparations from tissues was performed with the primer sets given above to produce the 396 bp product in the case of CG-2 and a published set for CG-14 resulting in a 411 bp product [20]. Non-cross-reactive polyclonal antibody preparations were obtained after immunizing rabbits and removing any contaminating cross-reactive material from the IgG fractions by chromatographic affinity depletion with the respective galectin as matrix-conjugated ligand, as described for mammalian lectins [31]. Complete removal of cross-reactivity was ascertained by ELISAs and Western-blot assays. The resulting IgG preparations were applied in Western blotting of tissue extracts and immunohistochemical processing using fluorescence microscopy, where rigorous specificity controls to spot any antigen-independent staining were included, as described in [31,32].

RESULTS AND DISCUSSION

Mass profiling of CG-2

After cloning and recombinant expression of CG-2, its properties were first compared with those of CG-14 and CG-16 by one- and two-dimensional gel electrophoresis. Using running gels with either a fixed acrylamide concentration (15%) or a linear gradient (4–15%), the known difference between CG-14 and CG-16 served as internal quality control. Molecular masses of 14030.9 Da (CG-14) and 16639.2 Da (CG-16) with S.D. of 3.6 and 4.4% respectively were determined (Figure 1). When run lane-by-lane in the same gels, the electrophoretic mobility of CG-2 was consistently slightly lower than that of recombinant CG-14 in both gel systems (Figure 1). The difference did not reach the level of statistical significance (P=0.11). Two-dimensional gel electrophoresis could barely resolve a mixture of the two galectins. A slight disparity between the pI values was measured (Figure 2). Its size reflects the theoretical difference between the calculated pI values of 6.51 (CG-2) and 6.58 (CG-14). This accordance excludes the presence of a charge-altering post-translational modification in the recombinant products, as also previously noted for the galectin purified from 100 g of chicken intestine with a yield of 2.5±0.5 mg (n=6) [16]. If we now try to identify the nature of the intestinal lectin on the basis of this experiment, any conclusion would inevitably be ambiguous. We therefore performed tryptic ‘fingerprinting’ combined with MS peptide characterization. Because the positions of arginine/lysine residues are not identical in the sequences of both galectins (see Table 1, bottom), this method should allow us to definitively decide on the nature of the intestinal galectin.

Electrophoretic mobility of the three prototype CGs

Figure 1
Electrophoretic mobility of the three prototype CGs

Purified proteins were separated by PAGE under denaturing conditions in the presence of 2-mercaptoethanol in a 15% running gel (A) or a 4–15% linear gradient gel (B). Positions of relevant marker proteins are given in the upper panels. Molecular masses were determined by the plots given in the bottom panels. Insets show the quantitative relationship between relative migration distance and known molecular mass of the six standard proteins (β-galactosidase, 116 kDa; BSA, 66 kDa; ovalbumin, 45 kDa; carbonic anhydrase, 29 kDa; β-lactoglobulin, 18.4 kDa; lysozyme, 14.2 kDa).

Figure 1
Electrophoretic mobility of the three prototype CGs

Purified proteins were separated by PAGE under denaturing conditions in the presence of 2-mercaptoethanol in a 15% running gel (A) or a 4–15% linear gradient gel (B). Positions of relevant marker proteins are given in the upper panels. Molecular masses were determined by the plots given in the bottom panels. Insets show the quantitative relationship between relative migration distance and known molecular mass of the six standard proteins (β-galactosidase, 116 kDa; BSA, 66 kDa; ovalbumin, 45 kDa; carbonic anhydrase, 29 kDa; β-lactoglobulin, 18.4 kDa; lysozyme, 14.2 kDa).

pI values of the prototype CGs CG-2 and CG-14

Figure 2
pI values of the prototype CGs CG-2 and CG-14

The relevant section of a gel after two-dimensional gel electrophoresis is shown; experimental assignment of pI values was performed using human galectins of known properties. The given section shows the two spots for CG-2 and CG-14 together with those of CG-16 (pI 4.95) and human galectin-1 (pI 5.34), -2 (pI 5.93) and -7 (pI 7.00).

Figure 2
pI values of the prototype CGs CG-2 and CG-14

The relevant section of a gel after two-dimensional gel electrophoresis is shown; experimental assignment of pI values was performed using human galectins of known properties. The given section shows the two spots for CG-2 and CG-14 together with those of CG-16 (pI 4.95) and human galectin-1 (pI 5.34), -2 (pI 5.93) and -7 (pI 7.00).

Table 1
Characteristics of the galectin-derived peptides identified by MS

For peptide numbering, please see Supplementary Figure 1. Detected sequences for each of the two galectins from recombinant production are highlighted. The identified peptide sequences of the intestinal galectin are pointed out by underlining them in the sequences shown undermeath the Table, revealing a pattern identical with that of CG-2.

     Position in: Intestinal 
Peak no. m/z Modifications* Missed cleavages Database sequence rCG-14 rCG-2 galectin 
911.62 M+16 SFVMNLGK 30–37   
995.52 ETVFPFQK 75–82 − − 
1165.50 MEEWGTEQR 66–74 − − 
1181.47 M+16 MEEWGTEQR 66–74 − − 
1293.55 KMEEWGTEQR 65–74 − − 
1309.52 M+16 KMEEWGTEQR 65–74 − − 
1393.64 DSTHLGLHFNPR 38–49 − − 
2103.02 LGLSVFDYFDTHGDFTLR 112–129 − − 
2270.16 SFVMNLGKDSTHLGLHFNPR 30–49 −  
10 2286.11 M+16 SFVMNLGKDSTHLGLHFNPR 30–49 − − 
11 2691.27 LGLSVFDYFDTHGDFTLRSVSWE 112–134 − − 
12 3191.59 GAPIEITFSINPPSDLTVHLPGHQFSFPNR 83–111 − − 
13 1224.92 ISYLNILGGFK − 113–123 
14 1256.88 SSDLALHFNPR − 38–48 
15 1360.81 MFEMFNLDWK − 3–12 
16 1376.79 M+16 MFEMFNLDWK − 3–12 
17 1392.78 2×M+16 MFEMFNLDWK − 3–12 
18 1409.90 LPDGHEVEFPNR − 97–108 
19 1791.08 GHISEDAESFAINLGCK − 22–38 
20 2497.5 Internal S–S bond FNESVIVCNSLCSDNWQQEQR − 50–70 
21 2513.48 Internal S–S bond and W+16 FNESVIVCNSLCSDNWQQEQR − 50–70 
22 2529.45 Internal S–S bond and W+32 FNESVIVCNSLCSDNWQQEQR − 50–70 
     Position in: Intestinal 
Peak no. m/z Modifications* Missed cleavages Database sequence rCG-14 rCG-2 galectin 
911.62 M+16 SFVMNLGK 30–37   
995.52 ETVFPFQK 75–82 − − 
1165.50 MEEWGTEQR 66–74 − − 
1181.47 M+16 MEEWGTEQR 66–74 − − 
1293.55 KMEEWGTEQR 65–74 − − 
1309.52 M+16 KMEEWGTEQR 65–74 − − 
1393.64 DSTHLGLHFNPR 38–49 − − 
2103.02 LGLSVFDYFDTHGDFTLR 112–129 − − 
2270.16 SFVMNLGKDSTHLGLHFNPR 30–49 −  
10 2286.11 M+16 SFVMNLGKDSTHLGLHFNPR 30–49 − − 
11 2691.27 LGLSVFDYFDTHGDFTLRSVSWE 112–134 − − 
12 3191.59 GAPIEITFSINPPSDLTVHLPGHQFSFPNR 83–111 − − 
13 1224.92 ISYLNILGGFK − 113–123 
14 1256.88 SSDLALHFNPR − 38–48 
15 1360.81 MFEMFNLDWK − 3–12 
16 1376.79 M+16 MFEMFNLDWK − 3–12 
17 1392.78 2×M+16 MFEMFNLDWK − 3–12 
18 1409.90 LPDGHEVEFPNR − 97–108 
19 1791.08 GHISEDAESFAINLGCK − 22–38 
20 2497.5 Internal S–S bond FNESVIVCNSLCSDNWQQEQR − 50–70 
21 2513.48 Internal S–S bond and W+16 FNESVIVCNSLCSDNWQQEQR − 50–70 
22 2529.45 Internal S–S bond and W+32 FNESVIVCNSLCSDNWQQEQR − 50–70 
*

Modifications of methionine (M) or tryptophan (W) by single oxidation (+16) or double oxidation (+32), S–S bond formation between internal cysteine residues.

graphic

Indeed, the results obtained bear out this expectation: two clearly different profiles were recorded (Table 1; please see Supplementary Figure 1 at http://www.BiochemJ.org/bj/409/bj4090591add.htm, for spectra). Sequence coverage by the detected peptides, as illustrated by shading the respective sequence stretches in the bottom part of Table 1, was 67 or 63% respectively. The individual characteristics of these peptides are compiled in Table 1. Cases of oxidative modifications of methionine (M) and tryptophan (W) and of internal disulfide formation were encountered in the analysis as registered in Table 1. Using the proven diagnostic power of ‘fingerprinting’, the question on the nature of the intestinal lectin was answered: its mass spectrum was identical with that of recombinant CG-2, all detected peptides from the lectin purified from intestinal tissue having matching partners present in the sequence of CG-2 (Table 1; Supplementary Figure 1). Of note, the previously determined mass of 14969 Da for the intestinal protein [16] describes the N-terminus as acetylated alanine (theoretical value at 14970.04 Da relative to 14974.98 of CG-14 with acetylated serine). Except for a processed N-terminus, no other sequence modification is present, in line with the protein's properties in two-dimensional gel electrophoresis regardless of the source of the protein (recombinant production or intestine). Thus the prototype galectin CG-14 purified from embryonic skin and the intestinal galectin originate from two different genes. This finding is readily reconcilable with the immunological evidence for a difference between the skin and intestinal lectins, outlined in the Introduction section. It also defines the intestinal protein as a natural product of the gene previously detected by database mining [3]. The implied phylogenetic proximity to mammalian galectin-2 should be reflected in experimental data, but the respective comparison, at present confined to gel filtration, revealed the following difference: human galectin-2 behaved as a homodimer in gel filtration on a Superdex 75 HR 10/30 column [33], while elution properties of a monomer were reported for the protein from chicken intestine on Sephadex G-100 [6]. This apparent disparity between two closely related proteins merits thorough inspection.

Quaternary structure of CG-2 and cell binding features

We first turned to gel filtration, re-analysing CG-2 in an HPLC system. In order to avoid any carbohydrate-inhibitable interaction of a lectin with the resin, the analyses were also performed in the presence of 100 mM lactose. Homodimeric CG-16/human galectin-1 were added as internal standards. Using a loading concentration of 1.2 mg/ml (estimated eluted concentration of 0.12 mg/ml), the elution time for CG-2 was 29.4±0.1 min. This value corresponds to an apparent molecular mass of 16.9±0.1 kDa for a globular protein. For comparison, the homodimeric CG-16 and galectin-1 had elution times of 27.4±0.1 and 27.5±0.1 min respectively, corresponding to an apparent molecular mass of approx. 30 kDa. Further experiments with CG-2 concentrations of 0.5 and 2.5 mg/ml, to exclude an influence of concentration, confirmed the marked deviation of CG-2 properties from those of the homodimeric galectins (results not shown). All results obtained are in accord with the initial evidence from gel filtration on a Sephadex G-100 column (1.5 cm×30 cm) [6]. The authors of that report have suggested that the measured molecular mass of 14.0±1.7 kDa indicated a monomeric status. If this were the case, the two galectin-2-type proteins would not share quaternary structure. In order to resolve this critical issue, CG-2 was studied by analytical ultracentrifugation.

Sedimentation-equilibrium data obtained for CG-2 in solutions at loading concentrations between 0.1 and 1 mg/ml could always be fitted to a single ideal component with an Mw (weight-average molecular mass) of 29.0±0.85 kDa. The observed mass was virtually independent of rotor speed, indicating that the sample was homogeneous, and was not influenced by the presence of 0.1 M lactose. For comparison, apparent molecular masses of 30.6±1.0 and 27.3±1.5 kDa were obtained for CG-16 and human galectin-1 respectively when examined under the given conditions. Thus, in sedimentation-equilibrium analysis, in the range of concentrations tested, CG-2 behaved entirely as a dimer. Its shape was studied by sedimentation-velocity analysis.

These experiments at a CG-2 concentration of 0.5 mg/ml revealed a predominant peak with a sedimentation coefficient of 2.03±0.1 S (2.08 S for s020,w, expressed in terms of the standard solvent of water at 20 °C). Comparable results were obtained in the absence or presence of 0.1 M lactose. Respective analysis of the homodimeric CG-16 and human galectin-1 gave significantly higher sedimentation coefficients (2.6±0.1 and 2.5±0.1 S respectively) than the one measured for CG-2. An estimation of the hydrodynamic size of hydration of the proteins was enabled by using the Sednterp program. The partial specific volume and degree of hydration of the proteins were calculated from the amino acid composition. The resulting s020,w value of CG-2 was higher than the maximally possible sedimentation coefficient predicted for a particle with the mass of a monomer. On the other hand, the ratio of the experimental frictional coefficient (f) to the minimum frictional coefficient of a compact anhydrous sphere with the mass and volume of the CG-2 dimer (f0) was 1.63. For CG-16 and human galectin-1, frictional ratios were of the order of 1.3, indicating that CG-2 does not behave as a globular protein, but as, for a galectin, an unusually extended protein, under these conditions. Accordingly, approximation of the shape of CG-2 to prolate or oblate ellipsoids of revolution gave axial shape ratios of 7.7 and 9.0. These numbers are >2-fold higher than those estimated for CG-16 and human galectin-1. The apparent retardation in gel filtration might be explained by a weak interaction of CG-2 with the two types of matrixes used, irrespective of the presence of lactose and/or a preferential orientation of CG-2 to access the porous beads. Taken together, the sedimentation analysis proved that CG-2 is a homodimer with a peculiar hydrodynamic behaviour. Its cell-binding capacity was analysed next.

Compared with CG-16, a very potent cross-linker active at a minimal concentration of 10 ng/ml, CG-2 led to carbohydrate-inhibitable agglutination at 160 ng/ml, confirming its bivalency. We next tested binding to a human cell system of marked reactivity for galectins [27]. Concentration-dependent and carbohydrate-inhibitable binding was ascertained to validate the system (Figures 3A and 3B). Despite the marked contrast in hydrodynamic properties, a similarity was observed between data for CG-2 and CG-16, with CG-14 showing less reactivity (Figures 3B–3D). On the level of avian immune cells, however, CG-2 preferentially bound to B-cells, whereas CG-16 stained different lymphocyte populations uniformly [16]. These results strengthen the notion of distinct functional profiles for different cell types. In this case, gene organization and expression profiles of the three prototype galectins should have dissimilar properties.

Cell surface staining by the three prototype CGs

Figure 3
Cell surface staining by the three prototype CGs

Semi-logarithmic representation of fluorescent surface staining of p16INK4a-reconstituted human Capan-1 pancreatic carcinoma cells by biotinylated CGs. Concentration dependence is documented for CG-14 showing fluorescent staining after incubation with 2, 5, 10 and 20 μg/ml biotinylated lectin (A). Staining profiles in the absence (solid line) and presence of glycoinhibitors (dotted line) are given for CG-2 (B), CG-14 (C) and CG-16 (D) at a constant lectin concentration of 10 μg/ml. The characteristics of background staining when using the fluorescent indicator without prior incubation of cells with a biotinylated marker are illustrated as reference in each panel as shaded area. Quantitative data on percentage of positive cells (%) and mean channel fluorescence are inserted into each panel.

Figure 3
Cell surface staining by the three prototype CGs

Semi-logarithmic representation of fluorescent surface staining of p16INK4a-reconstituted human Capan-1 pancreatic carcinoma cells by biotinylated CGs. Concentration dependence is documented for CG-14 showing fluorescent staining after incubation with 2, 5, 10 and 20 μg/ml biotinylated lectin (A). Staining profiles in the absence (solid line) and presence of glycoinhibitors (dotted line) are given for CG-2 (B), CG-14 (C) and CG-16 (D) at a constant lectin concentration of 10 μg/ml. The characteristics of background staining when using the fluorescent indicator without prior incubation of cells with a biotinylated marker are illustrated as reference in each panel as shaded area. Quantitative data on percentage of positive cells (%) and mean channel fluorescence are inserted into each panel.

Profiling of CG-2 gene structure

The gene for CG-2 is found on chromosome 1, approx. 42 kb upstream of the CG-14 gene, but in the opposite orientation. A similar arrangement is seen in the cases of mammalian galectin-1 and -2. Remarkably, the coding sequence for CG-16 is located on chromosome 4. On the level of exons/introns, the CG-2 gene is organized into four exons, a common trait of prototype galectins. The lengths of the first and third exons, the latter one comprising the carbohydrate-recognition domain, are identical with or very close to those of human galectin-2 with 6 and 159 bp compared with 6 and 160 bp respectively (see Supplementary Figure 2 at http://www.BiochemJ.org/bj/409/bj4090591add.htm). CG-14 as well as CG-16 and human galectin-1 share the length of 9 bp in the first exon and have 172 bp in the third exon. Admittedly, the corresponding data on the second and fourth exons do not entirely follow this two-group pattern. Based on sequence comparison, the level of divergence justifies assignment of CG-2 and CG-14/CG-16 to two separate branches of a phylogenetic tree [3].

Availability of the genomic sequence afforded comparative promoter analysis. Separately combing through the proximal promoter region of the CG-2 gene with two search algorithms to spot putative binding sites uncovered a panel of hits (see Supplementary Figure 3 at http://www.BiochemJ.org/bj/409/bj4090591add.htm). In the same systematic manner, we processed the corresponding regions of gene sequences in the case of CG-14 and CG-16 (for a detailed listing of putative sites, please see Supplementary Table 1 at http://www.BiochemJ.org/bj/409/bj4090591add.htm). This first comparative analysis of promoter sequences for non-mammalian galectins came up with further support that distinct, functionally relevant properties were acquired by sequence divergence of the prototype galectins. Pronounced differences concern the unique presence of potential binding motifs for AP-2 (activator protein-2), p53, GATA-3 (GATA binding protein 3) and the common co-activators c-Rel and p300 (a transcriptional activator protein required to drive p53 expression) in the CG-2 promoter, as well as comparatively frequent occurrence of sites for the activating homeobox protein Cdx-1 (caudal-type homeobox 1) and the MZF-1 (myeloid zinc finger-1) protein. In contrast, CG-2's proximal promoter region lacks the presence of site(s) potentially reactive with the octamer factor POU2F1 (POU class 2 homeobox 1)/Oct-1 (octamer-binding transcription factor-1). A similar picture emerges in the case of CG-14, its promoter region uniquely harbouring motifs for the common transactivator CP2, Oct-6, v-Maf, the co-activator YY1 (Yin and Yang 1), a glucocorticoid-responsive element and a barbiturate-inducible box (Supplementary Table 1).

Expression profiling of CG-2

To preclude false-negative results, we started at the highest level of sensitivity with RT–PCR analysis using specific primer sets. As internal controls, samples from skin with abundant expression of CG-14 and from intestine, a rich source for CG-2, were included. Indeed, strong signals in RT–PCR gels were recorded in these cases at expected positions (see Supplementary Figure 4 at http://www.BiochemJ.org/bj/409/bj4090591add.htm). Presence of galectin-type-specific mRNA, at least in minute quantities, was also detected in extracts of other organs after running 35 amplification cycles (Supplementary Figure 4). To separate spurious expression from cases with robust protein production, tissue extracts were processed by Western blotting. Because systematic ELISAs using the three IgG fractions from immunized rabbits revealed cross-reactivity, e.g. anti-CG-14 binding CG-16 and anti-CG-2 reacting both with CG-14 and CG-16, albeit at a low level, affinity depletion with resin-immobilized galectins had to be performed. Its success was rigorously controlled, using complete removal as the criterion. As internal controls in Western blots, the processed antibody preparations were routinely confronted with purified galectins and, additionally, we included a loading control by visualizing actin (Figure 4). As Figure 4 shows, the presence of CG-2 was not observed in skin and lungs, a qualitative difference from CG-14. On the other hand, jejunum proved to be a rich source of CG-2. The two structurally closely related galectins are thus without doubt immunologically distinct, and their expression profiles also reveal quantitative and qualitative differences.

Expression profiling of the prototype CGs CG-2 and CG-14 by Western blotting

Figure 4
Expression profiling of the prototype CGs CG-2 and CG-14 by Western blotting

Lectin presence was determined in tissue extracts (30 μg of protein per lane) using non-cross-reactive IgG fractions for CG-2 (A; 0.5 μg/ml) and CG-14 (B; 0.5 μg/ml). Specificity controls were run with 25 ng of purified lectin per lane, and actin presence was probed as internal loading control for each type of tissue extract. Positions of molecular mass markers (kDa) are indicated.

Figure 4
Expression profiling of the prototype CGs CG-2 and CG-14 by Western blotting

Lectin presence was determined in tissue extracts (30 μg of protein per lane) using non-cross-reactive IgG fractions for CG-2 (A; 0.5 μg/ml) and CG-14 (B; 0.5 μg/ml). Specificity controls were run with 25 ng of purified lectin per lane, and actin presence was probed as internal loading control for each type of tissue extract. Positions of molecular mass markers (kDa) are indicated.

In order to take our analysis from the level of extracts to that of individual cell types, fixed sections of organs of adult animals were immunohistochemically processed. After first defining the optimal conditions for tissue preservation and staining quality, the ensuing antibody application resulted in readily assignable positivity (Table 2). Without exception, the results of immunohistochemistry and Western blotting are in full accord. As examples for cell-type specificity, lungs and skin presented strong signals in the respiratory epithelium and in the epidermis, with already documented regional separation for CG-14 (Table 2). The same marked cell-type specificity was seen for CG-2 (Table 2; Figure 5). Epithelial lining of villi in the gut and collecting ducts in kidney were strongly positive. Examining the staining of cells more closely, cytoplasmic and nuclear lectin localizations were discerned (Table 2; Figure 5).

Table 2
Immunohistochemical profiling of the presence of the prototype CGs CG-2, CG-14 and CG-16 in various organs of adult animals

The intensity of staining is grouped into categories: –, no staining; +, weak staining; ++, medium staining; +++, strong staining.

 Staining intensity 
Organ Galectin… CG-2 CG-14 CG-16 
Larynx    
 Respiratory epithelium − − ++* 
 Lamina propria mucosae − − 
    
Trachea    
 Respiratory epithelium − − ++* 
 Lamina propria mucosae − ++ − 
    
Lung    
 Respiratory epithelium − +++* − 
 Connective tissue − − 
    
Oesophagus    
 Lamina propria mucosae − ++ − 
    
Gut    
 Epithelial lining of villi and intestinal +++* − − 
 glands    
 Lamina propria mucosae − − 
    
Liver    
 Hepatocytes (parenchyma) − − +++* 
    
Kidney    
 Epithelium    
  Collecting ducts (medulla) ++† − − 
  Proximal/distal tubules − − +++* 
  (MTN I, MTN II, RTN)‡    
    
Skin    
 Epidermis    
  Stratum corneum − − − 
  Stratum intermedium − +++† − 
  Stratum basalis − +* − 
 Dermis − ++ − 
 Subcutis − − 
 Staining intensity 
Organ Galectin… CG-2 CG-14 CG-16 
Larynx    
 Respiratory epithelium − − ++* 
 Lamina propria mucosae − − 
    
Trachea    
 Respiratory epithelium − − ++* 
 Lamina propria mucosae − ++ − 
    
Lung    
 Respiratory epithelium − +++* − 
 Connective tissue − − 
    
Oesophagus    
 Lamina propria mucosae − ++ − 
    
Gut    
 Epithelial lining of villi and intestinal +++* − − 
 glands    
 Lamina propria mucosae − − 
    
Liver    
 Hepatocytes (parenchyma) − − +++* 
    
Kidney    
 Epithelium    
  Collecting ducts (medulla) ++† − − 
  Proximal/distal tubules − − +++* 
  (MTN I, MTN II, RTN)‡    
    
Skin    
 Epidermis    
  Stratum corneum − − − 
  Stratum intermedium − +++† − 
  Stratum basalis − +* − 
 Dermis − ++ − 
 Subcutis − − 
*

Only cytoplasmic.

Cytoplasmic and nuclear.

MNT-I, mammalian-type nephron I (juxtamedullar); MNT-II, mammalian-type nephron II (midcortical); RTN, reptilian-type nephron (superficial).

Localization of the prototype CG CG-2 in tissue sections by immunohistochemistry and the comparison to CG-14 staining

Figure 5
Localization of the prototype CG CG-2 in tissue sections by immunohistochemistry and the comparison to CG-14 staining

Microphotographs of a cross-section through a medullary cone of a kidney of a 6-month-old chicken at two levels of magnification (AD). Control without an incubation step using antiserum ascertained lack of antigen-independent staining (A). Application of the non-cross-reactive antibody preparation at 1 μg/ml revealed the presence of CG-2 in epithelial cells of the collecting ducts, whereas sections of Henle's loop of mammalian type nephrons I were free of staining (B). On the cellular level, epithelial cells of collecting ducts (CD) were present with signal in cytoplasm and nuclei as well as at the membrane, in stark contrast with the negative cells in Henle's loops (LH) (C). The application of anti-CG-14 antibody under identical conditions yielded no signal (D). Corresponding processing of a cross-section through jejunum of a 6-month-old chicken resulted in microphotographs, which are also shown at two levels of magnification (EJ). Lectin presence (CG-2) was confined to epithelial cells of villi and intestinal glands (E). Intense cytoplasmic staining characterizes the epithelial cells of villi (F) and intestinal glands (G). In contrast, CG-14 was localized in the connective tissue (lamina propria mucosae), as shown in a serial section to allow direct comparison (HJ). Scale bars: 20 μm (A, C, D, F, G, I, J), 50 μm (B) and 250 μm (E, H).

Figure 5
Localization of the prototype CG CG-2 in tissue sections by immunohistochemistry and the comparison to CG-14 staining

Microphotographs of a cross-section through a medullary cone of a kidney of a 6-month-old chicken at two levels of magnification (AD). Control without an incubation step using antiserum ascertained lack of antigen-independent staining (A). Application of the non-cross-reactive antibody preparation at 1 μg/ml revealed the presence of CG-2 in epithelial cells of the collecting ducts, whereas sections of Henle's loop of mammalian type nephrons I were free of staining (B). On the cellular level, epithelial cells of collecting ducts (CD) were present with signal in cytoplasm and nuclei as well as at the membrane, in stark contrast with the negative cells in Henle's loops (LH) (C). The application of anti-CG-14 antibody under identical conditions yielded no signal (D). Corresponding processing of a cross-section through jejunum of a 6-month-old chicken resulted in microphotographs, which are also shown at two levels of magnification (EJ). Lectin presence (CG-2) was confined to epithelial cells of villi and intestinal glands (E). Intense cytoplasmic staining characterizes the epithelial cells of villi (F) and intestinal glands (G). In contrast, CG-14 was localized in the connective tissue (lamina propria mucosae), as shown in a serial section to allow direct comparison (HJ). Scale bars: 20 μm (A, C, D, F, G, I, J), 50 μm (B) and 250 μm (E, H).

Encouraged by these discriminatory results, we extended the scope of our analysis to CG-16. Again, affinity depletion of the IgG fraction was required due to its reactivity with CG-2. Even in cases of simultaneous presence of two prototype galectins in an organ such as kidney, the individual patterns of localization were qualitatively different when examined by immunohistochemistry (Table 2). This compiled evidence identified no overlap between prototype CGs on the level of cell-type expression in organs of adult animals. Of course, the presence of a fourth member in this galectin group with features deviating from this picture would quash the given inference. To address this issue, we carried out extensive database mining on the current version of the chicken genome using a suitably tailored consensus sequence. No evidence for the existence of a further gene for a prototype galectin was unearthed. This result solidifies the notion that CG-2 together with CG-14 and CG-16 and no further protein comprise the group of prototype CGs.

Implications for nomenclature and perspectives

It is now timely to suggest a nomenclature that reflects the current status of our understanding of the phylogenetic relationships as given in [3], and not electrophoretic mobilities: CG-16/-14 are therefore suggested to be called CG-1A/B to signal their proximity to galectin-1. As predicted by this concept, chemical mapping using synthetic derivatives of lactose had revealed closer similarity of CG-16 (CG-1A) to human galectin-1 than to the intestinal lectin CG-2 [34]. Thus, recalling the broad range of effector functions of galectins via ligand binding [1], diversity in lectin evolution has the potential to expedite establishment of a finely tuned system to turn sugar coding into specific biosignalling, and initial work on oligosaccharides and natural glycoproteins to map binding profiles of prototype CGs already tracked down differences [3539]. The underlying structural basis will have to be defined by extending crystallographic analysis beyond CG-16 [40].

On the level of gene regulation, our analysis was guided by ‘the hypothesis that families of electrolectin genes could be under the control of different promoters’ [10]. This first detailed analysis of putative sites for transcription factors in genes of avian galectins pinpointed a series of hits. Verifying the given prediction on the molecular level, a characteristic pattern is present in each proximal promoter region. How to relate the complex combinations of motifs to the actual expression profiles in organs of adult animals and also to developmental regulation will have to be addressed by functional dissection. At this stage, we took care by internal controls to ascertain the validity of the expression profiles, and previous results, e.g. on pectoral muscle [41], were in full accord with the expression profiling presented, in this case excluding a false-positive result. On the level of organs, zonally distinct galectin appearance determined by immunohistochemistry in organs of adult animals is a clear indication for acquisition of non-overlapping functions of the three prototype proteins in the different cell types including nuclear and/or cytoplasmic presence. The case of epithelial kidney cells of proximal tubules with their positivity for both CG-2 and CG-16, which is limited to days 12–14 of incubation located at distinct subcellular sites, might even be interpreted along this line [42]. This finding and the need for further crystallographic data give further research on the three prototype CGs a clear direction.

We are grateful to Professor Dr K.-i. Kasai (Department of Biological Chemistry, Faculty of Pharmaceutical Sciences, Teikyo University, Kanagawa, Japan) for his expert advice and approval of the suggested nomenclature, to Dr B. Friday and Dr S. Namirha for inspiring discussions, to the reviewers of our manuscript for their valuable input, to A. Helfrich, B. Hofer, L. Mantel and S. Fiedler for skilful technical assistance, to the Spanish Dirección General de Investigación (BQU2003-03550-C03-03 and BFU2006-10288/BMC), to the research initiative LMUexcellent, to the Verein zur Förderung des biologisch-technologischen Fortschritts in der Medizin e.V. and to a European Community Marie Curie Research Training Network grant (contract no. MRTN-CT-2005-019561) for generous financial support.

Abbreviations

     
  • CG

    chicken galectin

  •  
  • RT

    reverse transcriptase

References

References
1
Gabius
H.-J.
Cell surface glycans: the why and how of their functionality as biochemical signals in lectin-mediated information transfer
Crit. Rev. Immunol.
2006
, vol. 
26
 (pg. 
43
-
79
)
2
Gabius
H.-J.
Animal lectins
Eur. J. Biochem.
1997
, vol. 
243
 (pg. 
543
-
576
)
3
Houzelstein
D.
Gonçalves
I. R.
Fadden
A. J.
Sidhu
S. S.
Cooper
D. N. W.
Drickamer
K.
Leffler
H.
Poirier
F.
Phylogenetic analysis of the vertebrate galectin family
Mol. Biol. Evol.
2004
, vol. 
21
 (pg. 
1177
-
1187
)
4
Den
H.
Malinzak
D. A.
Isolation and properties of β-D-galactoside-specific lectin from chick embryo thigh muscle
J. Biol. Chem.
1977
, vol. 
252
 (pg. 
5444
-
5448
)
5
Nowak
T. P.
Kobiler
D.
Roel
L. E.
Barondes
S. H.
Developmentally regulated lectin from embryonic chick pectoral muscle. Purification by affinity chromatography
J. Biol. Chem.
1977
, vol. 
252
 (pg. 
6026
-
6030
)
6
Beyer
E. C.
Zweig
S. E.
Barondes
S. H.
Two lactose binding lectins from chicken tissues. Purified lectin from intestine is different from those in liver and muscle
J. Biol. Chem.
1980
, vol. 
255
 (pg. 
4236
-
4239
)
7
Pitts
M. J.
Yang
C. H.
Isolation of a developmentally regulated lectin from chick embryo
Biochem. Biophys. Res. Commun.
1980
, vol. 
95
 (pg. 
750
-
757
)
8
Beyer
E. C.
Barondes
S. H.
Quantitation of two endogenous lactose-inhibitable lectins in embryonic and adult chicken tissues
J. Cell Biol.
1982
, vol. 
92
 (pg. 
23
-
27
)
9
Oda
Y.
Kasai
K.-I.
Purification and characterization of β-galactoside-binding lectin from chick embryonic skin
Biochim. Biophys. Acta
1983
, vol. 
761
 (pg. 
237
-
245
)
10
Levi
G.
Teichberg
V. I.
Isolation and characterization of chicken thymic electrolectin
Biochem. J.
1985
, vol. 
226
 (pg. 
379
-
384
)
11
Hirabayashi
J.
Kawasaki
H.
Suzuki
K.
Kasai
K.-I.
Complete amino acid sequence of 14 kDa β-galactoside-binding lectin of chick embryo
J. Biochem.
1987
, vol. 
101
 (pg. 
775
-
787
)
12
Sakakura
Y.
Hirabayashi
J.
Oda
Y.
Ohyama
Y.
Kasai
K.-I.
Structure of chicken 16-kDa β-galactoside-binding lectin. Complete amino acid sequence, cloning of cDNA, and production of recombinant lectin
J. Biol. Chem.
1990
, vol. 
265
 (pg. 
21573
-
21579
)
13
Cooper
D. N. W.
Galectinomics: finding themes in complexity
Biochim. Biophys. Acta
2002
, vol. 
1572
 (pg. 
209
-
231
)
14
Barak Briles
E.
Gregory
W.
Fletcher
P.
Kornfeld
S.
Vertebrate lectins. Comparison of properties of β-galactoside-binding lectins from tissues of calf and chicken
J. Cell Biol.
1979
, vol. 
81
 (pg. 
528
-
537
)
15
Castagna
L. F.
Landa
C. A.
Isolation and characterization of a soluble lactose-binding lectin from postnatal chicken retina
J. Neurosci. Res.
1994
, vol. 
37
 (pg. 
750
-
758
)
16
Schneller
M.
André
S.
Cihak
J.
Kaltner
H.
Merkle
H.
Rademaker
G. J.
Haverkamp
J.
Thomas-Oates
J. E.
Lösch
U.
Gabius
H.-J.
Differential binding of two chicken β-galactoside-specific lectins to homologous lymphocyte subpopulations and evidence for inhibitor activity of the dimeric lectin on stimulated T cells
Cell. Immunol.
1995
, vol. 
166
 (pg. 
35
-
43
)
17
Lipsick
J. S.
Beyer
E. C.
Barondes
S. H.
Kaplan
N. O.
Lectins from chicken tissues are mitogenic for Thy-1 negative murine spleen cells
Biochem. Biophys. Res. Commun.
1980
, vol. 
97
 (pg. 
56
-
61
)
18
Pitts
M. J.
Yang
D. C. H.
Mitogenicity and binding properties of β-galactoside-binding lectin from chick-embryo kidney
Biochem. J.
1981
, vol. 
195
 (pg. 
435
-
439
)
19
Kakita
H.
Nakamura
K.
Kato
Y.
Oda
Y.
Shimura
K.
Kasai
K.-I.
High-performance affinity chromatography of a chick lectin on an adsorbent based on hydrophilic polymer gel
J. Chromatogr.
1991
, vol. 
543
 (pg. 
315
-
326
)
20
Kopitz
J.
Russwurm
R.
Kaltner
H.
André
S.
Dotti
C. G.
Gabius
H.-J.
Abad-Rodríguez
J.
Hippocampal neurons and recombinant galectins as tools for systematic carbohydrate structure–function studies in neuronal differentiation
Dev. Brain Res.
2004
, vol. 
153
 (pg. 
189
-
196
)
21
André
S.
Kaltner
H.
Lensch
M.
Russwurm
R.
Siebert
H.-C.
Fallsehr
C.
Tajkhorshid
E.
Heck
A. J. R.
von Knebel Doeberitz
M.
Gabius
H.-J.
Kopitz
J.
Determination of structural and functional overlap/divergence of five proto-type galectins by analysis of the growth-regulatory interaction with ganglioside GM1 in silico and in vitro on human neuroblastoma cells
Int. J. Cancer
2005
, vol. 
114
 (pg. 
46
-
57
)
22
Hirabayashi
J.
Oda
Y.
Oohara
T.
Yamagata
T.
Kasai
K.-I.
Immunological cross-reactivity between β-galactoside-binding lectins of vertebrates. Possible relationship of antigenicity to oligomeric structure
Biochim. Biophys. Acta
1987
, vol. 
916
 (pg. 
321
-
327
)
23
Lips
K. S.
Kaltner
H.
Reuter
G.
Stierstorfer
B.
Sinowatz
F.
Gabius
H.-J.
Correspondence of gradual developmental increases of expression of galectin-reactive glycoconjugates with alterations of the total contents of the two differentially regulated galectins in chicken intestine and liver as indication for overlapping functions
Histol. Histopathol.
1999
, vol. 
14
 (pg. 
743
-
760
)
24
Didier
E.
Zalik
S. E.
Didier
P.
Ledsham
I. M.
Bayle
D.
Different immunoreactivities of anti-soluble lactose lectin antisera to tissues from early chick embryos: a histochemical study
Histochemistry
1993
, vol. 
100
 (pg. 
485
-
493
)
25
Guay
C. K.
Zalik
S. E.
Expression of the 14 kDa and 16 kDa galactoside-binding lectins during differentiation of the chick yolk sac
Roux's Arch. Dev. Biol.
1994
, vol. 
204
 (pg. 
126
-
140
)
26
Purkrábková
T.
Smetana
K.
Jr
Dvořánková
B.
Holíková
Z.
Böck
C.
Lensch
M.
André
S.
Pytlík
R.
Liu
F.-T.
Klima
J.
, et al. 
New aspects of galectin functionality in nuclei of cultured bone marrow stromal and epidermal cells: biotinylated galectins as tool to detect specific binding sites
Biol. Cell
2003
, vol. 
95
 (pg. 
535
-
545
)
27
André
S.
Sanchez-Ruderisch
H.
Nakagawa
H.
Buchholz
M.
Kopitz
J.
Forberich
P.
Kemmner
W.
Böck
C.
Deguchi
K.
Detjen
K. M.
, et al. 
Tumor suppressor p16INK4a: modulator of glycomic profile and galectin-1 expression to increase susceptibility to carbohydrate-dependent induction of anoikis in pancreatic carcinoma cells
FEBS J.
2007
, vol. 
274
 (pg. 
3233
-
3256
)
28
André
S.
Kaltner
H.
Furuike
T.
Nishimura
S.-I.
Gabius
H.-J.
Persubstituted cyclodextrin-based glycoclusters as inhibitors of protein–carbohydrate recognition using purified plant and mammalian lectins and wild-type and lectin-gene-transfected tumor cells as targets
Bioconjugate Chem.
2004
, vol. 
15
 (pg. 
87
-
98
)
29
Gabius
H.-J.
Engelhardt
R.
Rehm
S.
Cramer
F.
Biochemical characterization of endogenous carbohydrate-binding proteins from spontaneous murine rhabdomyosarcoma, mammary adenocarcinoma, and ovarian teratoma
J. Natl. Cancer Inst.
1984
, vol. 
73
 (pg. 
1349
-
1357
)
30
Jiménez
M.
Sáiz
J. L.
André
S.
Gabius
H.-J.
Solís
D.
Monomer/dimer equilibrium of the AB-type lectin from mistletoe enables combination of toxin/agglutinin activities in one protein: analysis of native and citraconylated proteins by ultracentrifugation/gel filtration and cell biological consequences of dimer destabilization
Glycobiology
2005
, vol. 
15
 (pg. 
1386
-
1395
)
31
Lensch
M.
Lohr
M.
Russwurm
R.
Vidal
M.
Kaltner
H.
André
S.
Gabius
H.-J.
Unique sequence and expression profiles of rat galectins-5 and -9 as a result of species-specific gene divergence
Int. J. Biochem. Cell Biol.
2006
, vol. 
38
 (pg. 
1741
-
1758
)
32
Lohr
M.
Lensch
M.
André
S.
Kaltner
H.
Siebert
H.-C.
Smetana
K.
Jr
Sinowatz
F.
Gabius
H.-J.
Murine homodimeric adhesion/growth-regulatory galectins-1, -2 and -7: comparative profiling of gene/promoter sequences by database mining, of expression by RT-PCR/immunohistochemistry and of contact sites for carbohydrate ligands by computational chemistry
Folia Biol.
2007
, vol. 
53
 (pg. 
109
-
128
)
33
Gitt
M. A.
Massa
S. M.
Leffler
H.
Barondes
S. H.
Isolation and expression of a gene encoding L-14-II, a new human soluble lactose-binding lectin
J. Biol. Chem.
1992
, vol. 
267
 (pg. 
10601
-
10606
)
34
Solís
D.
Romero
A.
Kaltner
H.
Gabius
H.-J.
Díaz-Mauriño
T.
Different architecture of the combining site of the two chicken galectins revealed by chemical mapping studies with synthetic ligand derivatives
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
12744
-
12748
)
35
Wu
A. M.
Wu
J. H.
Tsai
M.-S.
Kaltner
H.
Gabius
H.-J.
Carbohydrate specificity of a galectin from chicken liver (CG-16)
Biochem. J.
2001
, vol. 
358
 (pg. 
529
-
538
)
36
Hirabayashi
J.
Hashidate
T.
Arata
Y.
Nishi
N.
Nakamura
T.
Hirashima
M.
Urashima
T.
Oka
T.
Futai
M.
Müller
W. E. G.
, et al. 
Oligosaccharide specificity of galectins: a search by frontal affinity chromatography
Biochim. Biophys. Acta
2002
, vol. 
1572
 (pg. 
232
-
254
)
37
Gabius
H.-J.
Siebert
H.-C.
André
S.
Jiménez-Barbero
J.
Rüdiger
H.
Chemical biology of the sugar code
ChemBioChem
2004
, vol. 
5
 (pg. 
740
-
764
)
38
Villalobo
A.
Nogales-González
A.
Gabius
H.-J.
A guide to signaling pathways connecting protein–glycan interaction with the emerging versatile effector functionality of mammalian lectins
Trends Glycosci. Glycotechnol.
2006
, vol. 
18
 (pg. 
1
-
37
)
39
Wu
A. M.
Singh
T.
Liu
J.-H.
Krzeminski
M.
Russwurm
R.
Siebert
H.-C.
Bonvin
A. M. J. J.
André
S.
Gabius
H.-J.
Activity–structure correlations in divergent lectin evolution: fine specificity of chicken galectin CG-14 and computational analysis of flexible ligand docking for CG-14 and the closely related CG-16
Glycobiology
2007
, vol. 
17
 (pg. 
165
-
184
)
40
Varela
P. F.
Solís
D.
Díaz-Mauriño
T.
Kaltner
H.
Gabius
H.-J.
Romero
A.
The 2.15 Å crystal structure of CG-16, the developmentally regulated homodimeric chicken galectin
J. Mol. Biol.
1999
, vol. 
294
 (pg. 
537
-
549
)
41
Beyer
E. C.
Tokuyasu
K. T.
Barondes
S. H.
Localization of an endogenous lectin in chicken liver, intestine, and pancreas
J. Cell Biol.
1979
, vol. 
82
 (pg. 
565
-
571
)
42
Stierstorfer
B.
Kaltner
H.
Neumüller
C.
Sinowatz
F.
Gabius
H.-J.
Temporal and spatial regulation of expression of two galectins during kidney development of the chicken
Histochem. J.
2000
, vol. 
32
 (pg. 
325
-
336
)

Supplementary data