BIOCHEMICAL AND STRUCTURAL ANALYSIS OF HELIX POMATIA

AGGLUTININ (HPA): A HEXAMERIC LECTIN WITH A NOVEL FOLD*

Jean-Frederic Sanchez‡1, Julien Lescar‡§1, Valérie Chazalet‡, Aymeric Audfray‡, Jean Gagnon¶, Richard Alvarez║, Christelle Breton‡, Anne Imberty‡ & Edward P. Mitchell‡‡ From ‡Centre de Recherches sur les Macromolécules Végétales, CNRS (affiliated with Université Joseph Fourier), Grenoble, France, §School of Biological Sciences, Nanyang Technological University, Singapore, ¶CNRS-FRE2685, Université Joseph Fourier, La Tronche, France, ║Dept. Biochem/Mol. Biol. University of Oklahoma, USA, ‡‡E.S.R.F.

Experiments Division, Grenoble, France Running Title: Crystal structure of Helix pomatia lectin

Address correspondence to Anne Imberty, CERMAV-CNRS BP53, 38041 Grenoble cedex 09, France ; Fax: +33-476547203; E-Mail: imberty@cermav.cnrs.fr or Edward P. Mitchel E.S.R.F. Experiments Division, BP 220, F-38043, Grenoble Cedex, France; Fax : ++33-476882542; E-Mail: mitchell@esrf.fr Helix pomatia agglutinin (HPA) is a N-acetylgalactosamine (GalNAc) binding lectin found in the albumen gland of the roman snail. As a constituent of perivitelline fluid, HPA protects fertilized eggs from bacteria and is part of the innate immunity system of the snail. The peptide sequence deduced from gene cloning demonstrates that HPA belongs to a family of carbohydrate-binding proteins recently identified in several invertebrates. This domain is also present in discoidin from the slime mould Dictyostelium discoideum. Investigation of the lectin specificity was performed with the use of glycan arrays, demonstrating that several GalNAc containing oligosaccharides are bound, and rationalizing the use of this lectin as a cancer marker. Titration microcalorimetry performed on the interaction between HPA and GalNAc indicates an affinity in the 10-4M range with an enthalpy driven binding mechanism. The crystal structure of HPA demonstrates the occurrence of a new β-sandwich lectin fold. The hexameric quaternary state was never observed previously for a lectin. The high resolution structure complex of HPA with GalNAc characterizes a new carbohydrate binding site and rationalizes the observed preference for αGalNAc containing oligosaccharides. Because of their ability to recognize complex carbohydrates on cell surfaces with high specificity, lectins are proteins that play important roles in the social life of cells. A growing repertoire of lectins has been identified

in invertebrates (1), where these molecules are involved in self/non-self recognition (2). Examples include the recognition of “sister” cells as part of the aggregation mechanisms in primitive organisms (e.g. slime molds (3), sponges and corals (4)), the specific binding of polysaccharide coated pathogenic bacteria in the innate immunity system of invertebrates (5, 6) and the mediation of symbiosis, for example between coral and their symbiotic algae (7).

In the edible roman snail, a lectin named Helix pomatia agglutinin (HPA)2 is secreted by the albumen gland that produces the perivitelline fluid, composed of protein and polysaccharide complexes that coat each fertilized egg (8). This lectin has the property to aggregate bacteria such as group C streptococci (9) and Listeria monocytogenes (10) and is also able to bind to the surface of the herpes virus (11).

Despite the potential antimicrobial activity of HPA, biochemical and structural data are scarce. An agglutination activity was identified in an extract of snail albumen gland at the end of 19th century by its ability to aggregate red cells and human milk particles (12). Later on, the purified lectin was shown to agglutinate blood group A red cells (13), but not those of blood group B or O (14). The binding preference was established to be Forssman antigen (αGalNAc1-3GalNAc-R) > blood group A substance (αGalNAc1-3[αFuc1-2]Gal] > Tn antigen (αGalNAc-Ser/Thr) > GalNAc > GlcNAc, therefore confirming the specificity for

1

http://www.jbc.org/cgi/doi/10.1074/jbc.M603452200The latest version is at JBC Papers in Press. Published on May 16, 2006 as Manuscript M603452200

Copyright 2006 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

terminal α-N-acetyl-D-galactosamine (αGalNAc) residues (15). HPA was identified as a hexamer with a total MW of 79 kDa (14). Further characterization led to the conclusion that the snail hemagglutinin consists of six identical polypeptide subunits, each containing one intra-chain disulfide bond and a carbohydrate binding site. In fact, two chains associate by a disulfide bridge and the hexamer therefore consists of a non-covalent trimer of covalent dimers (16). HPA is a glycoprotein with about 7% carbohydrate content and, since this glycoprotein could be fractionated into at least 12 components by isoelectric focusing (17), the presence of different glycoforms is probable.

EXPERIMENTAL PROCEDURES Materials ⎯ Helix pomatia agglutinin (HPA) was purchased from Sigma and Vector (Burlingame, CA). Oligonucleotides for PCR were purchased from MWG SA. Biotech (Courtaboeuf, France). One Helix pomatia snail was kindly donated by the snail farm, Escargot du Pont Chardon (Chevrieres, France). Peptide sequencing and mass spectrometry ⎯ CNBr fragments of HPA were prepared by addition of a 10-fold by weight excess of CNBr to the protein (10 mg/ml) in 70% HCOOH, and the mixture was kept in the dark for 20 hr at 4 °C. The mixture was then diluted 1:10 with water and freeze dried. The digest was dissolved in 6 M GuHCl pH 2.5 and separated by HPLC on a C4 column (2.1 x 100 mm) using a gradient of 0.1% TFA and 70% CH3CN containing 0.085% TFA. Amino acid sequence determination of collected fractions was performed using a Procise Sequencer model 492 (Applied Biosystems). Phenylthiohydantoin amino acid derivatives were identified and quantified by HPLC analysis on-line as recommended by the manufacturer. MALDI-TOF spectra were recorded on a Bruker Autoflex (CERMAV) and Applied Bioysystems Voyager 125 (IBS). Reverse Transcription PCR and cDNA Cloning of Helix pomatia Agglutinin ⎯ PCR was used to amplify the coding region of HPA using degenerated primers that were designed with regard to the N-terminal sequence of the mature protein as deduced from the high resolution crystal structure of the HPA/GalNAc complex. Total RNA was obtained by grinding one

albumen gland in liquid nitrogen followed by an extraction using TRI reagent (Sigma Chemical Co.). cDNAs were synthesized using rapid amplification of cDNA ends (3’RACE kit, Invitrogen) with the following primers: pr-hpa (sense 5’-TGYGGNAAYGAYGCNGGNTGG-3') which codes for the peptide sequence CGNDAGW and UAP (Universal Amplification Primer, Invitrogen). PCR was performed using cDNA as the template and Taq DNA polymerase (Promega) with the following temperature profile: 4 min at 95°C first and then 45 s at 95°C, 1 min at 45°C, 5 min at 72°C, for 39 cycles. These conditions permitted the amplification of a fragment of 650 bp. The purified PCR fragment was cloned into the pGEM-T easy vector (Promega). Five independent clones were sequenced on both strands (Genome Express, Meylan, France). The 5’ end of the sequence was obtained using the 5’RACE kit from Invitrogen. cDNA was synthesized using a HPA gene-specific antisense oligonucleotide (GSP1-HPA: 5’-TCTATGACAGTGCGAAGACG-3’) and PCR was then performed using this cDNA preparation as the template, Taq DNA polymerase and the following primers: AAP (Abridged Anchor Primer, Invitrogen) and a second HPA gene specific primer GSP2-HPA (antisense 5’-TGAGAATTCCCTGATCTCTCTCGAAAGTAAG-3’). PCR was performed with the following cycling temperature profile: 45 s at 94°C, 1 min at 50°C, 2min at 72°C, for 35 cycles. The purified PCR fragment was ligated in the pGEM-T easy vector (Promega). Six independent clones were sequenced on both strands. Crystallization and Data Collections ⎯ Crystallization trials were performed with Hampton crystallization screens I and II (Hampton Research, Laguna Niguel, CA). After optimization, crystals were obtained using the hanging drop method with precipitation solution containing lithium sulfate (2 M) and ammonium sulfate (3.5 M) in sodium citrate buffer (1.5 M, pH 6.5). Drops were made of an equal volume of reservoir solution and protein solution at 10 mg/ml. For the co-crystallization assays, the ligand was added to the protein solution with a 3-fold excess in molarity.

Crystals were cryo-cooled at 100 K, after soaking them for as the shortest possible time in precipitant solution with 25% PEG 3000

2

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

and 25% glycerol added. Data images were recorded on an ADSC Q4R CCD detector (Quantum Corp, Il, USA) at beamline ID14-2 at the ESRF (Grenoble, France). Diffraction images were processed using MOSFLM (18) and scaled and truncated to structure factors using the CCP4 (19) programs SCALA and TRUNCATE. Data processing statistics are presented in Table 1. The presence of zinc in one batch of HPA was confirmed using a XANES spectrum on beamline ID14-4 at the ESRF.

Structure Determinations ⎯ The crystal structure of the HPA/GalNAc complex was determined using the SAD technique with highly redundant data from the crystals of the HPA/GalNAc/Zn complex to 1.7 Å resolution. Harker sections of the anomalous difference Patterson map showed peaks corresponding to one zinc atom per monomer in the asymmetric unit. The location of the zinc ion was found with HySS (20) and experimental phases were calculated and the position of the zinc refined with autoSHARP (21). A first model was built automatically with Arp/Warp (22) and used to prepare primers for cloning of the hpa gene and then completed using the obtained HPA sequence. Cycles of refinement with REFMAC (26)and manual rebuilding with O (24) resulted in an initial model of the lectin.

This model was used as the search probe for molecular replacement for the native protein and the GalNAc/Zn complex using MOLREP (25). In both the native structure (2.4 Å) and the high resolution HPA/GalNAc complex (1.3 Å), several amino acids missing in the search model could be added and sugar residues clearly defined in density were positioned manually using O (24). Cycles of refinement with Refmac (26), including automatic water molecule placement using ARP/wARP (22), for the native structure and with ShelX (23) using anisotropic B-factors for the complex and manual rebuilding with O resulted in the final models. The refinement statistics are listed in Table 2. Stereochemical checks were performed with the PROCHECK program (27). Glycan Microarray Analysis ⎯ HPA was labeled with Alexa Fluor 488-TFP (Invitrogen, Carslbad, CA) according to the manufacturer’s

instructions. The labeled protein was purified on a D-Salt polyacrylamide desalting column (Pierce, Rockford IL)

Alexa-labeled HPA was used to probe the printed glycan arrays (28) following the standard procedure of Core H of the Consortium for Functional Glycomics (http://www.functionalglycomics.org/) Titration Microcalorimetry Measurements ⎯ ITC experiments were performed with a VP-ITC isothermal titration calorimeter (Microcal). The experiments were carried out at 25°C. GalNAc (Sigma-Aldrich, St Louis, MO) and HPA were dissolved in the same buffer (0.1 M Tris-HCl pH 7.5). Two experiments were conducted with HPA concentrations in the microcalorimeter cell (1.4478 ml) of 0.32 and 0.35 mM. GalNAc solution, in 30 injections of 10 µl at concentrations varying from 2.8 to 3.0 mM, was added at intervals of 5 min whilst stirring at 310 rpm. Control experiments performed by injections of buffer in the protein solution yielded insignificant heats of dilution. The experimental data were fitted to a theoretical titration curve using software supplied by Microcal with ∆H (enthalpy change), Ka (association constant) and n (number of binding sites per monomer) as adjustable parameters, from the classical relationship (29). All experiments were performed with c values 100 < c < 200 (c = Ka x M where M is the initial concentration of the macromolecule).

RESULTS

Sequence Determination and Biochemical Characteristics ⎯ On SDS-polyacrylamide gel electrophoresis, under strong reducing conditions, the commercial HPA preparation showed a single band with an apparent molecular weight of 13 kDa, whilst the MALDI-TOF spectra (data not shown) indicated the presence of major peaks corresponding to a monomer (12.7 kDa), dimer (24.6 kDa) and hexamer (76 kDa) in agreement with previously published data (16). Cyanogen bromide cleavage and subsequent peptide sequencing led to the determination of three peptide sequences (Fig. 1). The sequence of a heptapeptide from the N-terminal region of the mature protein (CGNDAGW) was

3

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

unambiguously deduced from the high resolution X-ray crystal structure of HPA/GalNAc complex and this information allowed for cloning the corresponding hpa gene. Sequence analysis revealed that the gene encodes for a 121 amino acid polypeptide including a peptide signal of 20 residues, as deduced from prediction servers (CBS, http://www.cbs.dtu.dk/services/). The mature protein consists of 101 amino acids, with a calculated molecular weight of 11.3 kDa and an estimated pI of 8.20 (Figure 1). The nucleotide sequence reported in this paper was deposited in GenBankTM/EBI databank with accession number DQ341310. The peptide sequence displays only 53% identity with a HPA sequence deposited by an independent study under accession number Q575S3 (discussed below).

HPA has been described as a glycoprotein and one potential N-glycosylation site is present at Asn34 (mature protein numbering). Mass spectrometry experiments (data not shown) indicated the existence of several glycoforms for the dimer with three major peaks corresponding to MW of 22,795 Da, 23,967 Da and 25,139 Da. These peaks are interpreted as non-glycosylated, monoglycosylated and diglycosylated dimers, respectively. Indeed, the difference of 1,172 Da observed between each form corresponds closely to the low molecular weight N-glycan (MW 1,170 Da) that has been characterized as the most represented on Helix pomatia α-hemocyanin (30). It has been described as a classical core pentasaccharide (Man3-GlcNAc2) with the addition of a α1-6-linked fucose residue to the inner GlcNAc and a β1-2-linked xylose residue to the branching mannose residue. The unusual presence of xylose in an animal glycoprotein was later confirmed in several gastropods ((31) and references therein). Heavier glycoforms were also observed in the mass spectrum and could correspond to Gal, GalNAc and MeGal-containing diantennary N-glycans that have also been characterized in H. pomatia α-hemocyanin (32).

Binding Specificity Study ⎯ Screening for oligosaccharide specificity (Figure 2) confirmed the binding of HPA to glycans with terminal αGalNAc residue, such as the blood group A and Forssman antigens and related oligosaccharides. HPA also appears to tolerate

substitution on position 3 of αGalNAc by a βGal residue as shown by its binding to Thomsen-Freidenreich (T) antigen and its sulfated or sialylated derivatives. Oligosaccharides containing a terminal β-GalNAc or α-GlcNAc residues are also recognized. These binding data confirm, although with some interesting variations (see discussion), the observations made earlier (15,33,34). The only discrepancy is the binding of βGlcNAc1-4βGal1-4βGlcNAc trisaccharide. Since HPA does not bind any of the other βGlcNAc-topped oligosaccharides present in the micro-array, this observation will need to be confirmed using a different methodological approach. Thermodynamics of the Interaction ⎯ In order to characterize the interaction between HPA and GalNAc, the Ka values and the thermodynamic binding parameters were determined using titration microcalorimetry, a method that is well suited to the characterization of protein-carbohydrate interactions (35). A typical titration curve for HPA binding to GalNAc is shown in Fig. 3. Using a mathematical model corresponding to a simple binding mode, the value of the dissociation constant was calculated to be 1.3 x 10-4 M, which is in agreement with the modest affinity reported previously for blood group A pentasaccharide (33). The interaction is strongly driven by enthalpy since the free energy of binding (∆G value of -22 kJ/mol) has an enthalpic contribution, ∆H, of -35 kJ/mol counterbalanced by a strong unfavourable entropy contribution, T∆S, of -13 kJ/mol. Such an entropy barrier is often observed for the binding of carbohydrates to lectins (35). Overall Structure and Oligomerization ⎯ Native HPA crystallized in lens shaped crystals of space group P6322 (a = b = 61.6 Å, c = 104.9 Å) isomorphous to that previously described in a crystallization report (36). In the 2.5 Å resolution X-ray crystal structure, the asymmetric unit contains one monomer consisting of a six stranded antiparallel β-sandwich (Fig. 4A). The β-strands are connected by short loops, except for one longer loop that connects strands A and B, forming a hairpin at one extremity of the β-sandwich. Strands A and E are connected by a disulfide bridge between cysteine residues 9 and 80. The refined electron density map suggested the

4

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

presence of glycosylation at Asn34 but was insufficient to be able to model the glycan, probably due to flexibility of the sugars.

The three-fold crystallographic symmetry axis generates a tightly associated trimer (Fig. 4B). Strand C plays a special role in the trimerization since the first half (amino acids 45 to 50) participates in a strong hydrophobic cluster with the amino acids in the N-terminal and C-terminal regions of the neighboring monomer, whilst its second half (52-56) creates a β-sheet with facing strand D (Fig. 5A).

The disulfide bridge between monomers is centered on a two-fold axis with the involvement of two facing Cys42 residues that are located on one loop of the monomer, close to both the N- and C-termini, creating a very elongated, tail-to-tail type of dimer (Fig. 5B). The resulting rod shaped hexameric molecule (trimers of dimers) is shown in Fig. 4C. The monosaccharide binding site ⎯ Complexes obtained for HPA from co-crystallization with GalNAc produced prismatic crystals of space group H32 (a = b = 48.4 Å, c = 286.2 Å). Again, the asymmetric unit consists of one monomer and the hexamer is created by the 3-fold and 2-fold axis. The overall hexameric structure is very similar to that of native HPA, albeit with few degrees difference in the orientation of one trimer relative to the other. In this high resolution structure (1.3 Å), electron density could be detected for one GalNAc residue in each monomer (Fig. 6A) in a binding site away from the dimerization interface. In addition, one Zn++ cation is present close to the GalNAc binding site (Fig. 6B). This cation lies on a special position on the two fold axis of symmetry and is hexa-coordinated by one nitrogen from the side chain of His84, the two oxygen atoms of one acetate ion and the symmetry equivalents of these three atoms. The presence of this cation occurred in only one commercial lot of protein and is likely to be due to an exposure of living snails to metal ions since these animals have the capacity to accumulate metals (including zinc) in their midgut gland and other organs (37). The presence of the divalent cation is accidental but promoted intermolecular contacts that yielded a different space group and higher resolution.

The GalNAc binding site consists of a shallow pocket at the interface between two monomers (Fig. 6C) made up by part of the long loop connecting strands A and B, the extremities of strands D and E of the same monomer and of strands C and F of the neighboring one. Eight hydrogen bonds are observed between the monosaccharide and the protein. The axial O4 oxygen is the most buried and establishes three hydrogen bonds with the side chains of Arg63, Trp83 and of Asp55 from the neighboring monomer. Arg63 is also hydrogen bonded to the GalNAc ring oxygen O5 and also to O6. This O6 atom of the hydroxymethyl moiety is directed inside the site and interacts with the side chains of Asn61 and Asp55 of the second monomer. The oxygen O3 forms hydrogen bonds with the main chain oxygen of Gly24 and bridges through a water molecule to the side chain of Arg25. The carbonyl oxygen of the sugar N-acetyl group is hydrogen bonded to the Gly24 main chain nitrogen. Interestingly, the methyl group of GalNAc is not involved in any hydrophobic contacts. Such contacts are rather limited since only the CH2 group of C6 makes van der Waals contacts with Tyr89 of the neighboring monomer and the CH group of C1 with His84.

The reported higher affinity for αGalNAc can therefore be attributed to two main contributions. Firstly, three hydrogen bonds are established with the axial O4 oxygen, therefore creating higher affinity for monosaccharides with the galacto configuration. Secondly, His84 establishes a strong contact with C1 and sterically favors the occurrence of the α-anomer at this position.

N-glycosylation Site ⎯ In the highest resolution crystal structure, i.e. the HPA/GalNAc complex, the refined electron density map confirmed the presence of glycosylation at Asn34. Only the first GlcNAc could be unambiguously located. The density was not sufficient to be able to model the whole oligosaccharide, probably due to flexibility of the sugars located in the solvent channel between HPA hexamer. Structural similarities with other proteins ⎯ Although many lectins adopt a β-sandwich, or β-barrel fold, none of them have the same topology and binding site location as HPA. The β-sandwich fold with six strands arranged in a Greek key is shared by a large number of

5

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

protein domains adopting the so-called immunoglobulin-like fold and structural similarity searches result in a large list of crystal structures of unrelated function. Most of them include one or more additional strands. Disulfide bridges creating tail-to-tail dimers of β-sandwiches have been previously observed in HSP 90 co-chaperone p23 although with a different connectivity of the strands in each monomer (38).

The trimeric arrangement is reminiscent to that observed in another lectin involved in innate immunity: the fucose binding agglutinin from Anguilla anguilla eel (AAA) (39). The β-barrel of AAA has a different fold to HPA and the binding site also has a unrelated architecture and specificity. Nevertheless, the presentation of three binding sites on the flat surface formed by the association of the three monomers has strong similarities with distances between the sites going from 20 Å for HPA to 25 Å for AAA. Occurrence of HPA-like proteins ⎯ When searching databases, the peptide sequence that presents the highest identity with HPA (60%) is the recently characterized CH-IIL, the GalNAc-binding lectin from the albumin gland of the garden snail Cepaea hortensis (40). HPA has also homology, albeit lower (54%), with a sequence deposited by the same group and described as Helix pomatia lectin (Q575S3_HELPO). Since this latter is much closer to the Cepaea hortensis lectin sequence (81% identity), it would appear to be from a different snail subspecies than the one from which the commercial lectin is available. The three gastropod lectins have a similar size and are predicted to share the same hexameric arrangement since the cysteine residue involved in making the intermolecular disulfide bridge is conserved as displayed in the alignment of Fig. 7.

SLL-2, a lectin isolated from the octocoral Sinularia lochmodes, also displays significant similarity in both size (94 amino acids) and sequence (23 to 25% depending on the isoforms) to HPA. This galactose-specific lectin is involved in the surface recognition of symbiotic dinoflagellate Symbiodinium and the sequence has been shown to be related with the C-terminal region of discoidins (41). These proteins, that mediate the intercellular cohesiveness of Dictyostelium discoideum cells,

are developmentally regulated lectins with a binding preference for GalNAc for discoidin-1 and for Gal or GalNAc for discoidin-2 (42).

The sequence alignment of all of the HPA-like invertebrate lectins or lectin domains is shown in Fig. 7. The similarity is stronger in the C-terminal region and all of the amino acids that are necessary for binding to O4 of galactose (Asp, Arg and Trp) are strictly conserved. There are more variations in the region predicted to correspond to the first β-strand and the long loop. This loop, which in HPA is responsible for binding the N-acetyl group of GalNAc, has the same length in the other gastropod lectins and in discoidin-1, which all bind GalNAc, but is shorter in coral lectins and in discodin-2 that are reported to bind preferentially to galactose.

CONCLUSION

Due to its specificity and multimeric molecular architecture, the HPA lectin appears to be a powerful weapon to agglutinate pathogens. Although the oligosaccharide target recognized by HPA on the surface of pathogens has not yet been identified, its specificity is close to that of tachylectin-2 that binds to GlcNAc and GalNAc and recognizes staphylococcal lipoteichoic acids and LPS from several Gram-negative bacteria (6). N-acetyl groups are very frequently observed in active substances present on pathogen surfaces and now referred as pathogen-associated molecular patterns (PAMPs). The architecture of HPA brings two sets of three binding sites on opposite hydrophilic and charged faces of the molecule, in an architecture that allows two bacteria to be bound at the same time, therefore promoting aggregation and efficient protection of the snail eggs.

HPA has been demonstrated to be a useful tool in histopathology (43). It is considered to bind GalNAc present on the Tn antigen, but not its galactosylated form, the T antigen. However, the glycan array data that we present here, demonstrate that the T antigen could also be recognized. This is in agreement with a previous study (34). The glycan array data presented here should not be considered as quantitative and therefore do not allow to determine range in affinity. Nevertheless, they indicate that binding specificities other than

6

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

terminal αGalNAc should not be ruled out when analyzing HPA binding to complex glycans at the surface of cancer cells. Interestingly, the ganglioside GD2, a well-known tumor-associated carbohydrate antigen (44), does also bind to HPA in our assay. This preliminary screening opens the route for finer characterisation of carbohydrate antigens that are specifically bound by HPA on different tumor cell lines.

The use of HPA as cancer marker is limited due to its origin. The purification from animals as the only available source is a drawback for several reasons, the main one being the presence of impurities as demonstrated here with the occurrence of zinc in the commercial sample. The knowledge of both the cDNA sequence and the structural features, now allows the production of a large quantities of recombinant HPA with constant quality. It also offers the possibility of lectin

engineering to develop new powerful tools for cancer research.

HPA-like proteins were identified in three families of invertebrates where the lectins seem to play different roles: innate immunity in snail, symbiosis mediator in coral and social life in slime mold. Sequence similarities have been pointed out between the coral lectins and some bacterial sequence regions (41). Indeed, using the HPA sequence as a probe, a search in microbial genomes indicates similarity with putative proteins from Silicibacter, Magnetococcus and Rhodobacter. It is therefore likely that HPA-like domains will be present in both animals and bacteria. By analogy with the new fucose-recognition fold that has been observed in Anguilla anguilla agglutinin (AAA) and in bacteria, and which is now referred to as F-type lectins (39), we herein propose the name of H-type lectin for the HPA-like lectins and lectin domains.

7

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

REFERENCES

1. Vasta, G. R., Ahmed, H., and Odom, E. W. (2004) Curr. Opin. Struct. Biol. 14, 617-630 2. Vasta, G. R., Ahmed, H., Fink, N. E., Elola, M. T., Marsh, A. G., Snowden, A., and Odom, E.

W. (1994) Ann. NY Acad. Sci. 712, 55-73 3. Barondes, S. (1984) Science 223, 1259-1264 4. Muller, W. E., Dorn, A., and Uhlenbruck, G. (1985) Acta Histochem. Suppl. 31, 37-46 5. Iwanaga, S., and Lee, B. L. (2005) J. Biochem. Mol. Biol. 38, 128-150 6. Iwanaga, S. (2002) Curr. Opin. Immunol. 14, 87-95 7. Koike, K., Jimbo, M., Sakai, R., Kaeriyama, M., Muramoto, K., Ogata, T., Maruyama, T., and

Kamiya, H. (2004) Biol. Bull. 207, 80-86 8. Prokop, O., Uhlenbruck, G., and Kohler, W. (1968) Vox Sang. 14, 321-333 9. Kohler, W., Prokop, O., and Kuhnemund, O. (1973) J. Med. Microbiol. 6, 127-130 10. Patchett, R. A., Kelly, A. F., and Kroll, R. G. (1991) J. Appl. Bacteriol. 71, 277-284 11. Slifkin, M., and Cumbie, R. (1989) J. Clin. Microbiol. 27, 1036-1039 12. Camus, M. L. (1899) C. R. Acad. Sci. 129, 233-234 13. Uhlenbruck, G., and Prokop, O. (1966) Vox Sang. 11, 519-520 14. Hammarström, S., and Kabat, E. A. (1969) Biochemistry 8, 2696-2705. 15. Wu, A. M., and Sugii, S. (1991) Carbohydr. Res. 213, 127-143 16. Hammarström, S., Westoo, A., and Bjork, I. (1972) Scand. J. Immunol. 1, 295-309 17. Vretblad, P., Hjorth, R., and Laas, T. (1979) Biochim. Biophys. Acta 579, 52-61 18. Leslie, A. G. W. (1992) Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography

26 19. Collaborative Computational Project Number 4 (1994) Acta Crystallogr., 760-763 20. Grosse-Kunstleve, R. W., and Adams, P. D. (2003) Acta Cryst. D59, 1966-1973 21. de La Fortelle, E., and Bricogne, G. (1997) Methods in Enzymology 276, 472-494 22. Perrakis, A., Morris, R., and Lamzin, V. S. (1999) Nat. Struct .Biol. 6, 458-463. 23. Sheldrick, G. M., and Schneider, T. R. (1997) Methods in Enzymoly 277, 319-343 24. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. A47,

110-119 25. Vagin, A., and Teplyakov, A. (1997) J. Appl. Cryst. 30, 1022-1025 26. Murshudov, G. N., A.A.Vagin, and E.J.Dodson. (1997) Acta Crystallogr. D53, 240-255 27. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl.

Crystallogr. 26, 283-291 28. Blixt, O., Head, S., Mondala, T., Scanlan, C., Huflejt, M. E., Alvarez, R., Bryan, M. C., Fazio,

F., Calarese, D., Stevens, J., Razi, N., Stevens, D. J., Skehel, J. J., van Die, I., Burton, D. R., Wilson, I. A., Cummings, R., Bovin, N., Wong, C. H., and Paulson, J. C. (2004) Proc. Natl. Acad. Sci. USA 101, 17033-17038

29. Wiseman, T., Williston, S., Brandts, J. F., and Lin, L. N. (1989) Anal. Biochem. 179, 131-137 30. van Kuik, J. A., van Halbeek, H., Kamerling, J. P., and Vliegenthart, J. F. (1985) J. Biol.

Chem. 260, 13984-13988 31. Gutternigg, M., Ahrer, K., Grabher-Meier, H., Burgmayr, S., and Staudacher, E. (2004) Eur. J.

Biochem. 271, 1348-1356 32. Lommerse, J. P., Thomas-Oates, J. E., Gielens, C., Preaux, G., Kamerling, J. P., and

Vliegenthart, J. F. (1997) Eur. J. Biochem. 249, 195-222 33. Hammarström, S., and Kabat, E. A. (1971) Biochemistry 10, 1684-1692. 34. Piller, V., Piller, F., and Cartron, J. P. (1990) Eur. J. Biochem. 191, 461-466 35. Dam, T. K., and Brewer, C. F. (2002) Chem. Rev. 102, 387-429. 36. Lisgarten, J. N., Chattopadhyay, T. K., Pitts, J. E., Palmer, R. A., Reynolds, C. D., Dao-Thi,

M. H., Van Driessche, E., and Beeckmans, S. (1999) Acta Crystallogr. D55, 1903-1905. 37. Dallinger, R., and Wieser, W. (1984) Comp. Biochem. Physiol. C 79, 117-124 38. Weaver, A. J., Sullivan, W. P., Felts, S. J., Owen, B. A., and Toft, D. O. (2000) J. Biol. Chem.

275, 23045-23052

8

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

39. Bianchet, M. A., Odom, E. W., Vasta, G. R., and Amzel, L. M. (2002) Nat. Struct. Biol. 9, 628-634.

40. Gerlach, D., Schlott, B., Zahringer, U., and Schmidt, K. H. (2005) FEMS Immunol. Med. Microbiol. 43, 223-232

41. Jimbo, M., Koike, K., Sakai, R., Muramoto, K., and Kamiya, H. (2005) Biochem. Biophys. Res. Commun. 330, 157-162

42. Frazier, W. A., Rosen, S. D., Reitherman, R. W., and Barondes, S. H. (1975) J. Biol. Chem. 250, 7714-7721

43. Brooks, S. A., and Wilkinson, D. (2003) Acta Histochem. 105, 205-212 44. Hakomori, S. (2001) Adv. Exp. Med. Biol. 491, 369-402 45. Kraulis, P. (1991) J. Appl. Crystallogr. 24, 946-950 46. Merrit, E. A., and Murphy, M. E. (1994) Acta Crystallogr. D50, 869-873

FOOTNOTES

* This work was supported by a grant from the French Association ARC and from CEFIPRA. J-F. S. was supported by “La Ligue contre le Cancer”. The glycan microarray analysis has been provided by the Consortium for Functional Glycomics funded by the National Institute of General Medical Sciences grant GM62116 1 These two authors contributed equally to the work. 2 The abbreviations used are: HPA, Helix pomatia agglutinin; GalNAc, N-acetylgalactosamine; GlcNAc, N-acetylglucosamine 3Coordinates and structure factors have been deposited in the Protein Data Bank with accession codes 2CE6 and 2CCV for native and ligand-bound HPA, respectively

ACKNOWLEDGMENTS

We thank the ESRF, Grenoble, for access to synchrotron data collection facilities. Stéphanie Boullanger (CERMAV) and Bernard Dublet (IBS) are gratefully acknowledged for their expertise in mass spectroscopy.

9

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

LEGENDS FOR FIGURES

Fig 1. cDNA sequence and deduced amino acid sequence of HPA. The signal peptide

sequence is indicated in italics and the sequence used for designing degenerated primer in

bold. The three peptides sequenced after cyanogen bromide cleavage are underlined with

different styles. The N-glycosylated asparagine is indicated by a gray background. Numbering

of the protein starts with the first amino acid of mature protein.

Fig 2. Glycan array analysis of HPA as measured by fluorescence intensity. For clarity, part

of the array is not shown (indicated by //) as no significant binding was observed for these

sugars.

Fig 3. Titration calorimetry results of GalNAc (2.8 mM) binding to HPA (0.32 mM) at 25°C. Top,

data from 30 automatic injections of 10 µl GalNAc each into the HPA-containing cell. Lower, plot of

the total heat released as a function of ligand concentration for the titration shown above (squares).

The solid line represents the best least-square fit for the obtained data.

Fig 4. Ribbon representation of HPA crystal structure. A. Monomer with the intramolecular

disulfide bridge displayed as sticks. B. Trimer represented along the three-fold axis. C.

Hexamer with intermolecular disulfide bridges displayed as sticks. Plots are made using

MOLSCRIPT (45) and RASTER-3D (46), unless otherwise indicated.

Fig 5. Topology diagram displaying the arrangement of (A) the non-covalent trimer and (B)

the covalent dimer. The two b-sheets of each monomer are displayed with different gray

shades.

Fig 6. Crystal structure of the complex between HPA and GalNAc. (A) Final sigmaa

weighted 2Fo-Fc electron density map (blue contoured at 1σ) around the GalNAc residue. (B)

Representation of the binding site with GalNAc (as sticks) and Zn (as ball) and hydrogen

bonds represented by dashed lines. (C) Orthographic representation of the trimer.

Fig 7. Alignment of amino acid sequence repeats from HPA, Helix pomatia lectin with entry

code Q575S3, CH-IIL from snail Cepaea hortensis, SLL-2 isoforms from octocoral Sinularia

lochmodes and discoidins from Dictyostelium discoideum. The background is shaded

according to similarity between the sequences. Structural information from HPA includes the

extent of β-strands (arrows), the amino acids involved in hydrogen bonds through their side

chains (o) or backbone (x) atoms and those establishing hydrophobic contacts (+) to the

ligand.

10

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

Table 1: Data collection and phasing statistics

HPA/GalNAc/

Zn high redundancy HPA native HPA/GalNAc

Beamline ID14-2 ID14-2 ID14-2

Space group H32 P6322 H32

Resolution (Å) 21.5 – 1.72 53.4 – 2.40 11.2 – 1.30

Highest res. Shell (Å) 1.79 – 1.72 2.46 – 2.40 1.33 – 1.30

Wavelength (Å) 0.933 0.933 0.933

Cell dimensions (Å) a = b = 48.4

c = 284.6

a = b = 61.6

c = 105.0

a = b = 48.50

c = 286.47

Total number of hkl 581147 37072 243658

Numb. of unique hkl 14291 5044 32736

Average multiplicity 40.7 7.4 7.4

R merge (%)* 8.4 (54.2) 6.9 (38.0) 5.4 (31.0)

Average I/ σ(I)* 36.3 (8.3) 22.6 (2.7) 22.4 (4.3)

Completeness (%)* 99.6 (100.0) 99.8 (99.9) 99.8 (100.0)

Completeness for

anomalous data (%)* 99.6 (100.0) n/a n/a

Wilson B-factor (Ų) 18.2 56.3 12.9

Phasing power (anom)** 0.797 (2.389) n/a n/a

Rcullis (anom)** 0.884 (0.487) n/a n/a

Average figure of merit** 0.320 (0.637) n/a n/a

* : values in parenthesis refer to the highest resolution shell

**: over all reflections to 1.72 Å resolution and values in parenthesis refer to the lowest

resolution shell of 21.5 – 6.81 Å

R merge = Σ│I-<I>│/ Σ<I>, where I=observed intensity

Rcullis(anom) = r.m.s. (DANOcalc – DANOobs) / r.m.s. (DANOobs)

Phasing Power(anom) = r.m.s. (DANOcalc) / r.m.s.(DANOcalc – DANOobs)

11

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

Table 2 : Refinement statistics

* Angle distances and out of plane distances for SHELX refinement.

HPA native HPA/GalNAc

Resolution limits (Å) 53.4 – 2.40 11.2 – 1.30

Working set R (observation) 22.2 (4785) 14.43 (30995)

Test set R (observation) 26.6 (236) 17.90 (1643)

Average Biso (Ų) 53.0 27.0

Number of residues/protein

atoms 100/784 99/819

Number of

solvent/sugar/Zn/other

ligands

24/0/0/0 99/29/1/14

Number of residues in

alternative conformations 0 7

Number of outliers on

Ramachandran plot

92% most favoured, 0

disallowed

94% most favoured, 1

disallowed (Ser39)

Overall G-factor from

Procheck 1.8 -0.4

RMSD distance and angle

deviations

Bond distances (Å) 0.012 0.013

Bond angles (°)* 1.46 0.032

Planar groups (Å) 0.004 0.032

12

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

Table 3: Contacts between HPA and GalNAc.

GalNAc HPA distances (Å)

Hydrogen bonds O7 Asp26.N 2.95 O3 Gly24.O 2.79

Wat1$ 2.80 O4 Trp83.NE1 2.93

Asp55.OD2* 2.67 Arg63.NE 3.00

O5 Arg63.NH2 2.88 O6 Asn61.ND1 2.84

Asp55.OD1* 2.62 Arg63.NH2 3.24

Van der Waals C1 His84 C6 Tyr89*

$ bridging to Arg25.NE * Neighboring monomer

13

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

acgatacatcataaaca ATG CAG GCT TTC TAC TCC GGA CTC CTG TTA CTC TGT GTT 56 M Q A F Y S G L L L L C V GTG GCG TTC GCT GCA GCT CAG CGC GTA CAG TCG GGA AAA ATC GAC TGT GGC AAC 110 V A F A A A Q R V Q S G K I D C G N 11 GAT GCC GGC TGG GCC AAA GTC CCG TCT GAC GAC CCA GGC AGA GAC AAC ACC AGG 164 D A G W A K V P S D D P G R D N T R 29 GAA CTG GCT AAG AAC ATC ACG TTT GCT TCC CCT TAC TGC AGA CCC CCA GTT GTG 218 E L A K N I T F A S P Y C R P P V V 47 CTT CTG TCC ATC ACA CAA CTG GAC GTT GAA CAG AGC CAG AAC CTG AGA GTC ATC 272 L L S I T Q L D V E Q S Q N L R V I 65 GCT CGT CTG TAC TCT GTC TCA CCT ACA GGG TTC AAG GCC AGC TGT TAC ACT TGG 326 A R L Y S V S P T G F K A S C Y T W 83 CAC AAC ACC AAG GTT TAC AGC ATG AGT ATT AGC TGG ATT TCA ATT GAA AAT TAT 380 H N T K V Y S M S I S W I S I E N Y 101 TAA cttactttcgagagagatcaggtgtcgtcttcgcactgtcatagaaatatgcctctataatttcagg 450 * 122 tcttgctagtctttggagccgatgtagctgtcaataatctttattgtacacaaactaccttgaaaactttct 522 atacataacactttcttatctactataagctgttcgtgataattccgtgagcgttccttgctcctgttgttt 594 gcacaattaaaatacacacgcgtcggtacacaaatgtctcttggttagtctgcagtcgaagatatgtttatt 666 acaagcgaaatcatacacaaataaaccactaacaacctaaaaaaaaaaaaaaaaaaaaaaaaaa 730

Figure 1 14

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

0

60000

1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 162 169 176 183 190 197 204 211 260

β3

α

β3

3-Su β4α3

β4α6

β

β4β4

α

α

β

β

β

α3

β

α3

α8

β4 β4

β3α

GalNAcGal

GlcNAc

Fuc

NeuAc

α3α

α3β

α3β

α3β

α2

β3β

α2

α4 β

α2

Fluo

resc

ence

Carbohydrate array

Figure 2 15

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

0.0 0.5 1.0 1.5 2.0-6

-4

-2

0

- 2.5

0 20 40 60 80 100 120 140 160Time (min)

����

��

Molar Ratio

��

kcal

/mol

e of

inje

ctan

tµc

al/s

ec

���

��

�

0

Figure 3 16

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

A B

C

AA

BB

CC

DD

EE

FF

NN--termterm CC--termterm

Figure 4 17

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

NH2

2

98

A

45

FC9

87

C

57

11

12

COOH

DEB

28

3775

C42

72

6183

N34

Ligand

NH2

2 98

A

11

45

F

87

C

57

COOH

D E BC80

28

3775

72

6183

N34

Ligand

C42

C9

310

310

310

310

15

2225

C80

Y

Y

1215

2225

BC

D

E

A

F

Ligand

COOH

NH2

B

CD

E

A

FLiga

nd

NH2

B

C

DE

A

FLigand

COOH

NH2

COOH

A B

Figure 5

18

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

D26

G24

W83

D55

N61

R63

H84

R25

O4O6

O1

O5

O3

ZnA B

C

Figure 6 19

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from

βA βB βC βD βE βF HPA RVQSGKIDCGNDAGWAKVPSDDPGRDNTRELAKNITFASPYCRPPVVLLSITQLDVE-QSQNLRVIARLYSVSPTGFKASCYTWHNTKVYSMSISWISIENY- : 101 Q575S3 RAQTGEIDCGSDSSWPKDPSDNPNRDNTREIVKHVTFSPQFCREPKVILSVTTLDAE-HTKNLRYEARLKSVSPSGLSASCYSWHDGTIYRMKLSWLAVED-- : 100 CH-IIL RAQTGEIDYGSDSSWPKVPSDDPNKDRTRELVKNVTFPSPYCRPPKVILSVTTLDSD-HTKNLRYQARLKSVSPSGFSASSYTWHDTTIYRMKLSWLAVED-- : 100 SLL-2a RLIHVSRCEMGTSSHRCWPREC-----DTSSDEAISFWPPFENTPKVIVSFGMLDVD-NSHNLRVNSSADDVTVGGFTLHYNSWYTTIVWNYKLIWIACD--- : 94 SLL-2b RLIHVSRCEMGTSTHRCWPRPC-----DTSSDEPISFWPPFENTPNVIVSFGMLDVD-NSNNLRVNSSADDVTVGGFTLHYNSWYTTTVWNYKLIWIACD--- : 94 SLL-2c RLTAVSRCVMGSSTHYCWPSQC-----DTSSDEAISFQLPFENTPNVIVSFGMLDVD-STRNLRVNSSADDVDEEGFTLHFNTWANTIVWNYKLIWIACD--- : 94 disco_1 QPVQSSVTQVGADIYTGDNCALNTGSGKREVVVPVKFQFEFATLPKVALNFDQIDCTDATNQTRIGVQPRNITTKGFDCVFYTWNENKVYSLRADYIATALE- : 253 disco_2 KSYSNPSVQVGEVSIGDRS--LNSGTGSRTIVRHVKFPVEFLSVPIVSIGCKKVDAHTDNGQMRWEGKSENITTKGFDLTFITWGNNAVYDLTFDYVAVEFNN : 257 x x o o o o+ +

Figure 7 20

by guest on June 13, 2020 http://www.jbc.org/ Downloaded from

Richard Alvarez, Christelle Breton, Anne Imberty and Edward P. MitchellJean-Fréderic Sanchez, Julien Lescar, Valérie Chazalet, Aymeric Audfray, Jean Gagnon,

hexameric lectin with a novel foldBiochemical and structural analysis of Helix pomatia agglutinin (HPA): A

published online May 16, 2006J. Biol. Chem. 

  10.1074/jbc.M603452200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on June 13, 2020http://w

ww

.jbc.org/D

ownloaded from