Glycobiology Advance Access originally published online on January 19, 2005
Glycobiology 2005 15(6):625-636; doi:10.1093/glycob/cwi044
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Glycobiology vol. 15 no. 6 © Oxford University Press 2005; all rights reserved.
Interaction of Helicobacter pylori with sialylated carbohydrates: the dependence on different parts of the binding trisaccharide Neu5Ac
3Galß4GlcNAc
Institute of Medical Biochemistry, Göteborg University, PO Box 440, SE 405 30 Göteborg, Sweden
1 To whom correspondence should be addressed; e-mail: petra.johansson{at}medkem.gu.se
Received on October 12, 2004; revised on December 23, 2004; accepted on January 17, 2005
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsReferences |
|---|
We have recently shown that binding of Helicobacter pylori to sialylated carbohydrates is dependent on the presence of the carboxyl group and the glycerol chain of Neu5Ac. In this work we investigated the importance of GlcNAc in the binding trisaccharide Neu5Ac
3Galß4GlcNAc and the role of the N-acetamido groups of both Neu5Ac and GlcNAc. An important part of the project was epitope dissection, that is chemical derivatizations of the active carbohydrate followed by binding studies. In addition we used a panel of various unmodified carbohydrate structures in the form of free oligosaccharides or glycolipids. These were tested for binding by hemagglutination inhibition assay, TLC overlay tests, and a new quantitative approach using radiolabeled neoglycoproteins. The studies showed that the N-acetamido group of Neu5Ac is important for binding by H. pylori, whereas the same group of GlcNAc is not. In addition, Fuc attached to GlcNAc, as tested with sialyl-Lewis x, did not affect the binding. Free Neu5Ac was inactive as inhibitor, and Neu5Ac3Gal turned out to be active. The binding preference for neolacto structures was confirmed, although one strain also was inhibited by lacto chains. The combined results revealed that an intact Neu5Ac is crucial for the interactions with H. pylori. Parts of Gal also seem to be necessary, whereas the role of the GlcNAc is secondary. GlcNAc does influence binding, however, primarily serving as a guiding carrier for the binding epitope rather than being a part of the binding structure. Key words: De-N-acetylation / glycosphingolipids / Helicobacter pylori / sialic acid
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
The bacterium Helicobacter pylori colonizes more than one-half of the world’s human population, causing gastrointestinal diseases, including gastritis, gastric ulcer, and gastric cancer (Peek and Blaser, 2002
). Clinical symptoms appear in about 10–20% of the infected individuals resulting in worldwide medical problems. In 1994 this pathogen was classified as a carcinogenic agent by the International Agency for Research on Cancer (IARC, 1994). The mechanism for infection is not yet fully understood, but it probably includes binding to glycoconjugates on the host cell surface. Several binding specificities are expressed by the bacterium, including interactions with some neutral, sulfated, and sialylated carbohydrates (Karlsson, 2000). Interaction of H. pylori with sialic acid was first described by Evans et al. (1988) and later investigated by various research groups. Our own studies showed the presence of appreciable amounts of H. pylori–binding sialylated glycans on human neutrophils (Miller-Podraza et al., 1999) and interactions with highly complex polyglycosylceramides (Johansson et al., 1999; Miller-Podraza et al., 1997a). The preferred interaction of H. pylori with sialylated glycans is with 3-linked sialic acid (Evans et al., 1988; Hirmo et al., 1996; Johansson and Miller-Podraza, 1998; Miller-Podraza et al., 1997b), whereas glycans having 6-linked Neu5Ac are nonbinding.
The present work is limited to the sialic acid–dependent binding and represents a continuation of the recently published data on interactions of H. pylori with various carbohydrate structures and chemical derivatives (Miller-Podraza et al., 2004). We have used sialyl-3-paragloboside, S-3-PG (Table I), isolated from human erythrocytes, as a model substance for our derivatization experiments because it has a simple and well-defined structure. This glycolipid is furthermore biologically interesting because it is present on the surface of human neutrophils, which are inflammatory cells involved in H. pylori infections (Karlsson, 2000). Thus, S-3-PG is a potential in vivo binding epitope for H. pylori (Miller-Podraza et al., 1999).
|
In this context it should be mentioned that the interaction between neutrophils and H. pylori is believed to be important for survival of the bacterium in the stomach area. H. pylori is relatively resistant to phagocytosis, and it is likely that this bacterium uses neutrophils and their degradation products as a nutritional source (Blaser, 1992; Karlsson, 2000). Previously we have shown that S-3-PG binds H. pylori on thin-layer plates and we have proven, using chemical derivatizations, that the binding is dependent on the presence of two sialic acid parts, the carboxyl group and the glycerol chain (Miller-Podraza et al., 2004). Surprisingly, one of the amide derivatives of S-3-PG turned out to be an excellent binder indicating that only one oxygen of -COOH in Neu5Ac is crucial for the interaction (Miller-Podraza et al., 2004). In the present work we continued "dissection" of the sialylated epitopes adding to the list of the previously analyzed structures new derivatives and new carbohydrate sequences. Chemical and enzymatic derivatizations of carbohydrate epitopes followed by binding studies are frequently applied to define minimum binding sequences and show the importance of different functional groups in the receptor (Lanne et al., 1994, 1995, 1999; Miller-Podraza et al., 2004).
The main purpose of this article was to investigate the role of N-acetamido groups on both Neu5Ac and GlcNAc of S-3-PG and generally the role of sugars other than Neu5Ac in binding of H. pylori to sialylated structures. The interaction of H. pylori with carbohydrates containing Neu5Gc has been reported (Hirmo et al., 1996), however, the influence of N-acetamido groups on the binding has not yet been explained in detail. Besides, a growing volume of information indicates that the role of core chains in natural binding saccharides cannot be ignored. One good example is influenza virus, where the binding to sialic acid is well documented. The structures of natural binding sequences are still a subject of discussion and various experimental data show that the binding strength differs considerably between sialylated saccharides, depending on the core chain (Gambaryan et al., 2002; Matrosovich et al., 1999; Miller-Podraza et al., 2000).
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
S-3-PG, chosen as a model compound, was the simplest ganglioside of human granulocytes that interacted with H. pylori on thin-layer chromatography (TLC) plates (Figure 1; for structure see Table I and Figure 2A). Granulocytes contain a series of binding components with increasing complexity, however, only S-3-PG contains a simple epitope based on one lactosamine unit, suitable for chemical derivatizations. More complex species with repeated Galß4GlcNAc units and possible fucose branches (Müthing et al., 1996
; Stroud et al., 1995) may give complicated patterns of derivatives and may carry additional epitopes associated with inner parts of the core chains (Karlsson, 2000). Figure 1 shows that the relative binding strength to species with extended core chains is higher compared with S-3-PG as discussed further next.
|
|
Chemical dissection of S-3-PG included the following derivatizations: formation of lyso glycolipids, N-acetylation and N-benzoylation of lyso glycolipids, and de-N-acetylation and propionylation of de-N-acetylated sites.
Chemical modifications followed by binding studies
De-N-acylation
The lyso glycolipids were obtained through complete de-N-acetylation/acylation of the S-3-PG molecule. The reaction resulted in loss of the fatty acid from the ceramide part and the loss of two acetyl groups present in Neu5Ac and GlcNAc, respectively, as confirmed by negative fast atom bombardment mass spectrometry (FAB MS) (Table II). The obtained lyso glycolipid with three free amino groups was subsequently N-acetylated or N-benzoylated (Figures 2B and C, respectively, and Table II for FAB MS data). As shown, the S-3-PG derivative with three free amino groups and the N-benzoylated variant of this molecule were negative for binding of H. pylori on TLC plates (Figure 3). There was some binding to the N-acetylated derivative (not shown); however, this interaction was less reproducible and was apparently weaker requiring higher amounts of the glycolipid material.
|
|
Partial de-N-acetylation
De-N-acetylation of S-3-PG resulted in three bands when separated by TLC. The uppermost band corresponded to unmodified S-3-PG, whereas the two others represented mono- and di-de-N-acetylated products. The S-3-PG band and the band right below (mono-de-N-acetylated S-3-PG) were active in binding tests against H. pylori, whereas the third (slowest migrating, di-de-N-acetylated S-3-PG) was not (Figure 4 and Table II). The mono- and di-de-N-acetylated derivatives were separated and purified using preparative TLC and the structures were checked using mass spectrometry. The FAB MS data and binding results are summarized in Table II and the original spectra are shown in Figure 2 (Figures 2D and E for mono-de-N-acetylated and di-de-N-acetylated S-3-PG, respectively). The mono-de-N-acetylated fraction was identified as Neu5Ac-Hex-HexN-Hex-Hex-Cer, that is, with unmodified Neu5Ac and de-N-acetylated HexNAc. This was shown by the mass differences between the molecular ion at m/z 1586.8 and fragment ion Y4 at m/z 1295.8 (presence of Neu5Ac) and between the fragment ions Y3 at m/z 1133.8 and Y2 at m/z 972.6 (presence of HexN) and by the lack of the fragment ion at m/z 1337 (lack of NeuN-Hex-HexNAc-Hex-Hex-Cer). No signals corresponding to lyso S-3-PG (product after de-N-acylation) could be found in any of the spectra.
|
N-propionylation
Mono- and di-de-N-acetylated S-3-PG derivatives purified by preparative TLC were N-propionylated, which resulted in mono- and di-N-propionylated S-3-PG, respectively (Figures 2F and G). The mono-N-propionylated derivative migrated on TLC plates just above S-3-PG, and the di-N-propionylated S-3-PG appeared even higher up on the plate. As expected, the mono-N-propionylated S-3-PG contained the propionyl group only on GlcN. This was confirmed by FAB MS of the purified bands by the mass differences between the molecular ion at m/z 1643.1 and fragment ion Y4 at m/z 1351.9 (presence of Neu5Ac) and between the fragment ions Y3 at m/z 1189.8 and Y2 at m/z 972.7 (presence of HexProp, Figure 2F). The mono-N-propionylated S-3-PG turned out to be binding positive, whereas the di-N-propionylated S-3-PG hardly was recognized by the bacteria. The results are summarized in Table II and an example of binding of H. pylori to de-N-acetylated S-3-PG derivatives and a mixture of N-propionylated S-3-PG derivatives is shown in Figure 4.
Hemagglutination inhibition assay
The hemagglutination inhibition assay was used to compare the binding strengths of various sialylated oligosaccharides including 3'-sialyl-N-acetyllactosamine, 3'-sialylneolactotetraose and the sialyl-Lewis x tetrasaccharide to H. pylori. The assay showed that there was no significant difference between the aforementioned structures regarding the inhibitory strength (Table III). The chromatogram binding assays with dilution series of S-3-PG and sialyl-Lewis x also showed that these two potential receptors bound the bacterium equally well. Both bound at the lower pmol level and the lowest amount required for detection was 1 pmol (Table II). Furthermore, 3'-sialyl-N-acetyllactosamine, S-3-PG, and sialyl-Lewis x oligosaccharides inhibited the binding somewhat better than 3'-sialyllactose. This was confirmed by experiments with three different sialic acid–binding strains (Table III). An interesting observation was made that one of the three strains used in these studies (strain J99) also was inhibited by 3'-sialyllactotetraose (LSTa; for structure, see Table I), which is a saccharide of the lacto series.
|
Hemagglutination inhibition assays were also performed using N-acetyl S-3-PG, N-glycolyl S-3-PG, and GM4 (glycolipids) as inhibitors (see Table I for structures and Table III). Glycosphingolipids are expected to be more powerful inhibitors than free oligosaccharides due to micelle formation and multivalent presentation of epitopes. The results should be interpreted with caution because of possible differences in solubility of the glycolipid species in a water milieu. However in accordance with other results N-acetyl S-3-PG turned out to be a good inhibitor, whereas N-glycolyl S-3-PG was not (Table III). Evident inhibition was also observed using GM4 ganglioside, although the effect was weaker as compared with N-acetyl S-3-PG (Table III). This again indicate that GlcNAc and other sugars of the core chain may be important for the final results of the recognition. An important conclusion drawn from this experiment was that Neu5Ac
3Gal displays the activity and that GlcNAc is not a prerequisite of the binding. It should be noted that gangliosides with very short carbohydrate chains, like GM4 and GM3, are not suitable for TLC overlay assays because of poor accessibility on TLC surfaces.
Binding studies using radiolabeled neoglycoproteins
In these experiments the inhibition of binding of H. pylori to neoglycoproteins carrying sialylated epitopes was investigated using free oligosaccharides as inhibitors. Three different glycoprotein probes, sialyl-dimeric Lewis x-nona–acetyl phenylenediamine (APD)–human serum albumin (HSA), sialyl-Lewis x-hexa-APD-HSA and sialyl-LNnT-penta-APD-HSA, were 125I-labeled and tested for binding to H. pylori, strain CCUG 17874. The sialyl-di-Lewis x–HSA probe was found to be a better binder of H. pylori than the two probes with shorter carbohydrate moieties and was chosen for the inhibition experiments. Depending on H. pylori batch, 5–35% of the probe was bound by the bacterial cells. In comparative inhibition studies the same bacterial batch was therefore used for all runs. The inhibition of binding of H. pylori to the sialyl-di-Lewis x probe by 3'-sialyllactose, 3'-sialyl-N-acetyllactosamine, and sialyl-Lewis x oligosaccharide is shown in Figure 5. Like in the hemagglutination inhibition experiments, the three saccharides displayed similar binding/inhibitory strength with a tendency to a better binding for 3'-sialyl-N-acetyllactosamine and sialyl-Lewis x as compared with 3'-sialyllactose by about 10–20%. 6'-Sialyllactose was used as a negative control and did not displace added probe to any significant extent in accordance with the 3'-sialic acid specificity of H. pylori. Free sialic acid was also used and did not affect binding of the bacteria to the probe (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
One of the goals of the present study was to investigate the role of GlcNAc and N-acetamido groups in the binding trisaccharide Neu5Ac
3Galß4GlcNAc involved in the interaction with H. pylori. A simple approach for analysis of biological roles of hexosamines is de-N-acetylation of the saccharides followed by various N-acylations and functional studies. Our initial work included complete de-N-acetyl/acylation of S-3-PG (for structures, see Table I) using strong alkali and high temperature (Schwarzmann and Sandhoff, 1987). This simple and straightforward method allows preparation of various lyso glycolipid species and their derivatives. The observed nonbinding of H. pylori to the de-N-acetyl/acylated and N-benzoylated derivatives of S-3-PG and poor binding to N-glycolyl S-3-PG (Table II) indicated that N-acetamido group(s) are of importance for the receptor activity. However, further experiments were necessary to support this conclusion because the ceramide of N-acetylated lyso S-3-PG was also modified, which influenced the binding negatively (Table II). The importance of the lipid moiety in the glycolipid molecule has been discussed by other investigators (Lanne et al., 1998; Tang et al., 2001), and it has been shown that the fatty acid 2'OH in glycosphingolipids may interact with Glc/Gal at the reducing end, thus influencing the presentation of the binding epitope (Ångström et al., 1998; Tang et al., 2001).
Further work was thus focused on synthesis of S-3-PG derivatives with an unmodified ceramide part. Using milder alkali conditions, we could prepare S-3-PG derivatives with one or two free amino groups with the lipid moiety intact. Especially useful for final interpretation of our results was the mono-de-N-acetylated S-3-PG having its free amino group on GlcNAc. De-N-acetylation of S-3-PG can theoretically give rise to three different products, one with de-N-acetylated sialic acid, another with de-N-acetylated N-acetylglucosamine, and the third where both sialic acid and N-acetylglucosamine are de-N-acetylated. The derivative where only the GlcNAc was de-N-acetylated was found to be a dominant component in the mono-de-N-acetylated fraction. Higher resistance of Neu5Ac to hydrolysis as compared with GlcNAc and the following predominance of the second derivative may be explained by the presence of the glycerol chain in sialic acid. Nuclear magnetic resonance and molecular modeling studies indicate that the N-acetamido group on the sialic acid is stabilized by a hydrogen bond between the carbonyl oxygen of the N-acetamido group and the hydroxyl at C7 of the glycerol chain (Poppe et al., 1989).
The fact that H. pylori binds to S-3-PG with N-modified GlcNAc and not to S-3-PG with N-modified Neu5Ac and GlcNAc (Table II) shows that the N-acetamido group of the sialic acid is important for the interaction and may be part of the binding area, whereas the same group on GlcNAc is not. A secondary role of the GlcNAc is further supported by the fact that this sugar may be modified by Fuc at C3 without loss of the activity. As shown in the results, Neu5Ac3Galß4GlcNAc-R and Neu5Ac3Galß4(Fuc3)GlcNAc-R (Tables II and III, Figure 5) are equally strong binders of H. pylori. On the other hand, Gal seems to be involved in the binding epitope because Neu5Ac3Gal was inhibitory while Neu5Ac was inactive. It is also evident that the Fuc3 of Neu5Ac3Galß4(Fuc3)GlcNAc-R limits the accessibility of the Galß4 residue, suggesting that the latter monosaccharide contributes only partially to the binding epitope (Figure 6). The groups that appear to be accessible for interactions are hydroxyl functions at C2 and C6 of the Gal.
|
/
–155°/–25°), which from a previous study was suggested to be the one recognized by the H. pylori adhesin (Miller-Podraza et al., 2004
The role of GlcNAc is not quite clear; however, it seems likely that this sugar is important for spatial presentation of the binding structure rather than being a part of the interacting area. Most H. pylori strains display a binding preference for neolacto structures (Neu5Ac3Galß4GlcNAc-R) over lacto sequences (Neu5Ac3Galß3GlcNAc-R)(Table III) (Miller-Podraza et al., 2004), indicating either an ineffectual epitope presentation and/or sterical hindrance due to Gal being ß3-linked to GlcNAc in the lacto case. Of relevance in this regard is that the GlcNAc residue may rotate the sialylated epitope by 180° depending on the type of the glycosidic bond between Gal and GlcNAc (Figure 6, note the position of N-acetamido group of GlcNAc in sialyl-Lewis a versus sialyl-Lewis x, marked with arrows).
It should also be noted that in the extended structures (Neu5Ac3[Galß4GlcNAcß3]nGalß4Glcß1Cer with or without 3-linked fucoses; Müthing et al., 1996; Stroud et al., 1995), a significantly enhanced receptor affinity is evident (Figure 1), arguing that a subepitope far removed from the primary Neu5Ac3Gal epitope exists (Roche et al., 2004). Thus extended sequences having further N-acetyllactosamine segments will also affect the binding epitope presentation favorably in TLC overlay assay, where the plastic layer to some degree mimics conditions existing in cell membranes (Miller-Podraza et al., 2004; Roche et al., 2004).
As shown, H. pylori binds in the first place to sialylated neolacto structures. It should be noted, however, that one of the three strains tested in these studies (strain J99) interacted with sialylated lactotetraose (LSTa, Neu5Ac3Galß3GlcNAcß3Galß4Glc) in addition to the neolacto structures. This is interesting because this strain was earlier reported to bind also to the sialyl-Lewis a antigen (Neu5Ac3Galß3[Fuc4]GlcNAcß-R) (Mahdavi et al., 2002) (see Figure 6, bottom). A comparison of the spatial arrangements of the sialyl-Lewis x and sialyl-Lewis a tetrasaccharides thus reveals that the only significant difference is to be found in the orientation of the GlcNAc (
180° rotation). This confirms that the third sugar of the terminal trisaccharide may be drastically modified without loss of the activity. According to Mahdavi et al. (2002) 40% of all H. pylori strains that recognize sialic acid bind to the sialyl-Lewis a antigen, and it is therefore likely that strains recognizing both types of sialylated glycans, that is, with Gal ß4- and ß3-linked to GlcNAc, display higher tolerance regarding the spatial orientation of GlcNAc and perhaps the core chain in general, which may be of relevance for the infection process as compared with strains binding in the first place to neolacto structures.
Recently it has been reported that some gangliosides of the ganglio-series may be inhibitory against H. pylori (Hata et al., 2004). In line with this we could observe an inhibitory effect of some brain gangliosides including ganglioside GD1a on the strain J99, although strain CCUG 17874 was hardly influenced by these structures (as compared by hemagglutination inhibition tests, data not shown). The interaction with ganglioside GD1a may indicate that some H. pylori strains recognize Neu5Ac3Galß3GalNAc sequence present in O-linked oligosaccharides. This is of interest because the Galß3GalNAc-based structures are common components of human gastric mucins (Slomiany et al., 1984a,b), which are believed to be involved in binding of H. pylori, as dicussed by others (Hirmo et al., 1998, 1999).
In addition to the hemagglutination inhibition assay and the chromatogram binding assay, a novel method using radiolabeled probes was used to study the binding of sialylated sugars to H. pylori. Neoglycoproteins consisting of serum albumin with sialylated carbohydrates coupled to surface lysine residues of the albumin were 125I-labeled and tested for binding to the bacteria. Such multivalent affinity probes have previously been used for binding to H. pylori (Mahdavi et al., 2002). In our version, radiolabeled probe and saccharides were added simultaneously to the bacteria. The probe (sialyl-dimeric Lewis x-nona-APD-HSA) and saccharide are thus in equilibrium while competing for binding to the adhesin and the capability of different saccharides to displace the probe from the binding sites can be assayed. The results obtained using this approach reinforced our conclusion that Fuc coupled to GlcNAc has no negative influence on the activity (Figure 5). The concentrations of 3'-sialyllactose, 3'-sialyl-N-acetyllactosamine, and sialyl-Lewis x required for 50% inhibition of the binding were similar, with a tendency to be somewhat higher for 3'-sialyllactose (by 10–20%) as shown by repeated experiments. In hemagglutination inhibition assays the difference between 3'-sialyl-N-acetyllactosamine and 3'-sialyllactose regarding inhibitory strength could be seen more clearly (Table III), and this again indicates that the influence of the third sugar on the binding depends on the environmental conditions and possible secondary interactions between participating agents. The results obtained using radiolabeled neoglycoproteins were generally consistent with those obtained from the hemagglutination inhibition and the TLC overlay assays. It should also be mentioned that the new affinity probe methodology may have a potential for screening of high-affinity (possibly polyvalent) inhibitors of various bacterial adhesins. Using this methodology we have already shown that the binding efficiency of polyglycosylceramides versus 3'-sialyllactose was 50 times better (unpublished data) in accordance with TLC overlay and hemagglutination inhibition assays (Miller-Podraza et al., 1997a, 2004).
In summary, the sequence required for effective binding of H. pylori to natural sialylated oligosaccharides is a trisaccharide rather than a disaccharide unit as shown by experiments with carbohydrates containing different core chains (Table III). Neu5Ac and its three side chains are critical for the interaction, which indicates their participation in direct binding to the adhesin. Parts of the penultimate Gal also seem to be involved in this process because Neu5Ac3Gal (tested as ganglioside GM4) but not free Neu5Ac turned out to be active as inhibitor (Table III). The third monosaccharide (GlcNAc), however, seems to have an auxiliary role, probably as a guiding carrier for the binding area. As shown in the results, this sugar may be modified at C2 and C3 without loss of the activity. Other sugars of the core chain, especially additional N-acetyllactosamine units (for structures, see Table I), may also be of fundamental importance for the definition of the extended epitope and its presentation on cell membranes and in vivo recognition by H. pylori. Presently it is not possible to more precisely define the interacting area of the epitope. However, the sialic acid–recognizing H. pylori adhesin has been identified (Mahdavi et al., 2002), and this may inspire future studies of the epitope–adhesin complexes by use of nuclear magnetic resonance, MS, and crystallographic approaches.
The aim of this article in a larger perspective is to create a basis for further research on H. pylori receptors and receptor analogs. Such analogs could be used as oral medicines or food supplements. The present treatment for H. pylori includes use of antibiotics, which is associated with risks of development of antibiotic-resistant bacterial strains. This has already been detected in several cases (van der Wouden et al., 2000) and the need for an alternative treatment for H. pylori-associated diseases is urgent. The effectiveness of carbohydrates in inhibition of binding of pathogenic bacteria to host cells has been proven (Mouricout et al., 1990) and the search for effective inhibitors of H. pylori is a current issue. In one such study, an unidentified component of cranberry juice was shown to inhibit the sialic acid–dependent binding of H. pylori to human gastric mucus and human erythrocytes (Burger et al., 2000).
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Materials
H. pylori strains used were strain CCUG 17874 from Culture Collection, University of Göteborg, strain J99 and the babA1A2-knockout mutant derivative of CCUG 17875 (denoted DM in this paper). The two latter strains were kindly provided by Dr. Thomas Borén, Umeå University, Sweden. S-3-PG and N-glycolyl S-3-PG were prepared in our laboratory from human erythrocytes and rabbit thymus, respectively (Karlsson, 1987
). 3'-sialylneolactotetraose was prepared from S-3-PG using ceramide glycanase. Sialyl-Lewis x (glycosphingolipid) was synthesized de novo by Symbicom AB and given to us as a gift. Gangliosides GM1, GD1a, GD1b and GT1b of bovine brain were purchased from Calbiochem (La Jolla, CA) and 3'-sialyllactose from Glycorex AB (Lund, Sweden). Sialyl-Lewis x (saccharide) was purchased from Calbiochem and Dextra Laboratories (Reading, U.K.) and lactose from J. T. Baker Chemical (Phillipsburg, NJ). 3'-Sialyl-N-acetyllactosamine was from Dextra and lactosamine from TRC, Toronto Research Chemicals (Toronto, Canada). Fetuin and asialo-fetuin were purchased from Sigma Chemical (St. Louis, MO) and GM4 from HyTest Ltd (Turku, Finland). LSTa, sialyl-dimeric Lewis x-nona-APD-HSA, sialyl-Lewis x-hexa-APD-HSA and sialyl-LNnT-penta-APD-HSA were from IsoSep (Tullinge, Sweden). Na125I and the PD-10 columns were from Amersham (Uppsala, Sweden).
Preparation of gangliosides
Gangliosides used in these studies were isolated according to a standard procedure (Karlsson, 1987) that included extraction of tissues with mixtures of chloroform, methanol, and water, mild alkaline hydrolysis, dialysis, ion exchange chromatography (using DEAE-cellulose) and silica gel chromatography.
Total ganglioside fractions of human granulocytes, obtained after DEAE chromatography was used in TLC overlay studies. A crude S-3-PG fraction (human erythrocytes), obtained after the silica gel separation was used for preparation of lyso S-3-PG and its derivatives. Highly purified fraction of S-3-PG (human erythrocytes) prepared by high-performance liquid chromatography (HPLC) was used in de-N-acetylation/N-propionylation studies.
Purification of S-3-PG by HPLC
A crude fraction of S-3-PG was further purified by HPLC on a Kromasil (5 micron) column (Phenomenex, Torrance, CA). The gradient used was 65:25:4 to 40:40:12 (C:M:H2O, v/v/v) over a period of 180 min, with a flow rate of 2 ml/min. Fractions of 2 ml were collected and analyzed by TLC using anisaldehyde staining.
Preparation of lyso S-3-PG derivatives
A crude fraction of S-3-PG (0.5–5 mg) was de-N-acylated with 0.8 M KOH in methanol (0.5 ml) at 100°C for 20 h (Schwarzmann and Sandhoff, 1987). The mixture was neutralized with 1 M acetic acid, dialyzed against distilled water for 2 days, and lyophilized. The N-acetyl derivative of S-3-PG was prepared according to Carter and Gaver (1967), as follows. Lyso S-3-PG (0.5 mg) was evaporated, dissolved in methanol:acetic acid anhydride (4:1, v/v, 1000 µl) and left at room temperature overnight. The reaction mixture was then evaporated and dissolved in C:M:H2O (60:35:8, v/v/v, 150 µl). Lyso S-3-PG was also converted to N-benzoyl derivative using N-acylation conditions essentially as described by Pacuszka and Panasiewicz (1995). The lyso glycosphingolipid (500 µg) was dissolved in 0.4 ml dimethylformamide to which benzoic anhydride (12 mg) and triethylamine (5 µl) were added. The mixture was stirred at room temperature for 2 h, after which the product was purified by Sephadex LH-20 column packed in methanol. The sample was eluted with methanol and the fractions checked by TLC. Anisaldehyde was used to stain the plates, and the sugar-positive fractions were collected. The material was next separated by preparative TLC using silica gel plates and C:M:H2O (60:35:8, v/v/v) as developing system. The main band (visualized by spraying with water) was scraped off the plate and extracted with C:M:H2O (60:35:8, v/v/v).
De-N-acetylation of sialyl-3-paragloboside
De-N-acetylation of sialyl-3-paragloboside was performed according to a modified procedure by Rebbaa and Portoukalian (1995). S-3-PG (0.5 mg, purified by HPLC) was dissolved in 0.1 M NaOH in 90% n-butanol (500 µl) and the reaction mixture was incubated at 80°C. After 20 min the products were analyzed by TLC, which showed that almost all of the S-3-PG was modified. Finally, ethylacetate (5 ml) was added to neutralize the hydroxide and the mixture was left at room temperature for at least 15 min before evaporation under a stream of nitrogen.
Purification on silica gel
The evaporated mixture of de-N-acetylated S-3-PG derivatives was dissolved in C:M (2:1, v/v, 2 ml) and purified on a small silica gel column (2 x 0.6 cm). The column was eluted with C:M (2:1, v/v, 5 ml), C:M:H2O (60:35:8, v/v/v, 10 ml), and C:M:H2O (50:40:10, v/v/v, 10 ml). The fractions were evaporated, dissolved in C:M:H2O (60:35:8, v/v/v, 300 µl), and checked on TLC plates for the presence of sugars using C:M:0.25% KClaq (50:40:10, v/v/v) as eluent. The material was recovered in fractions eluted with C:M:H2O (60:35:8, v/v/v).
Preparative TLC
A greater amount of S-3-PG (5 mg, crude fraction) was de-N-acetylated and purified on a silica gel column as described. The glycolipid-containing fractions were pooled, evaporated, and dissolved in C:M:H2O (60:35:8, v/v/v) to a concentration of about 1 mg/ml. From this solution 720 µl were applied on a TLC plate in an 18-cm band. The plate was then developed using C:M:0.25% KClaq (50:50:12, v/v/v). One strip in each end and one in the middle of the plate were cut out and chemically stained with anisaldehyde to indicate the positions of the products. Two bands were extracted in C:M:H2O (60:35:8, v/v/v), and the extracts were evaporated under a stream of nitrogen. The two remainders were dissolved in C:M:H2O (60:35:8, v/v/v, 3 ml) each and half of the volume from the samples was applied to a small column (2.5 x 0.6 cm) each, packed with Sephadex G-25 superfine (prewashed with 5 ml 60:35:8). The sample was eluted with C:M:H2O (60:35:8, v/v/v, 2.5 ml) and C:M (2:1, v/v, 2.5 ml), evaporated, and dissolved in a small volume of C:M:H2O (60:35:8, v/v/v).
Propionylation of de-N-acetylated S-3-PG
De-N-acetylated S-3-PG (fractions after silica gel column containing the de-N-acetylation product mixture), mono-de-N-acetylated S-3-PG, and di-de-N-acetylated S-3-PG were evaporated and dissolved in dry methanol (100 µl). Propionic anhydride (2 µl) was added, and the samples were left for about 1 h at room temperature. Then the reaction mixtures were evaporated and dissolved in C:M:H2O (60:35:8, v/v/v), and the obtained derivatives were analyzed by TLC using C:M:0.25% KClaq (50:40:10, v/v/v) as eluent.
Chromatogram binding assay
The glycolipids were separated by TLC using C:M:0.25% KClaq (50:40:10, v/v/v) as developing solvent, and the plates were overlaid with radiolabeled bacteria as described previously (Hansson et al., 1985). In short, after separation the plates were dried and immersed into a solution of polyisobutylmethacrylate (0.5%, w/v) in diethylether:n-hexane (5:1, v/v) for 1 min, dried, and incubated in phosphate buffered saline (PBS), containing bovine serum albumin (BSA) (2%, w/v) and Tween 20 (0.1%, v/v), for 1 h. The TLC plates were then overlaid with 35S-radiolabeled bacteria in PBS and incubated for 2 h followed by three washings with PBS. The plates were dried again and exposed to BIOMAX MR films for at least 12 h. All steps were performed at room temperature.
Hemagglutination assay
A twofold dilution series of a suspension of H. pylori in PBS was incubated at room temperature for 60 min together with PBS (22.5 µl) and human red blood cells suspended in PBS (0.7%, 25 µl). The extent of hemagglutination was graded with help of a microscope, and the lowest concentration of bacteria giving total hemagglutination was used for the hemagglutination inhibition assay.
Hemagglutination inhibition assay
The hemagglutination inhibition assay was performed as described (Miller-Podraza et al., 1997a). A suspension of H. pylori in PBS (12.5 µl, 108 cells/ml, see hemagglutination assay) and inhibitor (22.5 µl, twofold dilution series) were incubated in microtiter wells for 30–60 min at room temperature. Wells with PBS instead of inhibitor were used for control experiments. Human erythrocytes suspended in PBS (0.7%, 25 µl) were then added to each well, and the plates were incubated for 1.5 h at room temperature. The extent of hemagglutination was evaluated under a microscope and graded to a five-step scale. The lowest concentration of inhibitor giving 50% hemagglutination was noted. Three different sialic acid–binding strains of H. pylori were used, CCUG 17874, DM, and J99.
Binding studies using radiolabeled neoglycoproteins
Sialyl-dimeric Lewis x-nona-APD-HSA was iodinated by the Iodo-Gen methodology. The neoglycoprotein (1 mg/ml, 100 µl) in PBS was added to a glass vial coated with 1,3,4,6-tetrachloro-3,6-diphenylglycouril (10 µg). Na125I (5 µl, 100 mCi/ml) was then added, and after 10 min the iodinated protein was desalted on a PD-10 column with PBS (1% BSA) as eluent. The iodinated protein was collected in a 1.5-ml fraction and diluted 50 times before binding studies. The concentration of the probe was thus in the nanomolar range and the level of 125I incorporation was about 3000 cpm/µl. The specific activity of the probe was about 60 cpm/fmol as measured by a LKB Wallac 1277 GammaMaster gamma counter (Turku, Finland). Sialyl-Lewis x-hexa-APD-HSA and sialyl-LNnT-penta-APD-HSA were also 125I-labeled and tested for binding by H. pylori but did not give as high binding percentage as the probe containing sialyl-dimeric Lewis x and were not used in the binding studies.
The inhibition experiments were performed in the following way. H. pylori was suspended in PBS (0.5% BSA) to a concentration of 108 cells/ml, and 100 µl of the bacteria was added to each tube. The iodinated probe (5 µl) and a saccharide (10 µl, dilution series) were then added and the tubes were gently agitated for 30 min. Finally, the mixtures were centrifuged at 12,000 rpm for 2 min, and the supernatant was discarded. The bottom of the tubes was cut off and placed in the gamma counter.
FAB MS
The samples were analyzed by FAB MS in the negative ion mode on a JEOL SX-102A mass spectrometer (JEOL, Tokyo). Triethanolamine was used as matrix and the resolution set to 1000. The FAB gas used was xenon, the energy 6 keV, and the acceleration voltage was –8 or –10 kV.
Molecular modeling
Minimum energy conformations of the sialyl-Lewis x and sialyl-Lewis a tetrasaccharides were calculated within the Quanta2000/CHARMm25 software (Accelrys) on an Indigo2Extreme workstation (Silicon Graphics) using literature values for the glycosidic torsion angles (Imberty et al., 1991; Miller-Podraza et al., 2004).
|
Acknowledgements |
|---|
Prof. Karl-Anders Karlsson is gratefully acknowledged for valuable discussions. This work was supported by the Swedish Medical Research Council (grant number 06X-12628), the Swedish Cancer Foundation, the Wilhelm and Martina Lundgren’s Research Foundation, the Adlerbertska Research Foundation, the Ingabritt and Arne Lundberg Foundation, the Medical Faculty of Sahlgrenska Academy, and Biotie Therapies Corporation.
|
Abbreviations |
|---|
APD, acetyl phenylenediamine; BSA, bovine serum albumin; C, chloroform; CCUG, Culture Collection, University of Göteborg; Cer, ceramide; FAB MS, fast atom bombardment mass spectrometry; Fuc, fucose; Gal, galactose; Glc, glucose; GlcNAc, N-acetylglucosamine; GSL, glycosphingolipid; HPLC, high-performance liquid chromatography; HSA, human serum albumin; LNnT, neolactotetraose; LSTa, 3'-sialyllactotetraose; M, methanol; Neu5Ac, N-acetyl neuraminic acid; Neu5Gc, N-glycolyl neuraminic acid; PBS, phosphate buffered saline; S-3-PG, sialyl-3-paragloboside; TLC, thin-layer chromatography
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Ångström, J., Teneberg, S., Milh, M.A., Larsson, T., Leonardsson, I., Olsson, B.-M., Ölwegård Halvarsson, M., Danielsson, D., Näslund, I., Ljungh, Å., and others. (1998) The lactosylceramide binding specificity of Helicobacter pylori. Glycobiology, 8, 297–309.
Blaser, M.J. (1992) Hypotheses on the pathogenesis and natural history of Helicobacter pylori–induced inflammation. Gastroenterology, 102, 720–727.
Burger, O., Ofek, I., Tabak, M., Weiss, E.I., Sharon, N., and Neeman, I. (2000) A high molecular mass constituent of cranberry juice inhibits Helicobacter pylori adhesion to human gastric mucus. FEMS Immunol. Med. Microbiol., 29, 295–301.
Carter, H.E. and Gaver, R.C. (1967) Improved reagent for trimethylsilylation of sphingolipid bases. J. Lipid Res., 8, 391–395.[Abstract]
Evans, D.G., Evans, D.J. Jr., Moulds, J.J., and Graham, D.Y. (1988) N-acetylneuraminyllactose-binding fibrillar hemagglutinin of Campylobacter pylori: a putative colonization factor antigen. Infect. Immun., 56, 2896–2906.
Gambaryan, A., Webster, R., and Matrosovich, M. (2002) Differences between influenza virus receptors on target cells of duck and chicken. Arch. Virol., 147, 1197–1208.[CrossRef][ISI][Medline]
Hansson, G.C., Karlsson, K.-A., Larson, G., Strömberg, N., and Thurin, J. (1985) Carbohydrate-specific adhesion of bacteria to thin-layer chromatograms: a rationalized approach to the study of host cell glycolipid receptors. Anal. Biochem., 146, 158–163.[CrossRef][ISI][Medline]
Hata, Y., Murakami, M., and Okabe, S. (2004) Glycoconjugates with NeuAc-NeuAc-Gal-Glc are more effective at preventing adhesion of Helicobacter pylori to gastric epithelial cells than glycoconjugates with NeuAc-Gal-Glc. J. Physiol. Pharmacol., 55, 607–625.[ISI][Medline]
Hirmo, S., Kelm, S., Schauer, R., Nilsson, B., and Wadström, T. (1996) Adhesion of Helicobacter pylori strains to -2,3-linked sialic acids. Glycoconj. J., 13, 1005–1011.
Hirmo, S., Kelm, S., Iwersen, M., Hotta, K., Goso, Y., Ishihara, K., Suguri, T., Morita, M., Wadström, T., and Schauer, R. (1998) Inhibition of Helicobacter pylori sialic acid-specific haemagglutination by human gastrointestinal mucins and milk glycoproteins. FEMS Immunol. Med. Microbiol., 20, 275–281.
Hirmo, S., Artursson, E., Puu, G., Wadström, T., and Nilsson, B. (1999) Helicobacter pylori interactions with human gastric mucin studied with a resonant mirror biosensor. J. Microbiol. Methods, 37, 177–182.
IARC (1994) Schistosomes, liver flukes and Helicobacter pylori. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Lyon, 7–14 June. IARC Monogr. Eval. Carcinog. Risks Hum., 61, 1–241.
Imberty, A., Delage, M.-M., Bourne, Y., Cambillau, C., and Perez, S. (1991) Data bank of three-dimensional structures of disaccharides: part II, N-acetyllactosaminic type N-glycans. Comparison with the crystal structure of a biantennary octasaccharide. Glycoconj. J., 8, 456–483.
Johansson, L. and Miller-Podraza, H. (1998) Analysis of 3- and 6-linked sialic acids in mixtures of gangliosides using blotting to polyvinylidene difluoride membranes, binding assays, and various mass spectrometry techniques with application to recognition by Helicobacter pylori. Anal. Biochem., 265, 260–268.
Johansson, L., Johansson, P., and Miller-Podraza, H. (1999) Neu5Ac3Gal is part of the Helicobacter pylori binding epitope in polyglycosylceramides of human erythrocytes. Eur. J. Biochem., 266, 559–565.
Karlsson, K.-A. (1987) Preparation of total nonacid glycolipids for overlay analysis of receptors for bacteria and viruses and for other studies. Methods Enzymol., 138, 212–220.[ISI][Medline]
Karlsson, K.-A. (2000) The human gastric colonizer Helicobacter pylori: a challenge for host-parasite glycobiology. Glycobiology, 10, 761–771.
Lanne, B., Schierbeck, B., and Karlsson, K.-A. (1994) On the role of the carboxyl group of sialic acid in binding of cholera toxin to the receptor glycosphingolipid, GM1. J. Biochem. (Tokyo), 116, 1269–1274.
Lanne, B., Uggla, L., Stenhagen, G., and Karlsson, K.-A. (1995) Enhanced binding of enterotoxigenic Escherichia coli K99 to amide derivatives of the receptor ganglioside NeuGc-GM3. Biochemistry, 34, 1845–1850.
Lanne, B. Ångström, J., Ölwegård, M., Teneberg, S., Stenhagen, G., Karlsson, K.-A., and Nilsson, E. (1998) Binding of N-trifluoroacetyl-derivatized natural glycosphingolipids by uropathogenic Escherichia coli and Neisseria subflava. Glycoconj. J., 15, 1017–1031.[CrossRef][ISI][Medline]
Lanne, B., Schierbeck, B., and ÅngstrÖm, J. (1999) Binding of cholera toxin B-subunits to derivatives of the natural ganglioside receptor, GM1. J. Biochem. (Tokyo), 126, 226–234.
Mahdavi, J., Sondén, B., Hurtig, M., Olfat, F.O., Forsberg, L., Roche, N., Ångström, J., Larsson, T., Teneberg, S., Karlsson, K.-A., and others. (2002) Helicobacter pylori SabA adhesin in persistent infection and chronic inflammation. Science, 297, 573–578.
Matrosovich, M., Zhou, N., Kawaoka, Y., and Webster, R. (1999) The surface glycoproteins of H5 influenza viruses isolated from humans, chickens, and wild aquatic birds have distinguishable properties. J. Virol., 73, 1146–1155.
Miller-Podraza, H., BergstrÖm, J., Milh, M.A., and Karlsson, K.-A. (1997a) Recognition of glycoconjugates by Helicobacter pylori. Comparison of two sialic acid-dependent specificities based on haemagglutination and binding to human erythrocyte glycoconjugates. Glycoconj. J., 14, 467–471.
Miller-Podraza, H., Milh, M.A., Teneberg, S., and Karlsson, K.-A. (1997b) Binding of Helicobacter pylori to sialic acid-containing glycolipids of various origins separated on thin-layer chromatograms. Infect. Immun., 65, 2480–2482.
Miller-Podraza, H., Bergström, J., Teneberg, S., Milh, M.A., Longard, M., Olsson, B.-M., Uggla, L., and Karlsson, K.-A. (1999) Helicobacter pylori and neutrophils: sialic acid-dependent binding to various isolated glycoconjugates. Infect. Immun., 67, 6309–6313.
Miller-Podraza, H., Johansson, L., Johansson, P., Larsson, T., Matrosovich, M., and Karlsson, K.-A. (2000) A strain of human influenza A virus binds to extended but not short gangliosides as assayed by thin-layer chromatography overlay. Glycobiology, 10, 975–982.
Miller-Podraza, H., Johansson, P., Ångström, J., Larsson, T., Longard, M., and Karlsson, K.-A. (2004) Studies on gangliosides with affinity for Helicobacter pylori: binding to natural and chemically modified structures. Glycobiology, 14, 205–217.
Mouricout, M., Petit, J.M., Carias, J.R., and Julien, R. (1990) Glycoprotein glycans that inhibit adhesion of Escherichia coli mediated by K99 fimbriae: treatment of experimental colibacillosis. Infect. Immun., 58, 98–106.
Müthing, J., Spanbroek, R., Peter-Katalinic, J., Hanisch, F.-G., Hanski, C., Hasegawa, A., Unland, F., Lehmann, J., Tschesche, H., and Egge, H. (1996) Isolation and structural characterization of fucosylated gangliosides with linear poly-N-acetyllactosaminyl chains from human granulocytes. Glycobiology, 6, 147–156.
Pacuszka, T. and Panasiewicz, M. (1995) Photochemical labeling of human erythrocyte membranes with radioiodinatable azidosalicylic acid derivative of globoside. Biochim. Biophys. Acta, 1257, 265–273.[Medline]
Peek, R.M. Jr. and Blaser, M.J. (2002) Helicobacter pylori and gastrointestinal tract adenocarcinomas. Nat. Rev. Cancer, 2, 28–37.[CrossRef][ISI][Medline]
Poppe, L., Dabrowski, J., von der Lieth, C.-W., Numata, M., and Ogawa, T. (1989) Solution conformation of sialosylcerebroside (GM4) and its NeuAc(2-3)Galß sugar component. Eur. J. Biochem., 180, 337–342.[ISI][Medline]
Rebbaa, A. and Portoukalian, J. (1995) Optimal conditions to radiolabel (3H or 14C) aminosugar-containing glycosphingolipids by de-N-acetylation and re-N-acetylation. Biochim. Biophys. Acta, 1244, 418–422.
Roche, N., Ångström, J., Hurtig, M., Larsson, T., Borén, T., and Teneberg, S. (2004) Helicobacter pylori and complex gangliosides. Infect. Immun., 72, 1519–1529.
Schwarzmann, G. and Sandhoff, K. (1987) Lysogangliosides: synthesis and use in preparing labeled gangliosides. Methods Enzymol., 138, 319–341.[ISI][Medline]
Slomiany, A., Zdebska, E., and Slomiany, B.L. (1984a) Structures of the neutral oligosaccharides isolated from A-active human gastric mucin. J. Biol. Chem., 259, 14743–14749.
Slomiany, B.L., Zdebska, E., and Slomiany, A. (1984b) Structural characterization of neutral oligosaccharides of human H+Leb+ gastric mucin. J. Biol. Chem., 259, 2863–2869.
Spiro, R.G. (1960) Studies on fetuin, a glycoprotein of fetal serum. J. Biol. Chem., 235, 2860–2869.
Stroud, M.R., Handa, K., Ito, K., Salyan, M.E.K., Fang, H., Levery, S.B., Hakamori, S.-i., Reinhold, B.B., and Reinhold, V.N. (1995) Myeloglycan, a series of E-selectin-binding polylactosaminolipids found in normal human leukocytes and myelocytic leukemia HL60 cells. Biochem. Biophys. Res. Commun., 209, 777–787.[CrossRef][ISI][Medline]
Tang, W., Seino, K., Ito, M., Konishi, T., Senda, H., Makuuchi, M., Kojima, N., and Mizuochi, T. (2001) Requirement of ceramide for adhesion of Helicobacter pylori to glycosphingolipids. FEBS Lett., 504, 31–35.
van der Wouden, E.J., Thijs, J.C., Van Zwet, A.A., and Kleibeuker, J.H. (2000) Review article: nitroimidazole resistance in Helicobacter pylori. Aliment. Pharmacol. Ther., 14, 7–14.
CiteULike 
Connotea 
Del.icio.us What's this?
| |||||||||||||||||||||||||||||