A quantitative method to discriminate between non-specific and specific lectin–glycan interactions on silicon-modified surfaces
Introduction
Key players in the development and maintenance of living systems are cell-surface glycans (encompassing glycoproteins, glycolipids, proteoglycans, etc.). Specific interactions between cell-surface glycans and their protein receptors (for example, lectins) have also been implicated in numerous disease processes including bacterial and viral infections [1], [2], [3]. However, the wide-spread impact of lectin–glycan recognition events in nature is enigmatic in that they are often characterized by intrinsically weak monovalent affinity constants (Ka ≈ 102–103 M−1) [4], [5]. In biological settings this low binding has been shown to be augmented by exploiting multivalency. This latter phenomenon usually kicks in when multiple copies of a glycan are presented on a cell surface and interact simultaneously with a target protein partner featuring multiple glycan recognition sites. The operation of multivalent effects has been observed to lead to affinities in the Ka ≈ 105–108 M−1 range [6].
It is no surprise, given their importance, that the development of appropriate tools for studying glycan–lectin interactions such as microarrays has been intensively researched [7], [8], [9]. This technology requires only small amounts of glycan and in addition, offers the possibility that the glycan density on the array surface be modulated so as to favor the establishment of multivalent interactions with the correct lectin partners [10], [11], [12], [13], [14]. Nevertheless, fabrication of glycan microarrays continues to present a challenge. The need for efficient strategies for the immobilization of glycans onto selected interface material is primordial and must necessarily allow control of glycan presentation in terms both, of their density and orientation. An absolute prerequisite is that any non-specific binding of lectins with the glycan array be minimal so as to avoid analysis artefacts and to maximize the sensitivity of the interface. Proteins tend to spontaneously adsorb to most materials having potential for array development resulting in their denaturation with a corresponding loss of their biological function and/or substrate specificity [15], [16]. Physisorption remains a general concern in the study of proteins with high-affinity interactions (ie. antibody-antigen or aptamer/proteins) and even more so with low-affinity ones such as those encountered between lectins and their monomeric glycan ligands. Interfering physisorption leads to an overestimation of the “actual” protein amount bound to the surface ligand. It can therefore result in errors in the determination of binding affinity Ka of specific glycan–lectin interactions [17], [18], [19], [20]. The incorporation of hydrophilic poly(ethylene glycol) (PEG) molecules on interfaces is one of the most widely accepted strategies to minimize non-specific binding of proteins and its effectiveness is found to depend on both the length and density of the PEG units employed [21], [22], [23], [24]. An alternative strategy relies on the implementation of appropriate buffer recipes or surface cleaning protocols that efficiently disrupt weakly physisorbed proteins and facilitate their selective removal [25], [26], [27]. The surfactant sodium dodecyl sulfate (SDS) has been successfully employed to remove proteins such as Bovin Serum Albumin (BSA) physisorbed on oligo(ethylene glycol) alkyl chains grafted to silicon surfaces [28]. In present bioassays, the actual success of limiting the non-specific binding is usually difficult to evaluate so that a set of surface structures or rinse protocols has to be tested in order to obtain an optimal assay condition. Independent of the strategy employed for minimizing non-specific binding, quantitative information on the amount of surface-adsorbed proteins and their distributions would undoubtely help in optimizing glycan arrays.
This work sets out to interrogate the adsorption behavior of lectins on glycan-modified crystalline (1 1 1) silicon interfaces. The interest in using hydrogenated crystalline (1 1 1) Si is two-fold: angstrom-level flatness can easily be obtained and chemical modification of the surface in a controllable way via Si
C formation is possible [29], [30], [31]. In the present study quantitative IR spectroscopy in ATR geometry (IR-ATR) is exploited to assess the characteristics of the surface at a molecular level and contact-mode atomic force spectroscopy (CM-AFM) allowing detailed information to be obtained about the morphological changes that occur after a specific lectin–glycan interaction has been established. We have recently developed an efficient surface-functionalization strategy for the covalent attachment of simple glycans to crystalline silicon and shown that the glycan-terminated surface is effective in the detection of selective lectin interactions [32]. As an extension of this work, we interrogate here the performance of mannoside- and lactoside-modified Si(1 1 1) surfaces in evaluation of the interaction of the lectins Lens culinaris (LENS), specific for the mannosyl moieties, and Peanut agglutinin (PNA), specific for the lactosyl moieties. The effects of the oligo(ethylene glycol) OEG chain length on the effectiveness of the surface and the post-treatment with SDS rinses after interaction will be discussed on the basis of IR-ATR and CM-AFM data.
Section snippets
Materials
All chemicals were of reagent grade or higher and were used as received without further purification. All cleaning reagents (H2O2, 30%; H2SO4, 96%), and etching (HF, 50%; NH4F, 40%) reagents were supplied from Carlo Erba. Undecylenic acid (99%) was purchased from Acros Organics. All other chemicals and lectins (Lens culinaris, Peanut agglutinin, lyophilized powders free of salt highly purified by affinity chromatography) were purchased from Sigma–Aldrich and were used as received. Ultrapure
Results and discussion
Fig. 1 depicts the “click” chemistry strategy used for the successful integration of propargyl-terminated glycans to an azide-functionalized crystalline Si(1 1 1) surface. The strategy is based on the modification of carboxylic acid modified Si(1 1 1) via a peptide coupling with NH2CH2CH2(EG)nN3 (n = 2 or 8) followed by reaction with either propargyl mannoside or propargyl lactoside using a copper (I)-catalyzed cycloaddition (“click”) protocol. A typical IR-ATR spectrum of mannose-terminated silicon,
Conclusion
We have shown that crystalline Si(1 1 1) wafers are well-suited for the linking of glycans in a “click” chemistry approach whereupon the interactions with specific and non-specific lectins can be followed by FTIR-ATR and AFM techniques. We have demonstrated that the nonspecific adsorption can be eliminated using a long enough EG chain (EG8) incorporated in the glycan-terminated monolayer and an effective surfactinated rinse. We also proposed that the interaction of the surface glycans with
Acknowledgments
J.Y. thanks the Ecole Polytechnique for Ph.D financial support (EDX grant). He also thanks Catherine Henry de Villeneuve for fructuous help in AFM imaging. We gratefully acknowledge financial support from the Centre National de la Recherche Scientifique (CNRS). R.B. and S.S. wish to thank the Université Lille 1 and the région Nord Pas de Calais for financial support. S.S thanks the Institut Universitaire de France (IUF) for financial support.
References (55)
- et al.
Trends Biotechnol.
(2004) - et al.
Curr. Opin. Struct. Biol.
(2008) - et al.
Curr. Opin. Chem. Biol.
(2009) - et al.
Adv. Colloid Interface Sci.
(2011) Adv. Colloid Interface Sci.
(1986)- et al.
Bioelectrochemistry
(2010) - et al.
Prog. Surf. Sci.
(2012) - et al.
J. Biol. Chem.
(1995) - et al.
J. Mol. Biol.
(1996) - et al.
Biochim. Biophys. Acta
(2011)
J. Biol. Chem.
J. Colloid Interface Sci.
Biophys. J.
Chem. Rev.
Acc. Chem. Res.
Ann. Rev. Biochem.
Chem. Rev.
J. Am. Chem. Soc.
J. Am. Chem. Soc.
Adv. Mater.
Anal. Chem.
Chem. Soc. Rev.
J. Am. Chem. Soc.
Chem. Commun.
Biochem. J.
Langmuir
Anal. Chem.
Cited by (3)
Detection of the cancer-associated T antigen using an Arachis hypogaea agglutinin biosensor
2019, Biosensors and BioelectronicsCitation Excerpt :Under these conditions, each impedance spectrum was obtained in about 4 min and the total assay time (first measurement, incubation and second measurement) was around 20 min. In order to reduce nonspecific interactions, the following blocking sequence was included in the methodology (Silva and Rangel, 2017): 1) biosensor immersion in ETA 20 mmol l−1 solution for 30 min to block carboxyl groups of 16-MHDA and 6-MHA that did not react during the construction of the biosensor (Frederix et al., 2004); 2) biosensor immersion in EG 10% solution for 30 min to cover unoccupied areas of the gold surface that may interact with glycoproteins (Zheng et al., 2003); 3) biosensor washing with SDS 2% solution after sample incubation to remove proteins that were nonspecifically bound to the biosensor surface (Yang et al., 2016). The use of mixed SAMs also enables the reduction of protein adsorption at electrode surfaces (Frederix et al., 2003).
Label-free impedimetric glycan biosensor for quantitative evaluation interactions between pathogenic bacteria and mannose
2018, Biosensors and BioelectronicsCitation Excerpt :A high glycan density on the sensing surface induced multivalent interactions, and on the contrary, monovalent interactions occurred under low glycan density conditions (Liang et al., 2007; Mammen et al., 1998; Muller et al., 2016; Zeng et al., 2016). However, the super high glycan density impeded the match between carbohydrate and lectins due to the presence of steric hindrance, which weakened the interaction between them (Yang et al., 2014, 2016). The results showed that the adsorption performance of Man/MUA-MH/Au for S. typhimurium ATCC14028 was the best when the concentration of MUA was 4.8% at 1:20, the NIC（Bacteria-Man）was 46.5%, and E. coli JM109 and DH5α were also reach the maximum value of 21.2% and 9.4% respectively (Fig. 1C).