Interactions and Dissociation Constants of Galactomannan Rendered Cellulose Films with Concavalin A by SPR Spectroscopy

Interactions of biomolecules at interfaces are important for a variety of physiological processes. Among these, interactions of lectins with monosaccharides have been investigated extensively in the past, while polysaccharide-lectin interactions have scarcely been investigated. Here, we explore the adsorption of galactomannans (GM) extracted from Prosopis affinis on cellulose thin films determined by a combination of multi-parameter surface plasmon resonance spectroscopy (MP-SPR) and atomic force microscopy (AFM). The galactomannan adsorbs spontaneously on the cellulose surfaces forming monolayer type coverage (0.60 ± 0.20 mg·m−2). The interaction of a lectin, Concavalin A (ConA), with these GM rendered cellulose surfaces using MP-SPR has been investigated and the dissociation constant KD (2.1 ± 0.8 × 10−8 M) was determined in a range from 3.4 to 27.3 nM. The experiments revealed that the galactose side chains as well as the mannose reducing end of the GM are weakly interacting with the active sites of the lectins, whereas these interactions are potentially amplified by hydrophobic effects between the non-ionic GM and the lectins, thereby leading to an irreversible adsorption.


Preparation of Solutions
For all experiments, a 10 mM phosphate buffer prepared from Milli-Q water exhibiting an ionic strength of 100 mM sodium chloride (adjusted with 0.1 M hydrochloric acid or 0.1 M NaOH) was used. ConA was dissolved in the buffer at a concentration of 0.1 mg·mL −1 . The GM was dissolved in Milli-Q water by stirring for 12 h, at concentrations of 1 mg·mL −1 (for SPR adsorption) and 2 mg·mL −1 (for the spin coating process).

Substrate Cleaning and Film Preparation
The SPR gold sensor slides and silicon wafer substrates were immersed in a "piranha" solution containing H 2 O 2 (30 wt.%)/H 2 SO 4 in a proportion of 1:3 v/v and 1:2 respectively for 10 min, subsequently extensively rinsed with Milli-Q water and blow dried in a stream of N 2 gas as reported elsewhere [25]. For film preparation, 100 µL of TMSC solution (0.75 wt.% in chloroform) were deposited on the substrate and then rotated for 60 s at a spinning speed of 4000 rpm and an acceleration of 2500 rpm·s −1 . TMSC was converted to cellulose by treatment with hydrochloric acid (HCl) vapor. For this purpose, the TMSC films were placed in a petri dish of 5 cm in diameter containing 3 mL of 10 wt.% HCl solution. After 12 min exposure to the HCl vapors in the covered dish, the regeneration was verified by water contact angle and ATR-IR measurements (Section 2.4) as reported elsewhere [25][26][27][28]. The film thickness of the films was 55 nm after regeneration.

Infrared Spectroscopy
IR spectra were attained by an Alpha FT-IR spectrometer (Bruker, Billerica, MA, USA) using an attenuated total reflection (ATR) attachment. Spectra were obtained in a scan range between 4000 to 400 cm −1 with 48 scans and a resolution of 4 cm −1 . The data was analyzed with OPUS 4.0 software.

Profilometry
Film thicknesses were acquired with a DETAK 150 Stylus Profiler from Veeco. The scan length was set to 1000 µm over a duration of 3 seconds. Measurements were performed with a force of 3 mg, a resolution of 0.333 µm per sample and a measurement range of 6.5 µm. A diamond stylus with a radius of 12.5 µm was used. Samples were measured after scratching the film deposited on a silicon wafer. The resulting profile was used to calculate the thickness of the different films. All measurements were performed three times.

Atomic Force Microscopy-AFM
Surface morphology and roughness of the films were obtained in tapping mode in ambient atmosphere at room temperature by a Veeco Multimode Quadrax MM scanning probe microscope (Bruker; Billerica, MA, USA) and by a Tosca TM 400 atomic force microscope (Anton Paar, Graz, Austria) using Si-cantilevers (NCH-VS1-W from NanoWorld AG, Neuchatel, Switzerland) with a resonance frequency of 320 kHz respectively 285 kHz and a force constant of 42 N·m −1 . Root mean square (RMS) roughness calculation and image processing was performed with the Nanoscope software (V7.30r1sr3, Veeco, Plainview, NY, USA) and Gwyddion 2.53 (28.02.2019).

Multi-Parameter Sruface Plasmon Resonance Spectroscopy: MP-SPR
MP-SPR spectroscopy was performed with a MP-SPR Navi TM 210A Vasa instrument (from Bionavis Ltd., Tampere, Finland) equipped with two different lasers (670 and 785 nm, respectively) in both measurement channels, using gold coated glass slides as substrate (gold layer 50 nm, chromium adhesion layer 5 nm). All measurements were performed using a full angular scan (39-78 • , scan speed: Gold sensor slides coated with a cellulose thin film were mounted in the SPR, equilibrated in Milli-Q water and afterwards in a 10 mM PBS (phosphate buffered saline) buffer with an ionic strength of 100 mM NaCl at pH 7.4. After equilibration, a solution of the GM (1.0 mg·mL −1 ) was injected into the flow cell for a period of at least 10 min and subsequently rinsed with corresponding buffer. Afterwards, a protein solution (1.0 mg·mL −1 ) was injected and allowed to adsorb for at least 10 min to the GM deposited on the cellulose film. The same results in GM deposition can be obtained when a solution of GM dissolved in MQ-water (1 mg·mL −1 GM, filtered) was dropped onto the regenerated cellulose surface, allowed to adsorb for a period of 10 min and then rinsed carefully with MQ-water. Regardless of the deposition method, the cellulose-galactomannan thin films were mounted in the SPR and equilibrated in a PBS buffer. After equilibration a solution containing ConA at different concentrations was allowed to adsorb for 10 min. The flow rate for all experiments was set to 20 µL·min −1 . After the adsorption experiment the thin films were rinsed with buffer until the signal was completely stabilized. For the calculation of the amount of adsorbed galactomannan and protein, the shift of the SPR angle (∆SPR) was used for each adsorption experiment. After protein adsorption, all samples were rinsed with Milli-Q water and dried in a stream of nitrogen. All experiments with ConA were performed in three parallels and GM adsorption on cellulose was done in 11 repetitions.
Protein adsorption was quantified according to Equation (1), which considers the dependence of the angular response of the surface plasmon resonance in dependence of the refractive index increment (dn/dc) of the adsorbing layer.
For thin layers (<100 nm), k × d p can be considered constant and can be obtained by calibration of the instrument by determination of the peak wavelength of the resonance dip ld. For the MP-SPR NaviTM 210A Vasa used in this study, k × d p values are approximately 1.09 × 10 −7 cm/ • (at 670 nm) and 1.9 × 10 -7 cm/ • (at 785 nm) in aqueous systems. For ConA, dn/dc in water-based buffer systems was reported 0.187 cm 3 ·g −1 which was used to calculate the amount of adsorbed masses. For GM, the dn/dc value of 0.145 cm 3 ·g -1 was used for the calculation of adsorbed mass [29].

Results and Discussion
The cellulose substrates for the SPR studies were obtained by spin-coating an acid labile cellulose derivative, trimethylsilyl cellulose (TMSC) onto SPR gold sensor slides. After deposition of a TMSC film, the conversion to cellulose can be easily accomplished by exposure to HCl vapors as reported elsewhere [25,26,30,31]. This conversion accomplished by splitting the trimethylsilyl-groups, leads to a densification of the film, resulting in a reduction in film thickness by approximately 50% [26]. The reason for this reduction is a structural reorganization that occurs after the cleavage of the silyl groups due to the formation of inter-and intramolecular hydrogen bonds in cellulose [32]. The conversion is also associated with a change in the static water contact angles which changes from hydrophobic (95-100 • ) to hydrophilic (24-35 • ) [26,28,33,34]. The determination of contact angles with other liquids than water revealed disperse contributions to surface free energies, stemming from van der Waals interactions and London dispersion forces [35]. These films are particularly interesting as they provide a mostly amorphous matrix while featuring a smooth surface allowing for excluding roughness and morphology effects [36]. As the surfaces solely consist of cellulose, they represent an ideal model system to screen for interactions with polysaccharides and proteins [37,38].
After preparation of the cellulose film, the next step was to render these surfaces with GM using an in-situ adsorption experiment using SPR. This adsorption experiment was designed in a way that after equilibration of the cellulose thin film in PBS buffer, the GM was pumped over the cellulose surface for a period of at least 10 min. The GM adsorbs onto the cellulose thin film (Figure 1). Afterwards, the film was rinsed with PBS buffer, thereby removing loosely bound material. By employing the De Fejter equation, the amount of deposited material was determined (0.64 ± 0.22 mg·m −2 , averaged from 12 runs).
Polymers 2020, 12, x FOR PEER REVIEW 5 of 10 After preparation of the cellulose film, the next step was to render these surfaces with GM using an in-situ adsorption experiment using SPR. This adsorption experiment was designed in a way that after equilibration of the cellulose thin film in PBS buffer, the GM was pumped over the cellulose surface for a period of at least 10 min. The GM adsorbs onto the cellulose thin film (Figure 1). Afterwards, the film was rinsed with PBS buffer, thereby removing loosely bound material. By employing the De Fejter equation, the amount of deposited material was determined (0.64 ± 0.22 mg m −2 , averaged from 12 runs).   The AFM images ( Figure 3) do not reveal large differences in morphology between the different samples on first glance. There are, however, subtle variations between the different ConA concentrations. Particularly, the surfaces with the highest amount of ConA used for the adsorption look more covered with the lectin compared to those with lower ConA amounts.
The main adsorption mechanism of the lectins on the galactomannan rendered surfaces is not clear from these studies. The first thought was to check whether there is a correlation between the amount of adsorbed GM and the extent of irreversible protein deposition. There were slight deviations in the adsorption of the GM on the cellulose surface, probably due to slight variations in roughness of the cellulose film, but we could not observe any trend that more GM on the films led to more deposited protein nor the other way around. In all cases, the difference between the individual experiments was rather low. The second idea was to investigate which type of specific interactions could come into play.
ConA is a lectin which has a high binding affinity to terminal and α-linked-D-mannose and, to a lesser extent, to α-D-glucose [39]. Previous investigations based on allowed orientations of the ConA binding site showed that the terminal mannose residues have higher probability of reaching and form more stable complexes than internal mannose residues [40,41]. If the oligosaccharide fragments contained the same terminal monosaccharide and differs in the type of linkage it may align in a different way with respect to the protein, leading to favorable or unfavorable interactions that would limit the number of carbohydrate groups to which ConA can bind. The restrictions on the oligosaccharide fragments will be much more stringent if ConA has to bind to an internal mannose residue [42], and therefore ConA binds more effectively to terminal D-mannose residues.
Polymers 2020, 12, x; doi: FOR PEER REVIEW www.mdpi.com/journal/polymers     One option is that in this experiment the mannose reducing end of the GM interacts with one of the active sites of ConA, thereby forming specific interactions. However, in the GM under investigation here, the glycosidic linkages of the mannose residues are in a β-anomeric configuration. The other option would be that the ConA interacts with glucose moieties of the cellulose support. However, we found that ConA adsorption on cellulose is negligible (0.11 mg·m −2 ) and hardly reaches even the extent of BSA (bovine serum albumin) adsorption (0.22 mg·m −2 ), a commonly used marker for non-specific protein adsorption, on cellulose surfaces under the experimental conditions (PBS, pH 7.4, c = 13.6 nM).
Most interaction studies reported with lectins have been performed for mono-and oligosaccharides (very often in solution) and here only the most affine ones have been explored. However, it is clear that also carbohydrates with a lower affinity will bind to the lectins which may be stabilized in the case of polysaccharides by additional interactions [42]. For instance, galectin-3 selectively binds to β-galactosyl residues in its canonical binding site [43,44] but also binds with lower affinity with α-galactosyl residues of galactomannans in the noncanonical carbohydrate-binding site on the F-face of the Gal-3 CRD β-sandwich, and to a less extent, if at all, with the canonical carbohydrate-binding site on the S-face [41]. In the case of the lectin ConA, Goldstein et al. suggested already that it binds to the terminal D-mannose unit at the reducing end of the carbohydrate chain [41]. In the hemicellulose GM, there are a variety of side chains as well as the mannose reducing end available for this type of interactions. Although the specificity for these carbohydrate moieties might be quite small, cooperative effects may come into play. This involves a reduced solubility at the GM-lectin interface which may lead to association effects which are entropically driven [45][46][47]. Further, hydrophobic effects may play a role (e.g., via van der Waals interactions) since the GM is a non-ionic polysaccharide carrying hardly any charges. However, these effects may contribute, but cannot be the driving force of and probably do not completely control the interaction between GM and the lectins. Otherwise, BSA would interact stronger with the GM rendered surfaces, since it adsorbs spontaneously on nearly all surfaces. Therefore, one of the explanations for the interaction is that the sugar moieties of the GM interact with ConA in a very weak fashion, but the reduced solubility at the interface in combination with hydrophobic interactions lead to spontaneous assembly and stabilization on the GM rendered cellulose surfaces.
The value for the dissociation constant that we obtained (2.1 ± 0.8 × 10 −8 M) is lower as in one of the few reported mannan-lectin studies (2.9 × 10 −6 M), where QCM-D was used to study the interaction of the mannan in solution with immobilized ConA on a solid substrate [21]. The differences in K D may have several origins, including variations in the mannan type (i), immobilization method and experimental setup (ii) and the use of different techniques (iii). The influence of the mannan type can be easily rationalized as different side chains may offer a different binding site profile, leading to variations in the dissociation constant. Another important parameter is molar mass, which may lead to steric hindrance and restricted access to binding sites. The immobilization method itself can have an impact as usually the lectins are immobilized in lectin-carbohydrate assays on the surface as they are rather large molecules compared to the analytes such as mono-, oligo and, in some cases, polysaccharides. The disadvantage of immobilization can be partial denaturation of the lectin, concomitant with a loss of active sites. The advantage of the protein immobilization method includes the absence of diffusion limitation, at least when analytes are small in size such as mono and oligosaccharides. As our GM features a rather high molar mass (1 × 10 6 g·mol −1 ), there will be some diffusion limitation and this could be one reason why the dissociation constant is two magnitudes smaller than for the reported literature value where molar mass of the mannan (obtained from Saccharomyces cerevisiae) was reported 54 kDa. Different methods may also yield different results on the deposited amounts. For instance, QCM-D is a gravimetric method which requires to address bound water in biomolecules layers by modelling (e.g., Voight modelling) if the Sauerbrey condition (rigid layer, no dissipation) is not fulfilled. In contrast, SPR measures the optical properties of a layer and does not consider a priori the bound water in such layers. However, also here there are approximations required, particularly if, e.g., dn/dc values are not at hand.

Conclusions
The adsorption of the GM was studied in situ by MP-SPR spectroscopy and revealed irreversible adsorption on the cellulose thin films. Onto these surfaces, ConA was adsorbed at different concentrations and the dissociation constant K D was determined. Although K D is small, the deposited amounts of ConA onto the surface significantly exceeds the deposition on neat cellulose surfaces, which have been successfully studied as protein repellant surfaces in earlier reports. Furthermore, BSA adsorption on GM rendered cellulose surfaces was much lower than the ConA deposition. The mechanism is not completely clear, but the facts point at weak affinity of the carbohydrate moieties of the GM (i.e., mannose reducing end and galactose side chains) to the ConA, which cooperatively act with hydrophobic interactions (e.g., van der Waals) and entropic contributions [48].
It is well known that lectins are involved in many physiological processes whereas their action is often modulated by specific saccharide interactions. A better understanding of the interaction of lectins with mono-, oligo-and polysaccharides may provide a means to develop novel and more efficient drugs for a variety of diseases. Insights from this study can be used in further developments of affinity assays of different polymeric galactomannans with a variety of lectins. In particular, GMs on cellulose provide a system, whose manufacturing in the laboratory is facile and reproducible. The GM rendered surfaces can be screened for a variety of lectins to create a library of binding affinities, which can be relevant for those lectins that are involved in pathological processes.