A Universal Photochemical Method to Prepare Carbohydrate Sensors

polymers

Article A Universal Photochemical Method to Prepare Carbohydrate Sensors Based on Perﬂuorophenylazide Modiﬁed Polydopamine for Study of Carbohydrate-Lectin Interactions by QCM Biosensor

Lili Guo 1, Shuang Chao 1, Pei Huang 1, Xiukai Lv 1, Quanquan Song 1, Chunli Wu 2,*, Yuxin Pei 1 and Zhichao Pei 1,*

1 Shannxi Key Laboratory of Natural Products & Chemical Biology, College of Chemistry & Pharmacy, Northwest A & F University, Yangling 712100, China; [email protected] (L.G.); [email protected] (S.C.); [email protected] (P.H.); [email protected] (X.L.); [email protected] (Q.S.); [email protected] (Y.P.) 2 School of Pharmaceutical Sciences, Zhengzhou University, 100 Kexue Avenue, Zhengzhou 450001, China * Correspondence: [email protected] (C.W.); [email protected] (Z.P.)

Received: 8 May 2019; Accepted: 6 June 2019; Published: 10 June 2019

Abstract: A universal photochemical method to prepare carbohydrate sensors based on perfluorophenylazide (PFPA) modified polydopamine (PDA) for the study of carbohydrate–lectin interactions by a quartz crystal microbalance (QCM) biosensor was developed. The PFPA was immobilized on PDA-coated gold sensors via Schiff base reactions. Upon light irradiation, the underivatized carbohydrates were inserted into the sensor surface, including mannose, galactose, fucose and N-acetylglucosamine (GlcNAc). Carbohydrate sensors were evaluated for the binding to a series of plant lectins. A kinetic study of the interactions between mannose and concanavalin A (Con A), fucose and Ulex europaeus agglutinin I (UEA-I) were performed. This method can eliminate the tedious modification of carbohydrates, improve the experimental efficiency, and reduce the experimental cost, which is of great significance for the development of QCM biosensors and the study of biomolecular interactions.

Keywords: perﬂuorophenylazide; polydopamine; underivatized carbohydrates; QCM biosensor; carbohydrate-lectin interactions

1. Introduction Carbohydrates play important roles in cell communication, apoptosis, immune response and infection by interacting with proteins and other biomolecules [1–3]. The quartz crystal microbalance (QCM) is an ultrasensitive weighing device, consisting of a thin disk of single crystal quartz, with metal electrodes deposited on each side of the disk. The crystal can be made to oscillate at its resonant frequency, f, when connected to an external driving oscillator circuit [4]. Biosensors based on QCM have proven to be an efficient and powerful instrument for label-free detection of biomolecular interactions in real-time, permitting the determination of the kinetics and affinity parameters of the interaction, including the dissociation rate constant, association rate constant, and affinity constant. It can be used for analysis of interactions between protein and carbohydrate with carbohydrate sensors, protein sensors, and cell sensors [5–12]. Therefore, as an efficient analysis tool, QCM sensor technology has attracted widespread attention, especially carbohydrate sensors for carbohydrate-protein interactions [13,14]. Lectins are a heterogeneous group of proteins, have at least one non-catalytic domain that recognize selectively and bind reversibly to specific free carbohydrates or glycans presenting on glycoproteins

Polymers 2019, 11, 1023; doi:10.3390/polym11061023 www.mdpi.com/journal/polymers Polymers 2019, 11, 1023 2 of 13 and glycolipids. The interactions between carbohydrate and lectin are specific and the structure of the carbohydrates will not be altered [15]. QCM biosensor can be used to monitor the binding affinities of carbohydrate and lectin. The binding between an immobilized lectin Con A and polysaccharides was researched using a QCM sensor [16]. Moreover, we successfully immobilized amino-carbohydrates on the surface of the sensor to detect the interactions of carbohydrate-lectin [12]. The efficient immobilization of carbohydrates on a sensor surface is a primary condition for the preparation of QCM carbohydrate sensor surface. There are usually two strategies to immobilize the carbohydrate on the sensor surface, which include covalent immobilization and non-covalent immobilization [17]. The application of non-covalent immobilization is limited because of its poor stability. However, the covalent immobilization method connects the carbohydrate to the basal surface by a covalent bond, which leads to a high stability and wide application. In fact, the preparation of carbohydrate sensors by covalent immobilization frequently requires multi-step synthesis of the carbohydrates, which is generally not applicable to the direct immobilization of underivatized carbohydrates on the sensor surface [15], where the process for multi-step modification of the carbohydrates is usually time consuming [18–21]. Therefore, it is thus highly desirable to develop a universal method which is applicable to the preparation of carbohydrate sensors by directly using the underivatized carbohydrates without any modification. Perfluorophenylazide (PFPA) has special chemical properties. The azide functionality of PFPA can be converted to nitrene under irradiation or heating that can insert C-H, N-H or C=C bond to form a new covalent bond [22]. It has thus been used to thermo- and photochemically lead into functional groups in a range of substances, such as proteins, carbohydrates, and nanoparticles, since it can significantly improve insertion efficiency [23–27]. The versatility of PFPA makes it to be an ideal choice for surface modification. Tyagi et al. [3] used thiol-modified PFPA to covalent immobilize underivatized carbohydrates to gold surfaces by a fast photochemical reaction to prepare Au surface plasmon resonance (SPR) sensors, where the arrays for protein-binding effects were described by SPR imaging. Polydopamine (PDA) is a well-known biomimetic adhesive material, which can be covered on the surfaces of any kind of materials in a simple way [28,29]. Dopamine (DA) is able to react with oxygen to self-polymerize in an alkaline solution and form a thin PDA layer, which can steadfastly attach to a wide range of substrates, such as rocks, metals, polymers and wood. Under alkaline conditions [30,31], the catechol group of PDA will be oxidized to the ortho-benzoquinones, which can react with the nucleophilic group by Michael addition reactions and Schiff-base reactions, such as amino and thiol groups [32]. Compared with traditional methods [33,34], this method has the advantages of a mild reactive condition and user-friendly control, which is widely used on various material surfaces. In our previous work, we successfully prepared a protein sensor which was functionalized by PDA to analyze the interactions between proteins [35]. Moreover, we successfully immobilized amino-modified carbohydrates on the surface of PDA-coated sensors to detect the interactions of carbohydrate-lectin. However, the process for multi-step modification of the carbohydrates was usually time-consuming [12]. In this study, carbohydrate sensors were prepared with underivatized carbohydrates based on PFPA modified PDA sensors by a photochemical fabrication, which avoid the multi-step modification of the carbohydrates. The prepared carbohydrate sensors were further used to evaluate the interactions of carbohydrate-lectin by QCM biosensor. This method can eliminate the tedious modification of carbohydrates, improve the experimental efficiency, and reduce the experimental cost, which has great significance for the development of QCM and the study of biomolecular interactions. Firstly, PFPA derivatives were immobilized on the PDA-coated gold sensor surface by Schiff-base reactions, and subsequently four different underivatized carbohydrates (mannose, galactose, fucose and GlcNAc) were inserted on the sensor surfaces by photochemical fabrication under light conditions. Secondly, the prepared carbohydrate sensors were evaluated for the binding to a series of plant lectins, which include wheat germ agglutinin (WGA), soybean agglutinin (SBA), peanut agglutinin (PNA), UEA-I and Polymers 2019, 11, x FOR PEER REVIEW 3 of 13

Polymers 2019, 11, 1023 3 of 13 to a series of plant lectins, which include wheat germ agglutinin (WGA), soybean agglutinin (SBA), peanut agglutinin (PNA), UEA-I and Con A. In addition, two kinds of carbohydrates and their correspondingCon A. In addition, lectins two were kinds selected of carbohydrates for kinetic and stud theiry. This corresponding research provides lectins werea universal selected method for kinetic for thestudy. preparation This research of carbohydrate provides a universal sensors method based on for PFPA the preparation modified ofPDA carbohydrate sensors by sensors photochemical based on covalentPFPA modiﬁed immobilization PDA sensors of byunderivatized photochemical carbohydrates, covalent immobilization which can of be underivatized used for the carbohydrates, analysis of whichcarbohydrate-lectin can be used for interactions the analysis by of QCM carbohydrate-lectin biosensors. interactions by QCM biosensors.

2. Materials Materials and Methods

2.1. Materials Materials 2,2’-(ethylenedioxy)bis(2,20-(Ethylenedioxy)bis(ethylamine),ethylamine), dopaminedopamine hydrochloride hydrochloride and and D-mannose D-mannose were were obtained obtained from Xiandingfrom Xianding Biotechnology Biotechnology Co. (Shanghai, Co. (Shanghai, China). Chin Methyla). Methyl pentafluorobenzoate pentafluorobenzoate and aminoacetic and aminoacetic acid was obtained from Shanghai Aladdin Bio-Chem Technology Co. (Shanghai, China). N-acetyl-d-glucosamine acid was obtained from Shanghai Aladdin Bio-Chem Technology Co. (Shanghai, China). N-acetyl-D- and l-fucose were obtained from Shanghai Macklin Biochemical Co. (Shanghai, China). Bovine serum glucosamine and L-fucose were obtained from Shanghai Macklin Biochemical Co. (Shanghai, China). albumin (BSA) and d-galactose were obtained from Energy Chemical (Shanghai, China). SBA, WGA, Bovine serum albumin (BSA) and D-galactose were obtained from Energy Chemical (Shanghai, China).PNA, and SBA, UEA-I WGA, were PNA, obtained and from UEA-I Vector were Laboratories obtained from (Burlingame, Vector Laboratories CA, USA). Con (Burlingame, A, concanavalin CA, AUSA). fluorescein Con A, conjugateconcanavalin (FITC-Con A fluorescein A) and conjug tris(hydroxymethyl)amino-methaneate (FITC-Con A) and tris(hydroxymethyl)amino- (Tris) were obtained methanefrom Sigma-Aldrich (Tris) were (Shanghai,obtained from China). Sigma-Aldrich (Shanghai, China). 13 19 1 13C,C, 19FF and and 1HH NMR NMR spectra spectra were were recorded recorded by by using using a Bruker AVANCE IIIIII HDHD 500500MHz MHz Spectrometer (Karlsruhe, Germany) in CDCl . SEM-EDX were obtained using the S-4800 instrument Spectrometer (Karlsruhe, Germany) in CDCl33. SEM-EDX were obtained using the S-4800 instrument (Hitachi Ltd.)Ltd.) with with an an accelerating accelerating voltage voltage of 10.0 of kV.10.0 The kV. photochemistry The photochemistry in this work in wasthis undertakenwork was withundertaken a Rayonet with PRP-100 a Rayonet photochemical PRP-100 photochemica reactor (Branford,l reactor CT, USA).(Branford, The experimentsCT, USA). The were experiments carried out wereusing carried an Attana out QCM using Cell-A200 an Attana (Stockholm, QCM Cell-A200 Sweden) (Stockholm, (Figure1). Sweden) An Attana (Figure gold sensor1). An (Stockholm,Attana gold sensorSweden) (Stockholm, was used forSweden) further was modifications used for further to obtain modifications a carbohydrate to obtain sensor. a carbohydrate sensor.

FigureFigure 1. 1. SchemeScheme of of the the Attana Attana quartz quartz crystal crystal microbalance microbalance (QCM) (QCM) biosensor biosensor flow-through ﬂow-through system. system.

2.2. Sensors Sensors Cleaning QCM gold gold sensors sensors were were placed placed in in a 50 a 50mL mL beaker beaker and and immersed immersed with with 1% NaClO 1% NaClO solution. solution. The Thebeaker beaker was wasshaken shaken slowly slowly to clean to clean the sensors the sensors for 2–3 for min, 2–3 min,and then and thenthe sensors the sensors were werecleaned cleaned with with1% NaClO 1% NaClO solution solution 2–3 times 2–3 times until untilthe surface the surface of the of sensors the sensors were were bright bright yellow. yellow. Furthermore, Furthermore, the sensorsthe sensors were were washed washed with with Milli-Q Milli-Q water water 4–5 4–5times times and and dried dried with with nitrogen nitrogen stream. stream.

Polymers 2019, 11, 1023 4 of 13

The sensors mentioned above were cleaned for 1 h in 30% H2O2 (0.4 mL) and H2SO4 (1.2 mL) at 50 ◦C. At the same time, the beaker was shaken continuously during this process. The sensors were washed with a large amount of Milli-Q water until neutral, then washed with ethanol three times and dried with nitrogen stream.

2.3. Modification of Polydopamine on the Sensor Surfaces 50 µL dopamine solution (2 mg/mL) in 10 mM Tris buffer (pH 8.5) was added to the gold surfaces at room temperature for 1.5 h. After incubation, the sensor surfaces were washed with Milli-Q water and dried with nitrogen stream to obtain the polydopamine-modified QCM sensors.

2.4. Synthesis of Amino-PFPA Compound 7 was synthesized via the previously reported method, which was shown in the Supporting Information.

2.5. Preparation of the PFPA-modiﬁed Sensors Compound 7 (5 mg) was dissolved in 100 µL PBS of pH 8.5. The solution (50 µL) was added to the surfaces of the modiﬁed PDA-coated sensors, and incubated in a wet box for 4 h, then the sensor surfaces were further rinsed with PBS in order to remove compound 7, which was unimmobilized on the surfaces.

2.6. Preparation of Carbohydrate Sensors 50 µL carbohydrate solution (100 mM) was added to the PFPA-modiﬁed sensors, after which the photochemical fabrication of PFPA was carried out under ultraviolet light of 300 nm for 10 min. The sensor surfaces were washed three times with PBS to remove uninserted carbohydrates and dried with nitrogen stream to obtain the carbohydrate sensors.

2.7. SEM-EDX Measurement The samples were characterized by SEM-EDX to analyze the surface chemical composition of sensor surfaces after drying overnight.

2.8. QCM Analysis of Carbohydrate-Lectin Interactions The sensors were docked into the QCM Cell-A200 and equilibrated under a continuous flow rate of running buffer (100 µL/min). Subsequently, the flow rate was reduced to 25 µL/min to study interactions between carbohydrate and lectin. To block the areas which were not react with carbohydrate, BSA (100 µg/mL) was subsequently injected over the carbohydrate sensor surfaces until the surfaces were saturated. Then, 100 µg/mL lectins were injected into the surface of the sensors. The resonant frequency of the quartz crystal could be recorded with the Attester software in real time, just as the frequency shift (∆f ) coupled to the association or dissociation. The interactions of carbohydrate and lectin were monitored for an association period of 84 s and a dissociation period of 300 s. The evaluation software of the Attana QCM Cell-A200 was used to record data. Finally, the bound lectin was removed by injecting 10 mM glycine of pH 1.5 to regenerate the sensors. Attester Evaluation software was used to analyze the data.

3. Results and Discussion The QCM gold sensor surfaces were prepared employing a three-step process. As shown in Figure2, ﬁrstly, in alkaline environment, the gold surfaces were fabricated with PDA via dopamine self-polymerization. After coating of the PDA, the PFPA with an amino group was immobilized on the surfaces of the PDA-coated sensor by Schiﬀ-base reactions. Finally, PFPA of photochemical fabrication was used to immobilize underivatized carbohydrates on the sensor surfaces under an Polymers 2019, 11, 1023 5 of 13

ultraviolet light of 300 nm. In this work, the carbohydrate-lectin interactions were monitored by QCM Polymers 2019, 11, x FOR PEER REVIEW 5 of 13 biosensor. Setting the rate of flow to 25 µL/min of running buffer (10 mM PBS of pH 7.4), the interaction interactionbetween carbohydrates between carbohydrates and different and lectins different wasdetected. lectins was In addition, detected. the In kinetic addition, properties the kinetic of the propertiesinteractions of betweenthe interactions carbohydrate between and carbohydra lectin werete further and lectin performed, were further where performed, we selected where two kinds we selectedof carbohydrates two kinds of for carbohydrates the kinetic study. for the The kinetic interactions study. The between interactions carbohydrate between carbohydrate and lectin can and be lectinmonitored can be by monitored QCM biosensor, by QCM permitting biosensor, the determinationpermitting the of determ the affiinationnity and of the the kinetics affinity parameters and the kineticsof the interactions. parameters of the interactions.

FigureFigure 2. 2. SchematicSchematic illustration illustration of of preparing preparing the the carboh carbohydrateydrate sensor, sensor, based based on on the the universal universal method. method.

3.1.3.1. SEM-EDX SEM-EDX Analysis Analysis of of the the Sensor Sensor Surfaces Surfaces AsAs shown shown in in Figure Figure 3,3, the the chemical chemical composition composition of of these these sensor sensor surfaces surfaces can can be be analyzed analyzed by by SEM-EDXSEM-EDX [36]. [36]. Based Based upon upon the the spectra spectra analysis analysis of of SEM-EDX, SEM-EDX, the the elemental elemental peak peak of of C, C, N N and and O O appearedappeared on on the the PDA-coated PDA-coated gold gold sensors, sensors, compar compareded with with the the element element composition composition of of gold gold sensors sensors (Figure(Figure 3a).3a). The The spectrum spectrum of of SEM-EDX SEM-EDX indicated indicated th thee evident evident signal signal of of the the PDA’s PDA’s atomic atomic composition composition (Figure(Figure 3b).3b). As As shown shown in in Figure Figure S6 S6 of of Supporting Supporting Information, Information, the the region region of of peaks peaks below below 1 1 keV keV for for FigureFigure 3a,b3a,b werewere enlarged. enlarged. Compared Compared with with the the sensor sensor surfaces surfaces coated coated with PDA,with thePDA, SEM-EDX the SEM-EDX spectral spectralsignal of signal the chemical of the chemical composition composition F appeared F appeared on the sensor on the surfaces sensor after surfaces immobilizing after immobilizing compound 7 compound(Figure3c), 7which (Figure indicated 3c), which that indicated the PFPA that was the connected PFPA was to connected the surfaces to the of the surfaces sensor of successfully. the sensor successfully.As shown in As Figure shown3d, thein spectralFigure 3d, signal the ofspectral the chemical signal composition of the chemical C and composition O improved C markedly and O improvedcompared markedly to the PFPA-modiﬁed compared to the sensor, PFPA-modified which indicated sensor, the which immobilization indicated ofthe the immobilization carbohydrate of on thethe carbohydrate sensor surfaces. on the sensor surfaces.

Polymers 2019, 11, 1023x FOR PEER REVIEW 66 of of 13

Figure 3. SEM-EDXSEM-EDX spectra spectra of ofdifferent different sensor sensor surfaces: surfaces: (a) (unmodifieda) unmodified gold gold sensor sensor surface; surface; (b) (polydopamineb) polydopamine (PDA) (PDA) coated coated sensor sensor surface; surface; (c ()c perflourophenylazide) perflourophenylazide (PFPA) (PFPA) immobilized immobilized sensor surface; ( d) carbohydrate sensors. 3.2. FITC-Con A Immobilization on the Carbohydrate Sensor Surface and Regeneration Result 3.2. FITC-Con A Immobilization on the Carbohydrate Sensor Surface and Regeneration Result To verify the interactions between carbohydrate and lectin, FITC-Con A with green fluorescence To verify the interactions between carbohydrate and lectin, FITC-Con A with green fluorescence was used to observe whether the carbohydrate on the sensor surface binding with lectin successfully. was used to observe whether the carbohydrate on the sensor surface binding with lectin successfully. When the baseline was stable (frequency shift less than 0.2 HzHz/min),/min), BSA was injected into the carbohydrate sensor sensor surface surface to to bl blockock the the area area where where PFPA PFPA did did not not react react with with carbohydrate carbohydrate until until the sensorthe sensor surface surface was was saturated. saturated. Then, Then, 100 100 μg/mLµg/mL FITC-Con FITC-Con A Awas was injected injected and and observed observed by by a fluorescencefluorescence microscopemicroscope to to prove prove interactions interactions between between carbohydrate carbohydrate and and lectin. lectin. As As shown shown in Figurein Figure4A, 4A,the bindingthe binding response response of interactions of interactions between between mannose mannose and and FITC-Con FITC-Con A was A was recorded, recorded, resulting resulting in a infrequency a frequency shift aboutshift about 20 Hz. 20 Before Hz. andBefore after and the after injection the ofinjection the FITC-ConA, of the FITC-ConA, a fluorescence a fluorescence microscopic microscopicevaluation was evaluation performed. was Before performed. injection, Before only ainje blackction, field only was a observed, black field while was a observed, green fluorescence while a greencould befluorescence observed aftercould injection be observed of FITC-ConA after inject ontoion theof carbohydrateFITC-ConA onto sensor thesurface. carbohydrate Furthermore, sensor surface.regeneration Furthermore, of the carbohydrate regeneration sensor of the surface carboh wasydrate carried sensor out, surface where wa thes boundedcarried out, Con where A can the be boundedremoved byCon injecting A can be 10 removed mM glycine by ofinjecting pH 1.5 10 (Figure mM 4glycineB). As canof pH be seen,1.5 (Figure the baseline 4B). As reverted can be toseen, its theoriginal baseline state, reverted which indicates to its original that the state, carbohydrate which indicates sensor that surface the carbohydrate was regenerated sensor successfully. surface was regenerated successfully.

Polymers 2019, 11, 1023 7 of 13

Polymers 2019, 11, x FOR PEER REVIEW 7 of 13

FigureFigure 4. BindingBinding of of FITC-Con FITC-Con A Ato tocarbohydrate carbohydrate sensor sensor (A)( andA) and QCM QCM analysis analysis of the of binding the binding of lectin of lectinwith withcarbohydrate carbohydrate immobilized immobilized on the on thePFPA-modified PFPA-modiﬁed sensor sensor surface surface (B (B).). Frequency Frequency shift waswas recordedrecorded duringduring thethe associationassociation (84(84 s)s) andand dissociationdissociation (300(300 s)s) phasesphases ofof thethe interactioninteraction betweenbetween lectinlectin andand carbohydrate,carbohydrate, asas wellwell asas thethe subsequentsubsequent desorptiondesorption ofof thethe remainingremaining boundbound lectinlectin byby anan injectioninjection ofof thethe 1010 mMmM glycineglycine pHpH 1.5.1.5. ((aa)) baseline;baseline; ((bb)) association;association; ((cc)) dissociation;dissociation; ((dd)) regeneration.regeneration.

3.3. QCM Measurements of Carbohyhdrate-Lectin Interactions 3.3. QCM Measurements of Carbohyhdrate-Lectin Interactions AccordingAccording toto thethe reports,reports, ConCon A,A, SBA,SBA, PNA,PNA, WGA,WGA, andand UEA-IUEA-I cancan specificallyspecifically bindbind mannose,mannose, galactosegalactose/GalNAc,/GalNAc, galactose galactose/glucose,/glucose, GlcNAc, GlcNAc, and and fucose, fucose, respectively respectively [37– 42[37–42]]. In this. In study,this study, we chose we fourchose carbohydrates four carbohydrates (mannose, (mannose, galactose, galactose, GlcNAc GlcNAc and fucose) and fucose) to study to study five lectins five lectins interactions interactions with carbohydrates,with carbohydrates, respectively. respectively. TheThe preparedprepared carbohydratecarbohydrate sensorssensors were insertedinserted in the Attana QCM Cell-A200 biosensor. WhenWhen stablestable baseline baseline was was achieved achieved at 25at µ25L /μminL/min of the of PBSthe PBS running running buffer, buffer, we blocked we blocked the areas the which areas didwhich not did react not with react carbohydrate with carbohydrate by BSA. by The BSA. BSA Th wase BSA repeatedly was repeatedly injected injected to the carbohydrate to the carbohydrate sensor surfacessensor surfaces until the until surfaces the surfaces were saturated. were saturated. Then a Then series a ofseries lectins of lectins of 100 ofµg 100/mL μ WGA,g/mL WGA, SBA, PNA,SBA, UEA-I,PNA, UEA-I, and Con and A Con were A injected were injected to detect to detect the interactions the interactions of carbohydrate-lectin, of carbohydrate-lectin, respectively. respectively. After eachAfter injection each injection of the lectin,of the 10lectin, mM 10 glycine mM glycine (pH 1.5) (pH was 1.5) injected was injected to the surface to the forsurface regeneration, for regeneration, and the bindingand the lectinbinding was lectin removed was afterremoved the cycle. after Wethe usedcycle. Attester We used software Attester to software realize real-time to realize monitoring real-time ofmonitoring carbohydrate of carbohydrate with lectin interactions with lectin and interact regenerationions and processes. regeneration Figure processes.5 shows frequency Figure 5 change shows offrequency five diff erentchange lectins of five (WGA, different SBA, lectins PNA, UEA-I,(WGA, ConSBA, A) PNA, binding UEA-I, to GlcNAc-modified Con A) binding sensorto GlcNAc chip- surface.modified As sensor can be chip seen, surface. the binding As can of WGA be seen, is the th highest,e binding at aboutof WGA 20 Hz,is the indicating highest, thatat about other lectins20 Hz, areindicating basically that not other bound lectins except are for basically non-specific not bound adsorption. except for non-specific adsorption. The binding results between the other three carbohydrates and five different lectins were represented in the Figures6–8, respectively. As can be seen in Figure6, Con A has the highest binding with mannose, while the other lectins were basically not bound except that a small amount of non-specific adsorption was detected. As can be seen in Figure7, UEA-I has the highest binding with fucose, while the other lectins were basically not bound except that a small amount of non-specific adsorption was detected. As can be seen in Figure8, SBA has the highest binding with galactose, followed by PNA. This is because both SBA and PNA can bind to galactose, but the binding strengths are different, while the other lectins are not bound except for a small amount of non-specific adsorption. The experimental results show that the interactions between carbohydrate and lectin are specific, which are consistent with the results obtained by other methods, indicating the universality of this method.

Figure 5. Interaction of GlcNAc with a series of lectins.

Polymers 2019, 11, x FOR PEER REVIEW 7 of 13

Figure 4. Binding of FITC-Con A to carbohydrate sensor (A) and QCM analysis of the binding of lectin with carbohydrate immobilized on the PFPA-modified sensor surface (B). Frequency shift was recorded during the association (84 s) and dissociation (300 s) phases of the interaction between lectin and carbohydrate, as well as the subsequent desorption of the remaining bound lectin by an injection of the 10 mM glycine pH 1.5. (a) baseline; (b) association; (c) dissociation; (d) regeneration.

3.3. QCM Measurements of Carbohyhdrate-Lectin Interactions According to the reports, Con A, SBA, PNA, WGA, and UEA-I can specifically bind mannose, galactose/GalNAc, galactose/glucose, GlcNAc, and fucose, respectively [37–42]. In this study, we chose four carbohydrates (mannose, galactose, GlcNAc and fucose) to study five lectins interactions with carbohydrates, respectively. The prepared carbohydrate sensors were inserted in the Attana QCM Cell-A200 biosensor. When stable baseline was achieved at 25 μL/min of the PBS running buffer, we blocked the areas which did not react with carbohydrate by BSA. The BSA was repeatedly injected to the carbohydrate sensor surfaces until the surfaces were saturated. Then a series of lectins of 100 μg/mL WGA, SBA, PNA, UEA-I, and Con A were injected to detect the interactions of carbohydrate-lectin, respectively. After each injection of the lectin, 10 mM glycine (pH 1.5) was injected to the surface for regeneration, and the binding lectin was removed after the cycle. We used Attester software to realize real-time monitoring of carbohydrate with lectin interactions and regeneration processes. Figure 5 shows frequency change of five different lectins (WGA, SBA, PNA, UEA-I, Con A) binding to GlcNAc- Polymersmodified2019 sensor, 11, 1023 chip surface. As can be seen, the binding of WGA is the highest, at about 208 ofHz, 13 indicating that other lectins are basically not bound except for non-specific adsorption.

Polymers 2019, 11, x FOR PEER REVIEW 8 of 13

The binding results between the other three carbohydrates and five different lectins were represented in the Figures 6–8, respectively. As can be seen in Figure 6, Con A has the highest binding with mannose, while the other lectins were basically not bound except that a small amount of non- specific adsorption was detected. As can be seen in Figure 7, UEA-I has the highest binding with fucose, while the other lectins were basically not bound except that a small amount of non-specific adsorption was detected. As can be seen in Figure 8, SBA has the highest binding with galactose, followed by PNA. This is because both SBA and PNA can bind to galactose, but the binding strengths are different, while the other lectins are not bound except for a small amount of non-specific adsorption. The experimental results show that the interactions between carbohydrate and lectin are specific, which are consistent with the results obtained by other methods, indicating the universality of this method. Figure 5. Interaction of GlcNAc withwith aa seriesseries ofof lectins.lectins.

FigureFigure 6.6. InteractionInteraction ofof mannosemannose withwith aa seriesseries ofof lectins.lectins.

FigureFigure 7.7. InteractionInteraction ofof fucosefucose withwith aa seriesseries ofof lectins.lectins.

Polymers 2019, 11, 1023 9 of 13 PolymersPolymers 20192019,, 1111,, xx FORFOR PEERPEER REVIEWREVIEW 99 ofof 1313

FigureFigure 8. 8. InteractionInteraction ofof galactosegalactose withwith a a series series of of lectins. lectins.

ExperimentsExperiments werewere repeatedrepeated threethree times, times, and and the the error error bar bar gives gives a a general general idea idea of of how how precise precise a a measurementmeasurement is.is. Mannose,Mannose, fucose,fucose, galactosegalactose andand and GlcNGlcN GlcNAcAcAc werewere usedused toto study study specific speciﬁcspecific binding binding with with lectinslectins under under the the samesame concentrationconcentration (100 (100 μμµg/mL).gg/mL)./mL). As As shown shown in in Figure Figure 9 9,9,, thethethe errorerrorerror barbarbar resultsresultsresults showedshowedshowed thatthat each each carbohydratecarbohydrate andandand itsitsits corresponding correspondingcorresponding lectin lectinlectin can cancan bind bindbind speciﬁcally, specifically,specifically, illustrating illustratingillustrating the thethe feasibility feasibilityfeasibility of ofthisof thisthis method. method.method.

Figure 9. Error bar results of the four underivatized carbohydrates. FigureFigure 9. 9. ErrorError bar bar results results of of the the four underivatized carbohydrates.

3.4.3.4. KineticKinetic andand AffinityAffinity Affinity Studies Studies WeWe used used an an Attana Attana QCMQCM Cell-A200Cell-A200 biosensorbiosensor toto performperform inin situ situ real-time real-time measurements measurements of of the the bindingbinding kinetics kinetics of of the the interactionsinteractions between between lectin lectin and and carbohydrate carbohydrate sensor. sensor. ToTo To study study the the kinetic kinetic propertiesproperties of of the the interactions interactions between between UEA-I UEA-I an andandd fucose, fucose, a a diluted diluted solutionsolution (25, (25, 50,50, 7575 andand 100100 μμµg/mL)g/mL)g/mL) ofof UEA-I UEA-I in in the the running running buffer bubufferffer waswas injectedinjected to to thethe surface.surface. The The binding binding was was monitored monitored inin real real time time (the(the binding binding responses responses are are shown shown in in Figure Figure 10).1010).). Th TheThee overalloverall fitfit fit waswas performedperformed usingusing aa theoreticaltheoretical 1:11:1 interactioninteraction modelmodelmodel with withwith limitations limitationslimitations of mass ofof massmass transport transporttransport and employing andand employingemploying Attana QCM AttanaAttana Cell-A200 QCMQCM instrumentCell-A200Cell-A200 instrumentevaluationinstrument evaluation software.evaluation Furthermore,software.software. Furthermore,Furthermore, the sensor thethe surface sensorsensor was surfacesurface treated waswas with treatedtreated 25–100 withwith µ25-10025-100g/mL μ ofμg/mLg/mL UEA-I, ofof UEA-I,thenUEA-I, the thenthen instrument thethe instrumentinstrument recorded recordedrecorded the response thethe (blackresponseresponse line). (black(black The Attanaline).line). TheThe evaluation AttanaAttana softwareevaluationevaluation generated softwaresoftware a generatedtheoreticalgenerated a 1:1a theoreticaltheoretical fit that was 1:11:1 covered fitfit thatthat (red waswas line). coveredcovered We obtained (red(red line).line). constants WeWe obtainedobtained of interactions constantsconstants between ofof interactionsinteractions UEA-I and 4 −1 −1 betweenbetween UEA-IUEA-I andand fucose,fucose, suchsuch asas thethe associationassociation raterate 4constantconstant1 1 (K(Konon == 1.661.66 ×× 10104 MM−1ss−1),), thethe fucose, such as the association rate constant (Kon = 1.66 10 M− s− ), the dissociation rate constant dissociation rate constant (Koff = 1.27 × 10−−33 s−−11) and affinity× constant (KD = 76.5 nM). dissociation(K = 1.27 rate10 constant3 s 1) and (K aoffffi =nity 1.27 constant × 10 s (K) and= affinity76.5 nM). constant (KD = 76.5 nM). off × − − D

Polymers 2019, 11, 1023 10 of 13 PolymersPolymers 20192019,, 1111,, xx FORFOR PEERPEER REVIEWREVIEW 1010 ofof 1313

FigureFigure 10. 10. KineticKinetic evaluation evaluation of of the the intera interactioninteractionction between between fucose fucose and and UEA-I. UEA-I.

Similarly,Similarly, in inin order orderorder to to obtainto obtainobtain the the kineticthe kinetickinetic information informationinformation of the ofof interaction thethe interactioninteraction between betweenbetween Con A and ConCon mannose, AA andand mannose,amannose, diluted Con aa diluteddiluted A solution ConCon (6.25, AA solutionsolution 12.5, 20 (6.25,(6.25, and 4012.5,12.5,µg /2020mL) andand was 4040 injected μμg/mL)g/mL) to waswas the surfaceinjectedinjected of toto the thethe sensor. surfacesurface A globalofof thethe sensor.datasensor. fit AA was globalglobal performed datadata fitfit using waswas pepe Attanarformedrformed QCM usingusing Cell-A200 AttanaAttana QCMQCM evaluation Cell-A2Cell-A2 software,0000 evaluationevaluation based software, onsoftware, a theoretical basedbased 1:1onon ainteractiona theoreticaltheoretical model 1:11:1 interactioninteraction with a mass modelmodel transfer withwith limitation. aa massmass transfertransfer As shown limitation.limitation. in Figure AsAs 11shownshown, the blackinin FigureFigure line 11, indicates11, thethe blackblack the lineresponseline indicatesindicates when thethe we responseresponse injected 6.25–40 whenwhen weweµg injected/injectedmL of Con 6.25–406.25–40 A to theμμg/mLg/mL surface ofof ConCon of the AA sensor.toto thethe surfacesurface It covers ofof the thethe theoretical sensor.sensor. ItIt covers the theoretical 1:1 fit obtained by Attana evaluation software (red line). We obtained the 1:1covers fit obtained the theoretical by Attana 1:1 evaluationfit obtained software by Attana (red evaluation line). We obtainedsoftware the(red constants line). We of obtained interactions the constants of interactions between Con A and mannose, which are the association rate constant (Kon = betweenconstantsCon of interactions A and mannose, between which Con A are and the manno associationse, which rate are constant the associ (Kation= 2.71 rate constant104 M (K1son1 ),= 4 −1 −1 −3 −1 on − − 2.712.71 ×× 10104 MM−1ss−1),), thethe dissociationdissociation raterate constantconstant3 (K(Koff1off == 1.051.05 ×× 1010−3 ss−1)) andand affinityaffinity constantconstant× (K(KDD == 38.838.8 the dissociation rate constant (Koff = 1.05 10− s− ) and affinity constant (KD = 38.8 nM). nM).nM). ×

FigureFigure 11. 11. KineticKinetic evaluationevaluation of of the the interaction interaction between between mannose mannose and and Con Con A. A.

4. Conclusions 4.4. ConclusionsConclusions InIn this this study,study, aaa universaluniversaluniversal methodmethodmethod forforfor photochemicalphotochephotochemicalmical covalentcovalent fabricationfabrication ofof thethe underivatizedunderivatized carbohydratescarbohydrates tototo prepare prepareprepare the thethe carbohydrate carbohydratecarbohydrate sensors sensorssensors based basedbased on onon PFPA PFPAPFPA modified modifiedmodified PDA PDAPDA sensors sensorssensors for real-time forfor real-real- timedetectiontime detectiondetection of carbohydrate-lectin ofof carbohydrate-lectincarbohydrate-lectin interactions interaintera byctionsctions QCM biosensorbyby QCMQCM was biosensorbiosensor developed. waswas The developed.developed. photochemical TheThe photochemicalfabricationphotochemical of PFPA fabricationfabrication was able ofof toPFPAPFPA prepare waswas carbohydrate ableable toto prprepareepare sensors carbohydratecarbohydrate by directly sensorssensors inserting byby thedirectlydirectly underivatized insertinginserting thecarbohydratesthe underivatizedunderivatized on carbohydratescarbohydrates the sensor surfaces. onon thethe sensor Thesensor SEM-EDX surfaces.surfaces. energy TheThe SEM-EDXSEM-EDX spectrum energyenergy indicated spectrumspectrum that theindicatedindicated PFPA thatwasthat successfullythethe PFPAPFPA waswas connected successfullysuccessfully to the sensorconnectedconnected surface. toto thethe Four sensorsensor different surface.surface. underivatized FourFour differentdifferent carbohydrates underivatizedunderivatized were carbohydratesimmobilizedcarbohydrates on werewere the immobilizedimmobilized PFPA modified onon thethe PDA PFPAPFPA surfaces modifiedmodified by PDAPDA photochemical surfacessurfaces byby fabrication photochemicalphotochemical to prepare fabricationfabrication the tocarbohydrateto prepareprepare thethe sensors carbohydratecarbohydrate for further sensorssensors detecting forfor furtherfurther their detectidetecti interactionsngng theirtheir with interactionsinteractions five diff erentwithwith fivefive lectins differentdifferent (WGA, lectinslectins SBA, (WGA,PNA,(WGA, UEA-I, SBA,SBA, PNA, ConPNA, A). UEA-I,UEA-I, The experimental ConCon A).A). TheThe results experimentalexperimental show that resultsresults the carbohydrate-lectin showshow thatthat thethe carbohydrate-lectincarbohydrate-lectin interactions are interactionsspecific,interactions which areare arespecific,specific, consistent whichwhich with areare consistentconsistent the results withwith obtained thethe resultsresults by other obtainedobtained methods. byby otherother A kinetic methods.methods. study AA kinetickinetic of the studystudy ofof thethe interactionsinteractions betweenbetween mannosemannose anandd ConCon A,A, fucosefucose andand UEA-IUEA-I werewere performed.performed. OurOur workwork providesprovides anan efficientefficient wayway forfor thethe fabricationfabrication ofof aa QCMQCM carbohydratecarbohydrate sensorsensor byby underivatizedunderivatized

Polymers 2019, 11, 1023 11 of 13 interactions between mannose and Con A, fucose and UEA-I were performed. Our work provides an efficient way for the fabrication of a QCM carbohydrate sensor by underivatized carbohydrates based on PFPA modified PDA surfaces, avoiding a tedious modification of carbohydrates, which makes the QCM biosensor analysis more efficient.

Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4360/11/6/1023/s1. Author Contributions: L.G. performed experiments, analyzed data and assisted in drafting the manuscript; S.C. analyzed data and assisted in drafting the manuscript; P.H. and X.L. performed the experiments; Q.S. performed a small part of the experimental work; C.W. participated in the design of the paper and revised the manuscript; Y.P. assisted with the discussion and revision of the manuscript; Z.P. supervised the work presented in this paper and revised the manuscript. Acknowledgments: The National Natural Science Foundation of China (21877088 and 21572181) supported this research. Conﬂicts of Interest: No conﬂict of interest is declared by the authors.

References

1. Ohtsubo, K.; Marth, J.D. Glycosylation in cellular mechanisms of health and disease. Cell 2006, 126, 855–867. [CrossRef][PubMed] 2. Park, S.; Lee, M.R.; Pyo, S.J.; Shin, I. Carbohydrate chips for studying high-throughput carbohydrate-protein interactions. J. Am. Chem. Soc. 2004, 126, 4812–4819. [CrossRef][PubMed] 3. Tyagi, A.; Wang, X.; Deng, L.; Ramström, O.; Yan, M. Photogenerated carbohydrate microarrays to study carbohydrate-protein interactions using surface plasmon resonance imaging. Biosens. Bioelectr. 2010, 26, 344–350. [CrossRef][PubMed] 4. Rodahl, M.; Höök, F.; Krozer, A.; Brzezinski, P.; Kasemo, B. Quartz crystal microbalance setup for frequency and q-factor measurements in gaseous and liquid environments. Rev. Sci. Instrum. 1995, 66, 3924–3930. [CrossRef] 5. Li, X.; Pei, Y.; Zhang, R.; Shuai, Q.; Wang, F.; Aastrup, T.; Pei, Z. A suspension-cell biosensor for real-time determination of binding kinetics of protein-carbohydrate interactions on cancer cell surfaces. Chem. Commun. 2013, 49, 9908–9910. [CrossRef][PubMed] 6. Pei, Z.; Larsson, R.; Aastrup, T.; Anderson, H.; Lehn, J.M.; Ramström, O. Quartz crystal microbalance bioaffinity sensor for rapid identification of glycosyldisulfide lectin inhibitors from a dynamic combinatorial library. Biosens. Bioelectr. 2006, 22, 42–48. [CrossRef][PubMed] 7. Pei, Y.; Yu, H.; Pei, Z.; Theurer, M.; Ammer, C.; Andre, S.; Gabius, H.J.; Yan, M.; Ramström, O. Photoderivatized polymer thin films at quartz crystal microbalance surfaces: Sensors for carbohydrate-protein interactions. Anal. Chem. 2007, 79, 6897–6902. [CrossRef] 8. Li, X.; Song, S.; Shuai, Q.; Pei, Y.; Aastrup, T.; Pei, Y.; Pei, Z. Real-time and label-free analysis of binding thermodynamics of carbohydrate-protein interactions on unfixed cancer cell surfaces using a QCM biosensor. Sci. Rep. 2015, 5, 14066. [CrossRef] 9. Pei, Z.; Aastrup, T.; Anderson, H.; Ramström, O. Redox-responsive and calcium-dependent switching of glycosyldisulfide interactions with Concanavalin A. Bioorganic Med. Chem. Lett. 2005, 15, 2707–2710. [CrossRef] 10. Song, S.; Lu, Y.; Li, X.; Cao, S.; Pei, Y.; Aastrup, T.; Pei, Z. Optimization of 3D surfaces of dextran with different molecule weights for real-time detection of biomolecular interactions by a QCM biosensor. Polymers 2017, 9, 409. [CrossRef] 11. Lu, Y.; Song, S.; Hou, C.; Pang, S.; Li, X.; Wu, X.; Shao, C.; Pei, Y.; Pei, Z. Facile fabrication of branched-chain carbohydrate chips for studying carbohydrate-protein interactions by QCM biosensor. Chin. Chem. Lett. 2018, 29, 65–68. [CrossRef] 12. Shang, K.; Song, S.; Cheng, Y.; Guo, L.; Pei, Y.; Lv, X.; Aastrup, T.; Pei, Z. Fabrication of carbohydrate chips based on polydopamine for real-time determination of carbohydrate-lectin interactions by QCM biosensor. Polymers 2018, 10, 1275. [CrossRef][PubMed] 13. Norberg, O.; Deng, L.; Aastrup, T.; Yan, M.; Ramstrom, O. Photo-click immobilization on quartz crystal microbalance sensors for selective carbohydrate-protein interaction analyses. Anal. Chem. 2010, 83, 1000–1007. [CrossRef][PubMed] Polymers 2019, 11, 1023 12 of 13

14. Norberg, O.; Deng, L.; Yan, M.; Ramström, O. Photo-click immobilization of carbohydrates on polymeric surfaces-A quick method to functionalize surfaces for biomolecular recognition studies. Bioconj. Chem. 2009, 20, 2364–2370. [CrossRef][PubMed] 15. Peumans, W.J. Lectins as plant defense proteins. Plant Physiol. 1995, 109, 347–352. [CrossRef][PubMed] 16. Coulibaly, F.S.; Youan, B.B.C. Concanavalin a-polysaccharides binding affinity analysis using a quartz crystal microbalance. Biosens. Bioelectr. 2014, 59, 404–411. [CrossRef] 17. Wang, X.; Ramström, O.; Yan, M. A photochemically initiated chemistry for coupling underivatized carbohydrates to gold nanoparticles. J. Mater. Chem. 2009, 19, 8944–8949. [CrossRef] 18. Otman, O.; Boullanger, P.; Lafont, D.; Hamaide, T. New amphiphilic glycopolymers based on a polycaprolactone-maleic anhydride copolymer backbone: characterization by 15N NMR and application to colloidal stabilization of nanoparticles. Macromol. Chem. Phys. 2008, 209, 2410–2422. [CrossRef] 19. Pei, Z.; Dong, H.; Caraballo, R.; Ramström, O. Synthesis of Positional Thiol Analogs of β-d-Galactopyranose. Eur. J. Org. Chem. 2007, 2007, 4927–4934. [CrossRef] 20. Davis, B.G.; Maughan, M.A.; Green, M.P.; Ullman, A.; Jones, J.B. Glycomethanethiosulfonates: powerful reagents for protein glycosylation. Tetrahedron Asymmetry 2000, 11, 245–262. [CrossRef] 21. Sivapriya, K.; Suguna, P.; Chandrasekaran, S. Novel cyclic tetraselenides of mannose: Synthesis and mechanistic studies. Tetrahedron Lett. 2007, 48, 2091–2095. [CrossRef] 22. Norberg, O.; Lee, I.H.; Aastrup, T.; Yan, M.; Ramström, O. Photogenerated lectin sensors produced by thiol-ene/yne photo-click chemistry in aqueous solution. Biosens. Bioelectr. 2012, 34, 51–56. [CrossRef] [PubMed] 23. Keana, J.F.; Cai, S.X. New reagents for photoaffinity labeling: Synthesis and photolysis of functionalized perfluorophenyl azides. J. Org. Chem. 1990, 55, 3640–3647. [CrossRef] 24. Yan, M.; Bartlett, M.A. Micro/nanowell arrays fabricated from covalently immobilized polymer thin films on a flat substrate. Nano Lett. 2002, 2, 275–278. [CrossRef] 25. Liu, L.; Yan, M. A general approach to the covalent immobilization of single polymers. Angew. Chem. Int. Ed. 2006, 45, 6207–6210. [CrossRef][PubMed] 26. Bartlett, M.A.; Yan, M. Fabrication of polymer thin films and arrays with spatial and topographical controls. Adv. Mater. 2001, 13, 1449–1451. [CrossRef] 27. Liu, L.; Engelhard, M.H.; Yan, M. Surface and interface control on photochemically initiated immobilization. J. Am. Chem. Soc. 2006, 128, 14067–14072. [CrossRef][PubMed] 28. Toma, M.; Tawa, K. Polydopamine thin films as protein linker layer for sensitive detection of interleukin-6 by surface plasmon enhanced fluorescence spectroscopy. ACS Appl. Mater. Interface 2016, 8, 22032–22038. [CrossRef][PubMed] 29. Lee, H.; Dellatore, S.M.; Miller, W.M.; Messersmith, P.B.Mussel-inspired surface chemistry for multifunctional coatings. Science 2007, 318, 426–430. [CrossRef] 30. Liu, Y.; Ai, K.; Lu, L. Polydopamine and its derivative materials: Synthesis and promising applications in energy, environmental, and biomedical fields. Chem. Rev. 2014, 114, 5057–5115. [CrossRef] 31. Wei, Q.; Becherer, T.; Mutihac, R.C.; Noeske, P.L.M.; Paulus, F.; Haag, R.; Grunwald, I. Multivalent anchoring and cross-linking of mussel-inspired antifouling surface coatings. Biomacromolecules 2014, 15, 3061–3071. [CrossRef][PubMed] 32. Park, J.; Brust, T.F.; Lee, H.J.; Lee, S.C.; Watts, V.J.; Yeo, Y. Polydopamine-based simple and versatile surface modification of polymeric nano drug carriers. ACS Nano 2014, 8, 3347–3356. [CrossRef][PubMed] 33. Luo, R.; Tang, L.; Wang, J.; Zhao, Y.; Tu, Q.; Weng, Y.; Shen, R.; Huang, N. Improved immobilization of biomolecules to quinone-rich polydopamine for efficient surface functionalization. Colloids Surf. B Biointerfaces 2013, 106, 66–73. [CrossRef][PubMed] 34. Frick, C.P.; Merkel, D.R.; Laursen, C.M.; Brinckmann, S.A.; Yakacki, C.M. Copper-coated liquid-crystalline elastomer via bioinspired polydopamine adhesion and electroless deposition. Macromol. Rapid Commun. 2016, 37, 1912–1917. [CrossRef][PubMed] 35. Wu, C.; Li, X.; Song, S.; Pei, Y.; Guo, L.; Pei, Z. QCM biosensor based on polydopamine surface for real-time analysis of the binding kinetics of protein-protein interactions. Polymers 2017, 9, 482. [CrossRef][PubMed] 36. Yadav, M.; Sarkar, T.K.; Obot, I.B. Carbohydrate compounds as green corrosion inhibitors: Electrochemical, XPS, DFT and molecular dynamics simulation studies. RSC Adv. 2016, 6, 110053–110069. [CrossRef] Polymers 2019, 11, 1023 13 of 13

37. Williams, B.A.; Chervenak, M.C.; Toone, E.J. Energetics of lectin-carbohydrate binding. a microcalorimetric investigation of concanavalin a-oligomannoside complexation. J. Biol. Chem. 1992, 267, 22907–22911. [PubMed] 38. Lis, H. Specific inhibition by N-acetyl-d-galactosamine of the interaction between soybean agglutinin and animal cell surfaces. Biochim. Biophys. Acta 1970, 211, 582–585. [CrossRef] 39. Pereira, M.E.; Kabat, E.A.; Sharon, N. Immunochemical studies on the specificity of soybean agglutinin. Carbohydr. Res. 1974, 37, 89–102. [CrossRef] 40. Fox, T.; De Miguel, E.; Mort, J.S.; Storer, A.C. Potent slow-binding inhibition of cathepsin B by its propeptide. Biochemistry 1992, 31, 12571–12576. [CrossRef] 41. Banerjee, R.; Das, K.; Ravishankar, R.; Suguna, K.; Surolia, A.; Vijayan, M. Conformation, protein-carbohydrate interactions and a novel subunit association in the refined structure of peanut lectin-lactose complex. J. Mol. Biol. 1996, 259, 281–296. [CrossRef][PubMed] 42. Lis, H.; Sharon, N. Lectins: carbohydrate-specific proteins that mediate cellular recognition. Chem. Rev. 1998, 98, 637–674. [CrossRef][PubMed]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).