Bioactive and adhesive properties of multilayered coatings ... · Bioactive and adhesive properties of multilayered coatings based on catechol-functionalized chitosan/hyaluronic acid

International Journal of Biological Macromolecules 157 (2020) 119–134

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

j ourna l homepage: ht tp : / /www.e lsev ie r .com/ locate / i jb iomac

Bioactive and adhesive properties of multilayered coatings based oncatechol-functionalized chitosan/hyaluronic acid and bioactiveglass nanoparticles

Ana Catarina Almeida a,b,1, Ana Catarina Vale a,b,⁎,1, Rui L. Reis a,b,c, Natália M. Alves a,b,⁎a 3B's Research Group, I3Bs – Research Institute on Biomaterials, Biodegradables and Biomimetics, University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineeringand Regenerative Medicine, AvePark, 4805-017 Barco, Guimarães, Portugalb ICVS/3B's PT Associate Laboratory, Guimarães, Portugalc The Discoveries Centre for Regenerative and Precision Medicine, Headquarters at University of Minho, Avepark, 4805–017 Barco, Guimarães, Portugal.

⁎ Corresponding authors at: 3B's Research Group,Biomaterials, Biodegradables and Biomimetics of Univerthe European Institute of Excellence on Tissue EngineeriAvePark – Parque de Ciência e Tecnologia, 4805-017 Barc

E-mail addresses: [email protected] (A.C. V(N.M. Alves).

1 Co-first authors.

https://doi.org/10.1016/j.ijbiomac.2020.04.0950141-8130/© 2020 Elsevier B.V. All rights reserved.

a b s t r a c t
a r t i c l e i n f o
Article history:Received 28 February 2020Received in revised form 7 April 2020Accepted 13 April 2020Available online 23 April 2020

Keywords:Natural polysaccharidesBioactive glassMultilayered adhesive coatings

Chitosan and hyaluronic acid are the most attractive natural polysaccharides used for tissue regeneration, hereininnovative orthopedic coatings were constructed by dip-coating technique. Inspired by the tough nacre-likestructure, multifunctional (MF) films were constructed using bioactive glass nanoparticles (BGNPs) as the inor-ganic phase and hyaluronic acid (HA) and chitosan (CHT) polymers as the organic phase. Polymeric (CTR) filmswere also built with both polysaccharides. Inspired by themarinemussel's adhesive proteins, itwas the first timethat multilayered coatings containing both HA and CHT catechol conjugates were combined with BGNPs. Bothcatechol-conjugates were successful synthesized and, particularly for HA, it was possible to achieve the doubleof the substitution degree varying the reaction time. Prior to the LbL build-up, viscosity and Zeta potential mea-surements of the polyelectrolytes were conducted. The in-situ LbL growth of the films was monitored by quartzcrystal microbalance with dissipation monitoring. It was found that the combination of both catechol conjugatesresulted in amore compact LbL structure. It was also shown thatMF evidenced bioactivity, CTR presented an im-proved adhesion, and preliminary cellular tests confirmed the biomedical potential of these multilayered coat-ings being used in orthopedic implants.

© 2020 Elsevier B.V. All rights reserved.

1. Introduction

Mimicking structures and functions from natural organisms, as themarine mussels, to produce high-performing and environmentallyfriendlymaterials, has recently been a hot research topic due to their su-perior mechanical and biological properties compared to synthetic ma-terials [1–6]. Systems inspired on the surface chemistry and the roughnacre structure of molluscs have attracted much attention, since theyprovide the construction of robust layer-by-layer (LbL) films, withhigh mechanical stability and other interesting functionalities [7–11].Nacre is a natural composite of hard shells of molluscs with outstandingmechanical properties due to its hierarchical layered structure contain-ing an inorganic matrix of aragonite and an organic matrix of proteinsand polysaccharides [6–8]. Since its toughness is essentially provided

I3Bs – Research Institute onsity of Minho, Headquarters ofng and Regenerative Medicine,o, Guimarães, Portugal.ale), [email protected]

by the organicmatrix that plays an important role in dissipating theme-chanical energy, owing to its capability for undergoing inelastic defor-mations [8], in the present work were developed coatings with theinorganic–organic layered structure but with inverted nacre-like com-position. On the other hand, the strong underwater adhesion of marinemussels to almost all types of surfaces has attracted much attention[1,9,12]. It has been confirmed that the ortho-dihydroxyphenyl (cate-chol) moiety of the amino acid 3,4-dihydroxy-phenylalanine (DOPA)present in secreted mussel's adhesive proteins (MAPs) is responsiblefor the strong adhesion between adhesive pads ofmussels andopposingsurfaces [1,12–14]. The catechol group had an extremely strong affinityto various organic/inorganic surfaces such asmetals,metal oxides, poly-mer surfaces, and even biomacromolecules, such as pig gastric mucinglycoprotein [15,16]. In addition, researchers have been using catecholchemistry to modify several types of flat substrates [2,17], particles[18,19] and to improve the mechanical properties of polymer compos-ites [6].

Envisaging biomedical applications, namely the orthopedic implantsurface fabrication, surface coatings have been extensively investigatedfor the development ofmultifunctional implants capable to enhance therate of osseointegration, to improve corrosion resistance and to prevent

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijbiomac.2020.04.095&domain=pdfhttps://doi.org/10.1016/j.ijbiomac.2020.04.095mailto:[email protected]:[email protected]://doi.org/10.1016/j.ijbiomac.2020.04.095http://www.sciencedirect.com/science/journal/01418130http://www.elsevier.com/locate/ijbiomac

120 A.C. Almeida et al. / International Journal of Biological Macromolecules 157 (2020) 119–134

implant-related infections [20–27]. In this regard, layer-by-layer is aversatile and efficient technique for surface functionalization of differ-ent materials, such as polymers and metals, being a promising tool todevelop drugdelivered systems for controlled and local release of bioac-tive and/bactericidal agents in the surroundings of the implant[20,28–32]. Effectively, polyelectrolyte multilayered coatings can incor-porate organic and inorganic compounds, and be specifically tailored inorder to increase the biocompatibility of distinct implants or micro/nanoparticles. In this sense, chitosan- and hyaluronic acid-based multi-layers have attracted particular attention, essentially due to their biode-gradability, biocompatibility and also biofilm repelling capacity [20,21].Particularly, dip-coated CHT/HA multilayered systems have been stud-ied on different polymeric and metallic surfaces, resulting in very inter-esting properties: Richert et al. [33] constructed CHT/HA coatings ontoglass that reduced bacterial adhesion; Fu et al. [34] developed CHT/hep-arin coatings on polyethylene terephthalate films with contact killingcapacity; Jou et al. [35] treated polyethylene terephthalate fibers withCHT/HA, which exhibited antibacterial activity and improved fibroblastproliferation; Shi et al. [36] described the treatment of Ti alloy with do-pamine that allowed the posterior functionalization with chitosan coat-ing by covalent reaction with glutaraldehyde; Bongaerts et al. [37]developed CHT/HA constructions on polydimethylsiloxane thatprevented protein adsorption; Bai et al. [26] described the constructionof CHT/magnesium phosphate coatings on magnesium alloys as an al-ternative strategy to slow the degradation of magnesium alloys;Zhang et al. [38] described the built up of CHT/dopamine-modified HAfilms on the surface of Ti alloys, which evidenced enhanced osteoblastproliferation; Ly et al. [29] developed CHT/alginate multilayers on tita-nium substrate that showed antibacterial properties; Zhong et al. [25]described the deposition of cellulose acetate nanofibers coated by hy-droxyapatite nanoparticles and chitosan on stainless steel plates, thatpromoted the formation of apatite layer; Govindharajulu et al. [22] de-scribed an alternative technique to promote osseointegration, throughCHT/peptide aptamermultilayered coatings on Ti substrate, which com-bined enhanced osteoblast growing with antibacterial activity; andmore recently, Valverde et al. [20] reported CHT/HA multilayers loadedwith an antibacterial agent onto smooth and micropatterned Ti alloyspreviously functionalized with free amino moieties, which revealedgood antibacterial properties.

So, inspired by the inorganic–organic nacre-like structure and byMAPs, we proposed new robust multifunctional (MF) chitosan/hyaluronic acid LbL films developed by the alternate combination of in-organic nanoparticles with catechol-functionalized biopolymeric layersto obtain biomimetic biomedical adhesive coatings. In parallel,biopolymeric films containing catechol groups were also developed astheir controls (CTR). Conventional layer-by-layer through dip-coatingwas chosen for film build-up, since it is simple and versatile methodol-ogy to construct thin films with complex architecture andmultifunctionalities, allowing to tune the film properties at the nano-scale length [10,11,39–41], being suitable for coating substrates withdifferent compositions, topographies and complex geometries [42,43].In this study, natural polyelectrolytes were catechol-functionalized asthe organic phase in the LbL assembly, aiming to enhance thefilm stabil-ity and improving adhesion strength. Regarding the negatively chargedpolymer used in this study, HA was chosen because is also part of thenative ECM and plays an important role in mechanical and structuralsupport, and it has been used in several biomedical applications [44].On the other hand, as positively charged polymer, chitosan is a partiallydeacetylated derivative from chitin, with known biocompatibility andbiodegradability [45,46]. As Wang and coworkers [41] already demon-strated, the catechol functionalization of CHT and HA throughcarbodiimide chemistry preserved the cationic and anionic characteris-tics of these polymers, as well as their electrostatic interactions, provid-ing a successful construction free-standing adhesive films. In addition,as Lu et al. [47] demonstrated through QCM study, multilayered con-structions with CHT and HA functionalized with catechol groups

presented an intrinsic stabilization by catechol interactions and a pH re-sponsiveness that can be useful for a controlled drug delivery. As inor-ganic phase of the MF LbL coatings, bioactive glass nanoparticles(BGNPs) were incorporated, since is expected that their inclusion willimprove their mechanical stability and also provide a bioactive charac-ter for the produced coatings. Hence, similarly with some calciumphos-phate coatings, these BGNPs will induce bone-bonding ability to theseadhesive coatings, improving the osseointegration on the implant sur-face [48–52].

Moreover, previous studies of our group [53–55] have already re-ported dip-coated LbL films onto glass substrate based on catechol-modifiedHA (HA-C), CHT and BGNPs. Itwas found that the combinationof these materials resulted in nanostructured films with enhanced ad-hesion, bioactivity and mechanical performance [53–55]. Then, we hy-pothesized that the construction of multilayer films with bothbiopolymers functionalized with catechol group could result in filmswith superior adhesive strength than the ones developed in these previ-ous studies, as the content of catechol groups will be higher. Based onprevious works [56–58], we have optimized the modification of CHTwith catechol groups, in order to obtain a catechol-modified CHT(CHT-C) with similar degree of substitution, DS (%). We have also opti-mized HA-C synthesis to achieve higher DS (%). In a recent work [59],our group reported the construction of CHT-C/HA-C multilayered coat-ings through two LbL methodologies (dip and spin-coating), usingonly biopolymers with higher catechol content, and compare their ef-fects on glass, titanium and stainless steel substrates. Briefly, it wasdemonstrated that the combination of CHT-C, HA-C and bioactive nano-particles contributed for a modulation of their surface topographyproperties.

In the present work, we focused on dip-coating LbL methodologyonto glass substrate, and extensively studied the construction of distinctconfigurations of multifunctional (MF) and polymeric (CTR) multilayercoatings, evaluating the effect of the DS (%) on the physicochemical andbiological properties of the constructed bioadhesive coatings. We be-lieve that these biocompatible MF films could be used as protectivecoatings of a variety of orthopedic devices and bone tissue engineeringscaffolds, to create an environment compatible with osteogenesis thatpromotes a bone-friendly interface with improved adhesion properties.Thus, by using MF films, a natural bonding junction between the im-plant and host's bone could be established in a simple and versatileway, avoiding the traditional use of cements. On the other hand, an im-proved bonding between distinct implants and non-bioactive tissuescould be achieved by using the developed highly adhesive biopolymericCTR films, as an alternative to the synthetic tissue adhesives available,which still present some limitations, such as their relative low adhesiveproperties in contact with biological fluids [60].

2. Materials and experimental methods

2.1. Materials

Medium molecular weight chitosan (CHT, viscosity 200–800 cP,Mw = 190–310 kDa, 75–85% N-deacetylation degree), hyaluronic acidsodium salt from Streptococcus equi (HA,Mw=1500–1800 kDa), dopa-mine hydrochloride (DN, Mw = 189.64 Da), hydrocaffeic acid (Mw =182.17 Da), N-(3-(dimethylamino)propyl)-N′-ethylcarbodiimide hy-drochloride (EDC, Mw = 191.70 Da), dialysis tubing cellulose mem-brane (avg. flat width 33 mm), calcium nitrate tetrahydrate, citric acidmonohydrate, ammonium phosphate dibasic, ethanol absolute, ammo-niumhydroxide solution, phosphate buffered saline (PBS), sodiumchlo-ride, magnesium chloride hexahydrate, sodium sulfate, tris(hydroxymethyl)aminomethane, hydrochloric acid (HCl),polyethylenimine (PEI) and hydrocaffeic acid (HCA, 3,4-dihydroxyhydrocinnamic acid) were purchased from Sigma-Aldrich(St. Louis, Mo, USA). Tetraethyl orthosilicate (TEOS) was purchasedfrom Merck KGaA (Darmstadt, Germany). Potassium chloride (KCl),

121A.C. Almeida et al. / International Journal of Biological Macromolecules 157 (2020) 119–134

acetone and 2-propanol were obtained from VWR International(UK). Sodium hydroxide (NaOH) was purchased from Fisher Chemi-cal (Fisher Scientific UK, Leics, UK) and hydrogen peroxide 30% (w/v)was obtained from Panreac AppliChem (Darmstadt, Deutschland).Sodium of hydrogen carbonate, di‑potassium hydrogen phosphatetrihydrate and calcium chloride were purchased from Merck(Merck Sharp & Dohme Corp., Kenilworth, NJ, USA). Standard goldQCM-D sensor crystals (QSX 301 Gold, Au 100 nm, 14 mm diameter)were purchased from Q-Sense (BiolinScientific, Stockholm, Sweden).For the cellular behaviour assays, the mouse fibroblast cell line L929was obtained from European Collection of cell cultures (ECACC, UK).Dulbecco's modified minimum essential medium (D-MEM), fetal bo-vine serum (FBS), phalloidin–tetramethylrhodamine B isothiocya-nate and DAPI (4′,6-diamidino2phenylindole) were purchasedfrom Sigma-Aldrich (St. Louis, MO, USA). MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium)was obtained from VWR International(UK). CHT was the only reagent that was previously purified by re-crystallization. Borosilicate glass plates (3 mm thickness), andround coverglasses (Ø 18 mm, Agar Scientific, Stansted, UK), wereused as substrates for deposition of LbL coatings by dip-coatingmethod. Prior to coating deposition, all the substrates were cleanedin sequential ultrasonic water bath treatments (15 min) to removesurface impurities: acetone; ethanol; osmotized water; and finally,they were dried with nitrogen flow.

2.2. HA-C synthesis

HA-C was synthesized using the procedure proposed by Lee andco-workers [2] with some modifications. HA modification with cate-chol groups was performed using EDC as an activation agent of thecarboxyl groups on HA chains. HA solution (10 mg/mL) was pre-pared in phosphate buffered saline (PBS) solution and the pH wasadjusted to 5.5. To limit the oxygen interaction with the solution,HA solution was purged with nitrogen for 30 min. Then, 2.5 mmolof EDC and DN were added to the previous HA solution, and the pHwas maintained at 5.5 at 4 °C. The resulting solution was divided intwo different solutions with different reaction times, i.e., HA-C4hand HA-C36h. Unreacted chemicals and urea byproducts were re-moved by dialysis (Mwco = 14 kDa) against an acidic aqueous solu-tion (pH 5.0, HCl solution) for 4 days and osmotized water for 1 day.Finally, the HA-C conjugates were freeze-dried for 4 days and storedat −20 °C. The entire procedure and storage of the produced HA-Cwas performed at 4 °C and protected from light to prevent oxidationof catechol groups.

2.3. CHT-C synthesis

CHT-C synthesis was based on the procedure adapted from the pro-tocols proposed by Kim et al. [56], Xu et al. [57] and Ghadban et al. [58].As in HA-C synthesis, CHT modification with catechol groups was ac-complished by the carbodiimide chemistry. CHT solution [1% (w/v)]was prepared in HCl aqueous solution at pH 2.5, then a HCA solution(59 mg/mL) was prepared in osmotized water and an EDC solution(119 mg/mL) was prepared in a mixture of osmotized water and etha-nol. These two previous solutions were mixed and added to CHT solu-tion, under stirring at 4 °C, followed by the addition of 1 M NaOHsolution to obtain a final pH of 4.8. The reactionwas allowed to continuefor 18 h, then the conjugate CHT-C18hwas purified by dialysis (Mwco=14 kDa) against an acidic aqueous solution (pH 5.0, HCl solution) for3 days and finally, only osmotized water for 4 h. The resulting productwas freeze-dried and stored at −20 °C. As in the HA-C synthesis, thisprocedure was performed at 4 °C and protected from light, in order toprevent oxidation of catechol groups.

2.4. BGNPs production

The procedure to obtain the ternary system of BGNPs with the com-position SiO2:CaO:P2O5 (mol.%) = 50:45:5, was based on the sol-gelmethod already optimized by two previousworks [48,61]. First, a “solu-tion A”was prepared through a mixture of precursor's solutions. So, 6%(w/v) of calcium nitrate tetrahydrate, calcium precursor, was dissolvedin osmotized water at room temperature. Then, 9.8353 mL of TEOS, sil-ica precursor, together with 60 mL of ethanol absolute were added tothe previous solution. The pH of solution A was adjusted to 2 with citricacid solution (10% (w/v)), under stirring for 3 h. After that, a “solution B”was also prepared by adding 0.07% (w/v) of ammonium phosphate di-basic, phosphorus precursor, to osmotized water. The pH of solution Bwas adjusted to 11.5 with ammonia hydroxide solution. Under stirring,the solution Awas slowly added, drop-by-drop, to solution B and the pHwasmaintained at 11.5 by continuous supplement of ammonia hydrox-ide solution. This reaction mixture was left under stirring during 48 hand then, aging for 24 h to occur the gel particle precipitation. After-wards, the gel precipitate was washed three times with osmotizedwater and stored at −80 °C to be subsequently freeze-dried for7 days. Finally, the obtained white gel powder was calcinated at700 °C for 3 h, in order to obtain a white powder of BGNPs with im-proved bioactivity.

2.5. UV–Vis spectrophotometry characterization

The degree of substitution, DS (%), of catechol groups in the conju-gates were determined using a Synergy HT Multi-Mode MicroplateReader (BioTek Instruments, U.S.A.) with an absorbance measurementrange of 200 to 350nm, and a quartzmicroplatewith 96wells. Solutionsof HA-C4h, HA-C36h and CHT-C18hwith different concentrations, 0.5, 1, 2,3, 4, and 5mg/mL in 0.15MNaCl, were prepared for the UV analysis and0.15 M NaCl solution was used as blank.

2.6. Rheological characterization

The viscosity of the different polyelectrolytes was determinedthrough rheological measurements performed on a Kinexus Pro Rhe-ometer (Malvern Instruments Ltd., UK) fitted with cone-plate geome-try: a cone with 40 mm diameter and 4° angle (CP4/40: SR 1772SS)and a plate with 65 mm diameter (PL65: S1425SS) were used. Rota-tional measurements of CHT, HA, CHT-C18h, HA-C4h and HA-C36h solu-tions at different concentrations of 0.5, 1, 2, and 3 mg/mL, in 0.15 MNaCl, weremade. The steady-state flowmeasurementswere performedunder controlled-stress conditions, where the torque amplitude wasimposed using a logarithmic ramp of shear rate ranging from 10 to100 s−1. All experiments were performed at a controlled temperatureof 25 °C, with three replicates per condition, being the experimentaldata registered with rSpace for Kinexus Pro 1.7 software.

2.7. Zeta potential (ζ) characterization

The Zeta potential of the different polyelectrolytes was also deter-mined at 25 °C through a Zetasizer equipment (Nano ZS, Malvern, UK)and the results were given as an average of three measurements foreach 0.5 mg/mL solution in 0.15 M NaCl solution. Immediately prior tomeasurement, the BGNPs suspensionwas dispersed for 15min in an ul-trasonic water bath (DT100H SONOREX, Bandelin electronic GmbH &Co. KG, Berlin, Deutschland) to prevent the nanoparticle's agglomera-tion and precipitation.

2.8. Quartz crystal microbalance with dissipation (QCM-D) monitoring

The build-up of LbL coatings with different materials wasmonitoredin situ byQCM-D (Q-sense, E4 system, Sweden) onto gold-coated quartzcrystals (14 mm diameter, QSX301 Gold, Q-Sense). QCM-D is an


accurate technique which makes use of piezoelectric quartz crystals todetect adsorption of molecules, at nanoscale, by measuring frequencychanges [62]. In addition, from the decay monitoring of the crystal's os-cillation is quantified the dissipation, which represents the viscoelasticproperties of the adsorbed mass [63]. So, through the QCM-D monitor-ing, it is possible to simultaneously measure the adsorbed amount,given by the normalized resonance frequency (Δf/ν) of each overtoneto the fundamental resonant, where ν is the overtone number, andthe variation of the viscoelastic properties given by energy dissipation(ΔD) of the multilayer film in real time [63].

Previously to QCM-D experiments, all quartz crystals were cleanedin an ultrasonic water bath (DT100H SONOREX, Bandelin electronicGmbH& Co. KG, Berlin, Germany) with sequential sonication for5–10min in: 2% (v/v) acetic acid solution; a 5:1:1 mixture of osmotizedwater, ammonia hydroxide (25%) and hydrogen peroxide (30%) at75 °C; acetone; isopropanol; ethanol; and ultrapure water. Finally, thecrystals were dried with nitrogen flow and treated with UV/Ozone for10 min (Bioforce Nanoscience, ProCleaner 220).

To ensure that the crystals are perfectly clean, all experimentsstarted with a 0.15 M NaCl baseline. The crystals were excited at multi-ple overtones, 3rd, 5th, 7th, 9th, and 11th, which corresponds respec-tively, to 15, 25, 35, 45, and 55 MHz fundamental resonant f.Adsorption took place at 25 °C, and at a constant flow rate of 50 μL/min using a peristaltic pump. LbL coatings (Fig. 1) were producedusing fresh solutions prepared in 0.15MNaCl and with the pH adjustedto 5.5. LbL construction started with injection of CHT (1mg/mL with 2%(v/v) of acetic acid) or CHT-C18h (1 mg/mL) used as polycations, whileHA (0.5 mg/mL), HA-C4h (0.5 mg/mL), HA-C36h (0.5 mg/mL), andBGNPs (2.5 mg/mL) acted as polyanions and were used as ended-polymer layer. Before the QCM experiments, the BGNPs suspensionwas dispersed in an ultrasonic water bath for 15 to 20 min to avoid ag-glomeration and precipitation of nanoparticles. The injection of thepolymers was standing for 10 min and the nanoparticles for 20 min toallow the adsorption until the equilibrium was achieved in QCM-Dmonitoring. A rising step of 5 min, between the polyelectrolyte adsorp-tions, using 0.15 M NaCl solution with the pH adjusted to 5.5 wasincluded.

During this in situmonitoring, the normalized frequency (Δf/ν) anddissipation (ΔD) shifts were continuously recorded as a function oftime, and the thickness of each LbL construction was further estimatedusing a viscoelastic Kelvin-Voigt model provided by the Q-Tools soft-ware (Q-Sense), which is based on the model described by Voinova

Fig. 1. Schematic illustration of the different multifu

et al. [64,65] From the modelled results (theoretical values), the QToolssoftware was also used to compare them with Δf and ΔD experimentalvalues to obtain the bestfitting using a simplex algorithm (Total ChiSqr)based on the minimum in the sum of the squares of the scaled errors.For this purpose, ρB (the liquid's density), ηB (the liquid's viscosity)and ρL (the density of the absorbed layer) should be changed in orderto get the lower error possible. In this study, a fluid density of1000 kg·m−3, a fluid viscosity of 1 mPa·s and a layer density of1200 kg·m−3 was assumed, and the Δf/ν and ΔD shifts were fittedbased on the mean of experimental results from three overtones.

2.9. LbL assembly of the coatings

The LbL assembly of distinct coatings was performed by dip-coatingusing fresh polyelectrolyte solutions prepared with 0.15 M NaCl solu-tion, at room temperature. Distinct polyelectrolytes solutions were pre-pared: 1 mg/mL CHT with 2% (v/v) of acetic acid; 0.5 mg/mL HA;1 mg/mL CHT-C18h; 0.5 mg/mL HA-C4h; 0.5 mg/mL HA-C36h;2.5 mg/mL BGNPs and 5 mg/mL PEI. Except for PEI solution, the pHwas adjusted to 5.5. Once again, BGNPs suspensions were kept understirring and periodically subjected to an ultrasonic treatment during15–20 min, in order to avoid BGNPs agglomeration.

Prior to coating deposition, glass substrateswere cleaned following aprocedure similar with the cleaning protocol described for QCM crys-tals: all substrates were rinsed with acetone, ethanol and osmotizedwater and then dried with a nitrogen flow before use.

In this study, the HA, HA-C4h, HA-C36h and BGNPs were used aspolyanion, while CHT and CHT-C18h were used as polycation. PEI wasused as an initial layer precursor, by immersing each glass substratefor 20 min, prior to multilayer deposition. After immersion in PEI, thesubstrates were alternately dipped in the oppositely-charged polyelec-trolyte solutions, to produce LbL coatings with 11 bilayers, i.e. 22 layers,and to be able to compare these results with those obtained in previousworks. A polycationwas used to initiate alternating deposition betweenoppositely charged polyelectrolytes.

Distinct configurations were developed as shown in Fig. 1: the MFLbL coatings containing [CHT/HA/CHT/BGNPs]5 + [CHT/HA] (MF1),[CHT-C18h/HA/CHT-C18h/BGNPs]5 + [CHT-C18h/HA] (MF2), [CHT/HA-C4h/CHT/BGNPs]5 + [CHT/HA-C4h] (MF3), [CHT/HA-C36h/CHT/BGNPs]5 + [CHT/HA-C36h] (MF4), [CHT-C18h/HA-C4h/CHT-C18h/BGNPs]5 + [CHT-C18h/HA-C4h] (MF5), [CHT-C18h/HA-C36h/CHT-C18h/BGNPs]5 + [CHT-C18h/HA-C36h] (MF6), and respective controls (CTR)

nctional (MF) and control (CTR) LbL coatings.


containing [CHT/HA]11 (CTR1), [CHT-C18h/HA]11 (CTR2), [CHT/HA-C4h]11 (CTR3), [CHT/HA-C36h]11 (CTR4), [CHT-C18h/HA-C4h]11 (CTR5), and[CHT-C18h/HA-C36h]11 (CTR6).

All substrates used in the experiments were rinsed with acetone,ethanol and osmotized water and then dried with a nitrogen flow be-fore use. The dipping time for the CHT, HA and their respective conju-gates (CHT-C and HA-C) was 10 min, whereas for the BGNPssuspension it was 20 min. These dipping times were based on the twopreviously mentioned works, where it was verified by QCM-D studiesthat these times corresponded to themomentwhen the polyelectrolyteadsorption reached equilibrium. In addition, a rinsing stepwas includedfor 5 min with 0.15 M NaCl solution, between the adsorptions of eachpolyelectrolyte.

2.10. Tensile shear strength tests

The adhesive performance of LbL coatings fabricated onto glassplates was evaluated through the lap-shear strength tests, accordingwith the procedure described in the ASTM D1002 standard. Thesetests were carried out in a universal mechanical testing machine(Instron 5543, USA) with a 1 kN load cell. Immediately after the LbL de-position, pairs of glass plates were put in contact with an overlappingarea of 15 × 20 mm2. The specimens were tightly clamped usingmetal binder clips and maintained at 37 °C overnight. For adhesion ex-periments, samples were appropriately placed between the grips of themechanical testingmachine and anunaxial tensile loadwas applied, at aconstant cross-head speed of 5 mm/min, until their detachment. Fivespecimens of each conditionwere tested, at room temperature, to calcu-late the mean and standard deviation values of the ultimate adhesionstrength. The testing software used was the Bluehill2 (INSTRON Corpo-ration), and the stress-strain curves was obtained for each testing con-dition from the maximum tensile stress (σTmax) and the maximumtensile strain (εTmax) calculations.

The σTmax is the maximum stress that the film can support withoutdetachment, which is obtained from the load applied (F) data, acquiredduring the test, and the overlapping area (A0), Eq. (1). The εTmax corre-sponds to the strain value reached for theσTmax, which is obtained fromthe ratio between the displacement (ΔL) data and the gauge length (L0)Eq. (2) [55].

σTmax ¼ FA0 ð1Þ

εTmax ¼ ΔLL0 ð2Þ

2.11. In-vitro bioactivity studies

The bone-bonding ability of a material can be asserted in-vitro byevaluating the formation of apatite on its surface, when immersed insimulated body fluid (SBF), which contains ionic concentrations similarwith the human blood plasma [66].

Thus, standard in-vitro bioactivity studies were performed by im-mersing for 7 and 14 days at 37 °C, the dip-coated LbL coatings in aSBF solution with pH value adjusted to 7.4, which was prepared byfollowing the Kokubo and Takadama procedure. After removingfrom SBF solution, these coverglasses were cleaned three timeswith ultrapure water and dried at room temperature. The bone-likeapatite layer formation was characterized by a surface analysisusing the scanning electron microscopy (SEM, JSM-6010 LV, JEOL,Japan) coupled with energy dispersive Xray spectroscopy (EDS,INCAx-Act, PentaFET Precision, Oxford Instruments) and X-ray dif-fraction (XRD, Bruker AXS D8, Discover, USA) technologies. BeforeSEM analysis the samples were sputtered with a thin platinumlayer, using a sputter coater EM ACE600 (Leica Microsystems,

Germany). The XRD experiments were performed at 40 kV and40 mA using Cu Kα radiation (λ = 1.54060 Å). The XRD detectorwas scanned over a range of 2θ angles from 15° to 60° at a speed of0.04°/s. The crystalline phase identification was achieved using ananalytical software EVA and were indexed using the ICDD database(International Centre for Diffraction Data).

2.12. In-vitro biological studies

Studies of morphology and cellular activity were performed to eval-uate the in-vitro biological performance of the dip-coated LbL conditionsbuild-up onto round coverglasses with 1.33 cm2.

The mouse fibroblast cell line L929, from European Collection ofcell cultures (ECACC, UK), was chosen to perform direct contacttests with the developed LbL coatings. Before cell seeding, the sam-ples were treated with UV light for 30 min and immersed in ethanol70% for 30 min and then, they were washed twice with sterile PBS.Cells were cultured with Dulbecco's modified minimum essentialmedium (D-MEM) supplemented with 10% fetal bovine serum(FBS) and 1% antibiotic. The cultures were then incubated at 37 °C,in humidified air atmosphere of 5% CO2, and placed to grow untilconfluence. The culture medium was replaced every 2 days. When90% of confluence was reached, the cells were seeded onto the LbLcoatings, using triplicates (n = 3), at a density of 1 × 104 cells persample and then incubated at 37 °C for 1, 3 and 7 days. After 3 h, sup-plemented D-MEM was added to each well to nourish the adheredcells.

After specific time points (1, 3 and 7 days) a MTS (3-(4,5-dimethylthiazol-2-yl)-5(3-carboxymethoxyphenyl)2(4-sulfophenyl)-2H-tetrazolium) assay was performed to evaluate thecytotoxicity of the dip-coated LbL conditions, and to compare the rel-ative cellular viability (%) between each condition and the tissue cul-ture polystyrene (TCPS), a positive control. The LbL coatings wereimmersed with a solution composed by a 1:5 ratio of MTS reagentand D-MEM culture mediumwithout phenol red or FBS, and then in-cubated for period of 3 h at 37 °C. All cytotoxicity tests were con-ducted by using three replicates (n = 3). Finally, the opticaldensity (OD) was read at 490 nm on a multiwell microplate reader(Synergy HT, BioTek Instruments, USA).

Briefly, for each time point, the culture medium was removed, and10% formalin was added to each well in three replicates (n = 3) for30 min. Then, formalin was removed, and the wells of the plate werewashed with PBS. The samples were labelled with fluorescent stains:Phalloidin (phalloidin-tetramethylrhodamine B isothiocyanate), whichbinds to actin filaments staining the cytoskeleton of the cells in red,and DAPI (4′,6-diamidino2phenylindole), which binds to DNA regionsstaining cell nuclei in blue. For this, first, phalloidin was incubatedwith samples at 1:200 in PBS for 30 min and then, DAPI was added at1:1000 in PBS for 5 min. This procedure was done at room temperatureand protected from light. Finally, the samples were washed twice withPBS, left overnight, and then visualized in the dark using a fluorescencemicroscope (Transmitted andReflected LightMicroscopewithApotome2, Zeiss, Germany). The fluorescence images were acquired and proc-essed using AxioVision software version: Zeiss 2012 (Zeiss, Germany).

2.13. Statistical analysis

The results of all experiments were carried out at least in three rep-licates (n= 3) and were presented as mean± standard deviation. Sta-tistical significance between groups was determined by One-wayANOVA with Tukey's Multiple comparison test, using Graph Pad Prismversion 6.0 (GraphPad software, San Diego, CA). Statistical differenceswere represented and set to p b .05(*), p b .01(**), p b .001(***), andp b .0001(****).

Fig. 2. UV–Vis spectra for different concentrations of HA-C4h and HA-C36h.


3. Results and discussion

3.1. Synthesis and characterization of HA-C and CHT-C conjugates

HA and CHT were modified with catechol groups and, in order toconfirm if the modification of HA and CHT were successful, solutionsof HA-C4h, HA-C36h, and CHT-C18h with different concentrations wereanalyzed by UV–Vis spectroscopy. The results obtained for both HA-Cconjugates are shown in Fig. 2.

Both spectra of the HA-C conjugates exhibit a maximum absorbancepeak for all concentrations at a wavelength of approximately 280 nm,which ismore intense as the solution concentration increases. These re-sults confirm the presence of the catechol groups in the modified HA,due to the presence of their characteristic peak at a wavelength of280 nm [53,55], as opposed to theHA spectrum. In addition, the absenceof additional peaks at wavelengths longer than 300 nm demonstratesthat the synthesized conjugates were not oxidized [53,55,67].

Based on the results presented in Fig. 2, the DS (%) of the HA-C con-jugateswas estimated and their valueswere around 27% for HA-C4h and54% for HA-C36h. The DS (%) results obtained in the present work werehigher than those found in other works [2,53,55], which were onlyabout 11%. These differences can be explained by some different exper-imental conditions employed in the present HA-C synthesis, such as themolecular weight of HA and/or the time of the reaction. Based on thepresent study, it was found that the increase in the reaction time from4 to 36 h contributed to an increase in the DS(%) from 27% to 54%.

Fig. 3. UV–Vis spectra for different concentrations of CHT-C18h.

Similarly, the UV–Vis characterization was performed for CHT-C18hto confirm its successful modification and to obtain the DS (%). The ex-perimental results are presented in Fig. 3.

The CHT-C18h spectrum also exhibits a maximum absorbance peak,which increases its intensity with the increase of the solution concen-tration, at a wavelength around 280 nm, confirming the presence ofthe catechol groups in the modified CHT. Likewise, the absence of addi-tional peaks at wavelengths N300 nm shows that the CHT-C18h was notoxidized [56,57,68].

According to the DS (%) results obtained, the estimated value forCHT-C18h was approximately 9%. This value was slight lower thanthose found in other works [56–58], which could be due to differentmolar ratio of HCA/EDC, the molecular weight of CHT, dialysis proce-dure (time, pH, solvent) and also distinct reaction time used for theCHT-C synthesis in previous studies (4 h, 12 h, 18 h).

3.2. Rheological characterization

The effect of concentration of each PE used on its viscosity was eval-uated using a Kinexus pro+ rheometer (Malvern Panalytical Instru-ments, Lda, UK) equipped with a cone and plate geometry. The steadyflow curves (where the apparent viscosity is plotted in function of theshear rate) were obtained for different solutions of HA, HA-C4h, HA-C36h, CHT and CHT-C18h at 0.5, 1, 2 and 3 mg/mL, in 0.15 M NaCl, andthe relationship between the mean apparent viscosity and the concen-tration of each polymer was described by a trend line, and Fig. 4 illus-trates the trend line obtained for each PE at different concentrations.

As Fig. 4 presents, HA and HA-C evidence an exponential trend line,i.e. their viscosities increase exponentially with increasing concentra-tions, while CHT and CHT-C show a linear trend line, where viscosity in-creases almost linearlywith concentration. The exponential character ofHA has been reported in other studies [69,70], and this seems to happennot only with concentration but also with molecular weight. So, sincethe HA used in this study has a high molecular weight, it should exhibithighly viscous properties. For CHT, the linear increase of its viscositywith the concentration, aswell as,with the increase ofmolecularweightand the deacetylation degree has also been reported elsewhere [71,72].

Furthermore, the viscosity of all conjugates is lower than the corre-sponding values of the unmodified polymers. This tendency seems tobe more pronounced with increasing DS, since the viscosity reductionis higher for HA-C. The UV–Vis results showed that the HA-C had a DShigher than CHT-C. In turn, this effect is also evidenced by the higherviscosity reduction in HA-C36h compared to HA-C4h, which have a DSof 54% and 27%, respectively. The PE concentration used for LbL assem-bly was defined attending these results. Since the viscosity can directlyaffect the adsorbed amount of PE on the film during LbL process [73],similar viscosities were chosen to produce LbL coatings with a more

Fig. 4. Effect of concentration on the apparent viscosity of HA, HA-C4h, HA-C36h, CHT and CHT-C18h solutions and their respective trend lines: a) y= 2.2695e1.0442x (R2 = 0.9923); b) y =1.9618e0.8437x (R2=0.9989); c) y=1.9229e0.8375x (R2=0.9978); d) y=1.9802x+1.1596 (R2=0.9773); e) y=1.9973x+0.5683 (R2=0.93586). Dots represent themean values fromthree experimental points UV–Vis spectra for different concentrations of HA-C4h and HA-C36h.


homogeneous distribution. Therefore, 0.5 mg/mL solutions were pre-pared for HA and HA-C, and 1 mg/mL solutions for CHT and CHT-C.

3.3. Zeta potential (ζ)

Bymeasuring the Zeta potential, an indication of the degree of repul-sion between adjacent particles similarly charged, can be obtained. Avalue of±25mV can be taken as the arbitrary value to indicatewhetherthe particles have low or high charge. Thus, for Zeta potential valueshigher than±25mV, the particles tend to repeal contributing to the sta-bilization of the dispersion, while for values below ±25 mV, the parti-cles tend to aggregate leading to their precipitation [74]. The Zetapotential obtained from each PE used for the build-up of LbL coatingswas presented in Table 1.

As expected, Zeta potential values for CHT and its conjugate (CHT-C18h) were positive, whereas for HA and its conjugates (HA-C4h andHA-C36h) as well as for the BGNPs were negative, and with the sameorder of magnitude. Positive values for CHT and CHT-C are due to theirpositively charged amine groups.

On the other hand, negative charge values for HA and HA-C are dueto their negatively charged carboxylic groups. The negative charge ofthe Zeta potential for the BGNPs has already been reported in previousstudies [55,75]. BGNPs exhibited −20.5 ± 0.8 mV, which was similarto the value reported in a previous work [49], indicating that the pro-duced nanoparticles could present tendency to aggregation and precip-itation [74]. Lu et al. [76] demonstrated that a negative Zeta potential onthe BGNPs surface is crucial to promote the stable formation of an apa-tite layer. In addition, it has been suggested that surfaces with negativeZeta potential values have more important biological effects in-vivo[77], promoting bone cell attachment and proliferation than surfaceswithout or even with positive electric charge [78,79]. On the otherhand, both HA and CHT polymers presented lower Zeta potential valuesthan their conjugates (HA-C and CHT-C, respectively). In particular, HA-C4h showed a lower Zeta potential (−19.47 ± 1.70 mV) than HA-C36h(−21.17 ± 1.50 mV). These results suggest that the polymer modifica-tion with catechol groups contributes to the Zeta potential increase.

Table 1Zeta potential values for each polyelectrolyte solution with 0.5 mg/mL concentration.

CHT HA HA-C4H HA-C36H CHT-C18H BGNPS

MV 19.2± 1.0

−15.9± 1.2

−19.5± 1.7

−21.2± 1.5

24.1± 2.1

−20.5± 0.8

Hence, higher Zeta potential values of these polymers could provide amore stable LbL assembly.

3.4. QCM-D monitoring of LbL films construction

The LbL assembly of multilayered films with CHT, CHT-C, HA, HA-C,and BGNPs was monitored in situ using QCM-D, since this techniqueprovides the detection of small changes in the mass and viscoelasticproperties as the multilayers were adsorbed onto the quartz crystalsurface.

TheQCM-D results for the build-up of polymericfilms (controls) andfor their respective multilayered films with BGNPs are presented inFig. 5 and Fig. 6, respectively. These configurations were chosen tostudy the effect of the presence of BGNPs on the properties of the mul-tilayerfilmswith orwithout conjugates, namely on their final thickness,viscoelastic properties and on the stability of the LbL construction.

The normalized frequency, Δf/ν where ν is the overtone number,and the energy dissipation variation at the examined overtones, ΔD,are shown in Fig. 5 and Fig. 6 as a function of the deposited layers forthe different films. An Δf/ν decrease over time is observed after eachlayer deposition for each injection of CHT, HA, HA-C4h, HA-C36h, CHT-C18h and BGNPs, indicating a successful multilayered film construction.Also, a small increase of Δf/ν after each polymer adsorption can be ob-served, which is related to desorption of a small fraction of free PE dueto the rising step. On the other hand, each adsorbed layer is accompa-nied by a ΔD increase, due to the non-rigid layer structure of the filmthat is forming, typical of polymeric systems. Therefore, the multilay-ered films are expected to have higher damping properties for softerand more hydrated constructions [53,80].

As QCM-D results evidenced, the presence of the catechol groups ap-pears to decrease the viscous component of themultilayered films. ThisΔD decrease indicates that LbL films with catechol groups should prob-ably be thinner, denser, more rigid, less water-rich and, probably, ex-hibit a linear growth [53,80]. This effect was more pronounced insystems with CHT-C18h (CTR2 and MF2) than in systems containingone of the HA-C conjugates (CTR3, CTR4, MF3 and MF4). Also, the ΔDdecrease was noticeably more marked for the controls films that pres-ent both conjugates (CTR5 and CTR6). On the other hand, whenBGNPs were included in the multilayered films containing both conju-gates (MF5 and MF6), an opposite effect was observed (ΔD increase)when compared with the respective controls (CTR5 and CTR6). Thus,when both conjugates are combined, the addition of hydrophilic

Fig. 5. QCM-D results representing the mean and standard deviation of Δf/ν and ΔD variations from three different overtones as a function of the deposited layers in the build-up ofpolymeric films (CTR).


materials such as BGNPs seems to promote both the hydration and theviscoelastic properties of LbL films.

Based on the viscoelastic Kelvin-Voigt model [64,65], it was possibleto estimate the cumulative thickness evolution for all the LbL condi-tions, as a function of the number of deposited layers, which is repre-sented in Fig. 7.

Attending these modelled results, their trend line equations and re-spective square error calculation (R2) are shown in the Table 2. The re-sults evidence that the cumulative thickness for CTR1 and CTR2 have asimilar increase. As shown in the Table 2, both demonstrate an expo-nential growth, which could explain the higher Δf/ν values for thesetwo systems, with a final thickness for both around 500 nm. In

Fig. 6.QCM-D results representing themean and standard deviation ofΔf/ν andΔD variations from three different overtones as a function of the deposited layers in the build-up of theMFLbL coatings containing the BGNPs.


agreement with other authors [33,53,81], it was observed an exponen-tial growth of multilayered films containing CHT and HA. In fact, filmswith an exponential growth have been considered to occur by one oftwo general mechanisms: one attributed to the progressive increase inthe film surface roughness upon successive deposition of each newlayer, and another based on the ability of at least one PE component todiffuse “into” or “out” of a film during assembly [53,82]. In addition,

Sun et al. [82] reported that this type of film growth can be caused bythe molecular weight decrease of an anionic polymer, enhancing diffu-sion during assembly. On the other hand, as shown in Fig. 7 andTable 2, the presence of the catechol groups seemed to contribute to atendency of a linear growth, and the constructed films were thinnerand more rigid. This linear trend line has already been reported forCTR3 and CTR4 configurations [53,80], and in this work, it can be seen

Fig. 7. Cumulative thickness evolution of the multilayered films constructed as a functionof the number of deposited layers.


a linear growth followed by a plateau. In particular, the DS increase ofHA-C seems to further promote the multilayered film compaction(CTR4 N CTR3, MF4 N MF3, and MF6 N MF5). This effect was more pro-nounced in constructions with both conjugates, CHT-C18h and HA-C4hor HA-C36h, (CTR5 and CTR6), having a final thickness of 65 and66 nm, respectively.

Here, the higher DS of CTR6 compared to CTR5 did not show a con-siderable decrease in thickness, suggesting that multilayered filmshave probably reached their maximum compaction. In a previouswork [53], [CHT/HA]5 and [CHT/HA-C]5 films with a final thickness ofapproximately 130 nm and 75 nm, respectively, were obtained. Com-paring these results with those obtained in this work for 10 layers of[CHT/HA]11 (CTR1), [CHT/HA-C4h]11 (CTR3) and [CHT/HA-C36h]11(CTR4), it was found that these polymeric films had a higher thickness,207 nm, 206 nm and 182 nm, respectively, than those previously ob-tained [53]. This can be explained by the different polymer concentra-tions used, which in the case of Neto et al. [53] was 0.5 mg/mL for allPE solutions, and in the present work was 1 mg/mL for CHT and0.5 mg/mL for both HA and HA-C. As expected from previous works[54,55], the inclusion of BGNPs in the multilayered films MF1, MF2,MF3 and MF4 led to a final thickness lower than their respective con-trols (CTR1, CTR2, CTR3 and CTR4). In addition, it was previously re-ported the build-up of [CHT/HA/CHT/BGNPs]6 and [CHT/HA-C/CHT/BGNPs]6 films [55], where their thickness values were slightly higherthan the values obtained in this work. Thus, for 22 layers, a final thick-ness of 181 nm, 126 nm and 116 nm was obtained for MF1, MF3 andMF4, respectively. As expected from the ΔD values, the final thicknessdecreases with the combined presence of BGNPs and HA-C, as well aswith the DS increase (MF3 and MF4), when compared to their respec-tive controls, suggesting that these combinations result in a morecompacted films. These results could be also due to the polymeric

Table 2Trend line equations of each condition and their respective square error calculation (R2).

Trend line Equation Error

CTR1 Exponential y = 50.847e0.1193x R2 = 0.964CTR2 Exponential y = 48.217e0.1203x R2 = 0.9746CTR3 Linear y = 21.92x − 10.576 R2 = 0.9843CTR4 Linear y = 19.863x − 13.26 R2 = 0.9878CTR5 Linear y = 6.2818x + 18.367 R2 = 0.9515CTR6 Linear y = 5.6942x + 28.32 R2 = 0.9106MF1 Linear y = 7.492x + 27.988 R2 = 0.9701MF2 Linear y = 2.4424x + 32.291 R2 = 0.9657MF3 Linear y = 3.3686x + 71.812 R2 = 0.913MF4 Linear y = 8.1034x + 43.447 R2 = 0.9708MF5 Linear y = 10.854x − 31.912 R2 = 0.9536MF6 Linear y = 6.5431x + 41.503 R2 = 0.991

solutions with different concentration prepared by Rego et al. [55],which used 0.5 mg/mL for all PE, and also by their DS value (11%),which is lower than the values obtained in this work, 27% and 54%, forHA-C4h andHA-C36h, respectively. However, this behaviourwas not ver-ified for MF5 and MF6, which also had both conjugates and BGNPs. Inthese cases, the LbL assembly demonstrated a linear film growth andtheir final thicknesses were 220 nm for MF5 and 180 nm for MF6,beinghigher than the thickness values obtained for their respective con-trols, namely 67 nm for CTR5 and 71 nm for CTR6. These thicknesseswere higher than themultilayered films containing only one conjugate,CHT-C or HA-C conjugates (MF2, MF3, and MF4), or having the combi-nation of CHT, HA and BGNPs (MF1). Therefore, when films containingboth conjugates, it seems that the addition of the BGNPs contributesto a less compact, less rigid and more hydrated LbL film.

Based on the overall results, it was shown that the proposed config-urations could be successfully assembled by the versatile LbL technique.

3.5. Tensile shear strength tests

The adhesive strength obtained for the LbL films dip-coated on pairsof glass plates are shown in Fig. 8. MF films ending with an adhesivelayer were chosen, since it was found in a previous work [55] thatthese configurations exhibited higher adhesive strength than the MFfilms ending with BGNPs. These configurations were used to study theeffect of the presence of BGNPs on the adhesive properties of multilayerfilms, with or without conjugates (CHT-C, HA-C).

As Fig. 8 represents, polymeric films (controls) containing catecholgroups in their compositions (CTR2, CTR3, CTR4, CTR5 and CTR6) havehigher adhesive strength than the control film composedbyunmodifiedpolymers (CTR1). In fact, previous works [53,80] have already shownthat films containing HA-C showed an increase in the adhesive proper-ties and that multilayers films containing higher amounts of catecholgroups presented higher adhesive strength.

Adhesive strength results of [CHT/HA-C]11 coatings, represented bycontrols CTR3 and CTR4, were slightly higher than those obtained forsimilar system in a previous work [55]. These differences could be ex-plained by the different DS, which in the present work were higher,and also by the thickness of the polymeric coatings, since as it was al-readymentioned, these coatings were thicker than those previously ob-tained, resulting in LbL coatings with enhanced adhesion properties. Infact, there is no significant difference between CTR3 and CTR4 adhesionstrength, essentially due to their similar thicknesses. However, whenboth conjugates were combined (CTR5 and CTR6), a statistically signif-icant increase in the adhesion strength was observed in comparisonwith the formulations containing only one of the modified polymers(CTR2, CTR3 and CTR4). By comparing CTR4 and CTR6 with CTR3 andCTR5, it can be seen a slight increase in the adhesion properties of thecontrols with the DS increase of HA-C, which is also related with anabruptly decrease on their respective thickness (see Fig. S3).

Fig. 8 also shows that the inclusion of BGNPs in the LbL assembly de-creased the adhesive capability, except for MF1. As for the lamellarstructure of nacre, the nanosized inorganic phase, BGNPs, arranged ina biopolymer matrix should exhibit inelastic deformations, providingthe stress redistribution around strain concentration sites and also theelimination of stress concentration [55,83]. This is valid for MF1,where the rotations and deformations in the BGNPs were helped bythe biopolymer spacing composed by CHT and HA, resulting in an en-hanced adhesion [55,84]. The adhesion increase with the presence ofBGNPs was also observed in a previous work for CHT/HA based films[55]. However, for the remaining formulations containing BGNPs, theadhesive strength presented lower values than their respective poly-meric films (controls). Probably, when the inorganic phase of BGNPswas included in the LbL coatings containing catechol groups, the in-creased stiffness of the films together with their compaction led to alower interfacial adhesion between the multilayers.

Fig. 8. Adhesion strength values (MPa) measured for each LbL coating. Data are presented bymean± standard deviation (n=5; ****p b 0.0001; ***p b 0.001, **p b 0.01; *p b 0.05) [One-way ANOVA with Tukey's Multiple comparison test].


Despite the already demonstrated adhesive properties of CHT-C inprevious studies [56,57,68,85–88], thedecrease in the adhesive strengthwith the incorporation of BGNPs appears to be more pronounced whenCHT-C18h was present in the multilayer system, for example in MF2,MF5 and MF6. Additionally, all these formulations have an adhesivestrength lower than their respective controls. These effects could be re-lated with the lower content of adhesive polymers in the MF formula-tions, where some layers were replaced by BGNPs, which couldcontribute to a lower adhesive strength.

Thus, based on these adhesive properties, it was concluded thatCTR5 and CTR6were the configurations that demonstrated the best ad-hesive strength, and also the thinner films (see Fig. S3). The increasedstrength of CTR5 and CTR6 could be explained by the unique featureof the catechol groups being able to effectively reduce the mobility ofthe polymer chains during LbL assembly [89]. Such decrease in the in-terdiffusion of PE leads to a linear growth behaviour of the LbL filmsand their consequent compaction. Thus, in addition to the intrinsic ad-hesive nature of the two catechol-functionalized polymers (CHT-C andHA-C), this improved compaction probably confers increased stiffnessto LbL coatings, demonstrating a better response than the formulationscontaining BGNPs.

3.6. In-vitro bioactivity analysis

The main advantage of the BGs incorporation is their capability toprogressively dissolve onto the surrounding medium that reproducesthe human plasma (by having similar pH and ionic composition), re-leasing ions and form silanol groups on its surface that will act as a nu-cleating agent for a bone-like apatite deposition on the surface ofmaterials [48]. Hence, biomineralization studies in SBF were performed

with four LbL coatings configurations, namely CTR5, CTR6,MF5 andMF6(see Fig. 9) was assessed by analysing the apatite formation onto theirsurfaces under physiological-like condition. The selection of these con-ditionswas based on their adhesive strength,where CTR5 and CTR6 for-mulations exhibited the best values. MF5 and MF6 were also chosen toevaluate their bioactive properties conferred by BGNPs. Fig. 9 shows theSEM images and their respective EDS quantification, before and afterimmersion in SBF solution at 37 °C.

As expected from previousworks [48,55,90], after immersion in SBF,only the formulations containing BGNPs (MF5 and MF6) were able toinduce the formation of apatite-like structures on their surfaces. After7 days, nucleation and growth of apatite crystals were observed forboth formulations. Moreover, it was confirmed by the EDS analysis,which revealed changes in the elemental compositions related to silicon(Si), phosphorus (P) and calcium (Ca) on the film surface, as a result ofBGNPs dissolution [48]. Particularly, changes in the elemental composi-tion of MF6 remarkably greater than for MF5 were observed, demon-strating a decrease in Si concentration and an increase in Ca and Pconcentrations. After 14 days of immersion, both formulations evi-denced the formation of mineral agglomerates with cauliflower mor-phology and their respective EDS quantification showed Ca/P ratioswhich are close to the hydroxyapatite stoichiometry theoretical value(1.67) [91,92], similarly with those obtained by Luz et al. [93] On theother hand, the controls (CTR5 and CTR6) did not exhibit significantmorphological changes or presence of Si, P or Ca, as expected from pre-vious findings [48,55,90]. These results confirmed the bioactive behav-iour of the developed LbL configurations containing BGNPs, whichproved to be potentially able to promote the formation of a calciumphosphate (CaP) layer that would further crystallize in the form of car-bonated apatite when interacting with bone tissue [49,94].

Fig. 9. In-vitro bioactivity studies. Representative SEM images and respective quantitative EDS analysis of four LbL coatings configurations (see Fig. 1), before and after SBF immersion for 7and 14 days. The scale bar represents 5 μm.


Fig. 10 shows the XRD characterization of the LbL coatings, beforeand after immersion in SBF. Before the immersion in SBF, the XRD spec-trum showed a crystalline profile for all LbL coatings. According to liter-ature, the broad diffraction peak at around 2θ=20° can be attributed tothe presence of CHT, since its XRD pattern exhibited a characteristicpeak around this value [95–99].

Fig. 10. XRD diffractograms obtained for each LbL coating configuration before and afterimmersion in SBF for 14 days. *Peaks of hydroxyapatite.

This characteristic is generally attributed to the intermolecular hy-drogen bonding due to the presence of free NH2 groups within the mo-lecular structure, which results in the packing of the macromolecularpolymeric chains [100]. In addition, some works [96,101,102] have re-ported that the CHTmodification with catechol groups substantially re-duced the magnitude of the diffraction peak at 2θ = 20°, characteristicof CHT XRD pattern, indicating a noticeable decrease in the crystallinityof this compound, that could result from the partial breakage of the hy-drogen bonds in the original CHT, due to the grafting with catecholgroups [96,101,102].

According to Nath et al. [98], the sharp diffraction peak around 2θ=19° may confirm some complexation between CHT and HA. They alsoreported that the presence of some peaks at 14.17°, 17.31°, and 22.56°,may be attributed to a change in crystallinity, due to the interpen-etrating polymer network structure as a result of PE complexation[98]. On the other hand, other authors [103,104] reported that diffrac-tion peaks at 28° and 32°, can be attributed to the presence of NaCland, since all PE solutions were dissolved in NaCl, it could explainthese additional peaks.

After 14 days of immersion in SBF, both formulations containingBGNPs exhibited some typical crystalline peaks of hydroxyapatite,namely at 26° and 32° [48,49,55]. These characteristic peaks showedlower intensity in comparison with those reported in some previousworks [55], probably due to the lower thickness of the coatings con-structed in the presentwork. As expected, CTR filmswithout nanoparti-cles did not exhibit these characteristic peaks [55]. As expected, CTRfilms without nanoparticles did not exhibit these characteristic. There-fore, these results confirmed the bioactivity of the MF coatings,

Fig. 11. L929 viability results obtained through: a) MTS assay for 1, 4 and 7 days, where absorbance was read at 490 nm. Statistically significant differences between distinct samples byeach time point weremarkedwith ****, which represents p b 0.0001. Data are presented bymean± standard deviation (n=3) [One-way ANOVAwith Tukey'sMultiple comparison test];b) Representative images of L929 cells over differentmultilayer coatings (CTR5: [CHT-C18h/HA-C4h]11,MF5: [CHT-C18h/HA-C4h/CHT-C18h/BGNPs]5+ [CHT-C18h/HA-C4h], CTR6: [CHT-C18h/HA-C36h]11,MF6: [CHT-C18h/HA-C36h/CHT-C18h/BGNPs]5+ [CHT-C18h/HA-C36h]) represented by DAPI-phalloidin fluorescence assay at 1, 4 and 7 days. Cell nuclei are stained in bluewithDAPI and F-actin filaments are stained in red with phalloidin. Scale bar represents 50 μm.


supporting SEM/EDS results. Thus, such coatings could be potentiallyused for orthopedic applications.

3.7. In-vitro cellular tests

In order to evaluate the cellular behaviour when in direct contactwith the produced LbL coatings, some preliminary in vitro biological as-says were performed: MTS and fluorescence microscopy. By the chem-ical reduction of the MTS compound into formazan, the metabolicactivity of the cells could be determined. Fig. 11(a) shows the results ob-tained fromMTS assay for 1, 4 and 7 days, where the highest absorbancevalue corresponds to highermetabolic cellular activity indicating higher

number of cells. Four conditionswere evaluated: CTR5,MF5, CTR6,MF6,and TCPS, where cells were supposed to have a great proliferation.

TheMTS assay evaluated themetabolic activity of L929 cells and thetoxicity of themost promising CTR andMF constructions. After one day,the cells seemed to adhere to all constructionswith no significant differ-ences. After four and seven days, it was observed some significant differ-ences between MF conditions and their respective CTR ones, where MFcoatings exhibited better cellular viability. Similarly with other works[38,53–55,105] that have already reported better cellular responses onsubstrates modified with catechol groups, these results suggested thatthe combination of catechol-functionalized polymers with BGNPs hada positive influence on the cellular response contributing for bettercell viability and proliferation. Effectively, as it was mentioned


previously, MF coatings were slightly thicker than CTR, being less com-pact andmore hydrated and all these features could contribute for theirbetter biological results.

At day 7, comparing with the previous time points, it was clear animproved cell viability for all coatings, beingMF5 andMF6 the construc-tions that exhibited higher cell viability. Interestingly, coatings that ex-hibited better adhesive properties (CTR6 andCTR5), presented a cellularviability lower than their respective MF constructions. For one hand,these results evidenced that LbL coatings with higher catechol contentdid not have their cellular viability compromised. In fact, in a previouswork [53], the cellular behaviour of CHT/HA and CHT/HA-C multilayerswas compared and it was concluded that the catechol-functionalizedHA had a positive effect in terms of cell adhesion, proliferation and via-bility, comparingwith themultilayerfilmswithout such groups. Severalauthors [1,15,53] have attributed the enhancement of the cellular re-sponse to the fact that catechol groups can play an active anchor be-tween substrate surface and cells, allowing the formation of covalentand non-covalent bonds. On the other hand, the differences statisticallysignificant of both MF5 and MF6 in comparison with their respectivecontrols (CTR5 and CTR6) should be due to the BGNPs content. Thisfindingwas already observed in our previous work [55], being reportedthat their progressive dissolution could increase the pH of the culturemedium, interfering positively in cell proliferation and viability.

Furthermore, these MTS results were reinforced by the fluorescenceimages of L929 cells adhered on the surface of each LbL coating at 1, 4and 7 days, Fig. 11(b). Two fluorochromes, namely DAPI and phalloidin,were used to fix and stain the cells. Cell's nuclei were stained with blue(DAPI) and their cytoskeletons with red (phalloidin). The fluorescenceimages presented were consistent with the cell behaviour results previ-ously obtained. As it can be seen in Fig. 11(b), after 4 days, for all condi-tions, cells presented the typical morphology of the L929 cells and, after7 days, developed a kind of fibroblastic network, where the adheredcells occupied large space of the LbL coatings due to their intense prolif-erative activity. On the other hand, for both CTR5 and CTR6, the rate ofcell proliferation was lower. Generally, after 4 days, cells appeared tostretch and gain the L929 morphology, being more noticeable after7 days of culture. Moreover, the morphology of the adhered cells wassimilar in both MF5 and MF6, where the presence of BGNPs appearedto lead for a cell proliferation in cluster formations. This could be relatedto the typical heterogeneous distribution of BGNPs, already described byCouto et al. [49], and thatmight have been the reason for the presence ofa higher number of cells in a certain area.

4. Conclusion

Nanostructured coatings inspired by MAPs and by the inorganic–organic nacre-like structure were successfully developed through theLbL assembly technique. Multifunctional films were constructedthrough the combination of inorganic BGNPs with biopolymeric layersof CHT and HA functionalized with catechol groups, and as their con-trols, the polymeric films only composed by catechol-functionalizedbiopolymers.

QCM-D results showed that the proposed multilayered films weresuccessfully assembled, resulting in stable LbL films. Polymeric filmscontaining catechol groups showed a decrease in ΔD values, indicatingthat these films should be denser, more rigid and less water-rich. In ad-dition, it was observed that the presence of the catechol groups typicallycontributed to a linear film growth, resulting in thinner films. Except forMF5 andMF6, all othermultifunctionalfilms containing catechol groupsand BGNPs presented lower ΔD and thickness than their respectivecontrols.

Adhesive strength results confirmed that the presence of both CHT-Cand HA-C conjugates in CTR5 and CTR6 significantly improved their ad-hesion properties, when compared to the other CTR coatingswith lowercontent of catechol groups. On theother hand, besides the incorporation

of BGNPs in these LbL constructions slightly decreased their adhesionstrength, they provided bioactive potential after immersion in SBF.

Concerning the evaluation of cell behaviour, CTR and MF films withhigher content of catechol groups demonstrated to be nontoxic, butglobally, MF constructions exhibited better cell adhesion, proliferationand viability.

Finally, the produced LbL coatings could be used in a variety of appli-cations as biocompatible interface between the implant and host tis-sues. So, MF films could be potentially used as adhesive coatings oforthopedic implants, in order to promote osteogenesis and hydroxyap-atite deposition around the implant. On the other hand, thehighly adhe-sive polymeric coatings (CTR) could be used to improve the junctionbetween distinct orthopedic implants and a variety of hard tissues, ina simple and versatile way.

CRediT authorship contribution statement

Ana Catarina Almeida: Investigation, Data curation, Software, Writ-ing - original draft.Ana CatarinaVale: Conceptualization,Methodology,Investigation, Data curation,Writing - review& editing. Rui L. Reis:Val-idation, Resources. Natália M. Alves: Supervision, Validation, Writing -review & editing.

Declaration of competing interest

The authors declare no competing financial interest.

Acknowledgments

The authors acknowledge the Portuguese Foundation for Scienceand Technology (FCT) and the European program FEDER/FEEI for the fi-nancial support through projects PTDC/BTM-MAT/28123/2017 andPTDC/NAN-MAT/31036/2017. The authors also acknowledge FCT forthe financial support through the exploratory project MIT-EXPL/BIO/0089/2017. This paper has also been prepared with the support ofREMIX Project, funded by the European Union's Horizon 2020 Researchand Innovation programme under the Maria Sklodowska-Curie grantagreement n. 778078.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijbiomac.2020.04.095.

References

[1] H. Lee, S.M. Dellatore, W.M. Miller, P.B. Messersmith, Mussel-inspired surfacechemistry for multifunctional coatings, Science 318 (2007) 426–430.

[2] H. Lee, Y. Lee, A.R. Statz, J. Rho, T.G. Park, P.B. Messersmith, Substrate-independentlayer-by-layer assembly by using mussel-adhesive-inspired polymers, Adv. Mater.20 (2008) 1619–1623.

[3] Z. Tang, Y.Wang, P. Podsiadlo, N.A. Kotov, Biomedical applications of layer-by-layerassembly: from biomimetics to tissue engineering, Adv. Mater. 18 (2006)3203–3224.

[4] E. Faure, C. Falentin-Daudré, C. Jérôme, J. Lyskawa, D. Fournier, P. Woisel, C.Detrembleur, Catechols as versatile platforms in polymer chemistry, Prog. Polym.Sci. 38 (2013) 236–270.

[5] J. Longo, T. Garnier, M. Mateescu, F. Ponzio, P. Schaaf, L. Jierry, V. Ball, Stable bioac-tive enzyme-containing multilayer films based on covalent cross-linking frommussel-inspired adhesives, Langmuir 31 (2015) 12447–12454.

[6] P. Podsiadlo, Z. Liu, D. Paterson, P.B. Messersmith, N.A. Kotov, Fusion of seashellnacre and marine bioadhesive analogs: high-strength nanocomposite by layer-by-layer assembly of clay and L-3,4-dihydroxyphenylalanine polymer, Adv.Mater. 19 (2007) 949–955.

[7] G. Luz, J.F. Mano, Biomimetic design of materials and biomaterials inspired by thestructure of nacre, Philos. Trans Royal Soc. A Math. Phys. Eng. Sci. 367 (2009)1587–1605.

[8] J. Sun, B. Bhushan, Hierarchical structure and mechanical properties of nacre: a re-view, RSC Adv. 2 (2012) 7617–7632.

[9] J. Wu, L. Zhang, Y. Wang, Y. Long, H. Gao, X. Zhang, N. Zhao, Y. Cai, J. Xu, Mussel-inspired chemistry for robust and surface-modifiable multilayer films, Langmuir27 (22) (2011) 13684–13691.

https://doi.org/10.1016/j.ijbiomac.2020.04.095https://doi.org/10.1016/j.ijbiomac.2020.04.095http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0005http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0005http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0010http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0010http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0010http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0015http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0015http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0015http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0020http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0020http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0020http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0025http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0025http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0025http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0030http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0030http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0030http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0030http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0030http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0030http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0035http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0035http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0035http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0040http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0040http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0045http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0045http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0045


[10] X. Zhang, H. Chen, H. Zhang, Layer-by-layer assembly: from conventional to uncon-ventional methods, Chem. Commun. (14) (2007) 1395–1405.

[11] G. Decher, Fuzzy nanoassemblies: toward layered polymeric multicomposites, Sci-ence 277 (1997) 1232–1237.

[12] S.K. Madhurakkat Perikamana, J. Lee, Y.B. Lee, Y.M. Shin, E.J. Lee, A.G. Mikos, H.Shin, Materials from mussel-inspired chemistry for cell and tissue engineering ap-plications, Biomacromolecules 16 (2015) 2541–2555.

[13] Y. Lee, H.J. Chung, S. Yeo, C.-H. Ahn, H. Lee, P.B. Messersmith, T.G. Park, Thermo-sensitive, injectable, and tissue adhesive sol-gel transition hyaluronic acid/pluroniccomposite hydrogels prepared from bio-inspired catechol-thiol reaction, Soft Mat-ter 6 (2010) 977–983.

[14] Q. Lin, D. Gourdon, C. Sun, N. Holten-Andersen, T.H. Anderson, J.H. Waite, J.N.Israelachvili, Adhesion mechanisms of the mussel foot proteins mfp-1 and mfp-3, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 3782–3786.

[15] H. Lee, N.F. Scherer, P.B. Messersmith, Single-molecule mechanics of mussel adhe-sion, Proc. Natl. Acad. Sci. 103 (2006) 12999–13003.

[16] J.H. Waite, M.L. Tanzer, Polyphenolic substance of Mytilus edulis: novel adhesivecontaining L-Dopa and hydroxyproline, Science 212 (1981) 1038–1040.

[17] J.H. An, N.T. Huynh, Y. Sil Jeon, J.-H. Kim, Surface modification using bio-inspiredadhesive polymers based on polyaspartamide derivatives, Polym. Int. 60 (2011)1581–1586.

[18] C.J. Ochs, T. Hong, G.K. Such, J. Cui, A. Postma, F. Caruso, Dopamine-mediated con-tinuous assembly of biodegradable capsules, Chem. Mater. 23 (2011) 3141–3143.

[19] E. Amstad, T. Gillich, I. Bilecka, M. Textor, E. Reimhult, Ultrastable iron oxide nano-particle colloidal suspensions using dispersants with catechol-derived anchorgroups, Nano Lett. 9 (2009) 4042–4048.

[20] A. Valverde, L. Pérez-Álvarez, L. Ruiz-Rubio, M.A. Pacha Olivenza, M.B. GarcíaBlanco, M. Díaz-Fuentes, J.L. Vilas-Vilela, Antibacterial hyaluronic acid/chitosanmultilayers onto smooth and micropatterned titanium surfaces, Carbohydr.Polym. 207 (2019) 824–833.

[21] C. Pan, Z. Zhou, X. Yu, Coatings as the useful drug delivery system for the preven-tion of implant-related infections, J. Orthop. Surg. Res. 13 (2018) (N.PAG).

[22] J.P. Govindharajulu, C. Xi, L. Yuping, J.C. Rodriguez-Cabello, M. Battacharya, C.Aparicio, Chitosan-recombinamer layer-by-layer coatings for multifunctional im-plants, Int. J. Mol. Sci. 18 (2017) 369–384.

[23] M. Jäger, H.P. Jennissen, F. Dittrich, A. Fischer, H.L. Köhling, Antimicrobial andosseointegration properties of nanostructured titanium orthopaedic implants, Ma-terials 10 (2017) 1302.

[24] J.A. Jennings, J.D. Bumgardner, Chitosan based biomaterials, Fundamentals, volume1, Elsevier Science, 2016.

[25] Z. Zhong, J. Qin, J. Ma, Cellulose acetate/hydroxyapatite/chitosan coatings for im-proved corrosion resistance and bioactivity, Mater. Sci. Eng. C Mater. Biol. Appl.49 (2015) 251–255.

[26] K. Bai, Y. Zhang, Z. Fu, C. Zhang, X. Cui, E. Meng, S. Guan, J. Hu, Fabrication of chito-san/magnesium phosphate composite coating and the in vitro degradation proper-ties of coated magnesium alloy, Mater. Lett. 73 (2012) 59–61.

[27] J.D. Bumgardner, R. Wiser, P.D. Gerard, P. Bergin, B. Chestnutt, M. Marini, V.Ramsey, S.H. Elder, J.A. Gilbert, Chitosan: potential use as a bioactive coating for or-thopaedic and craniofacial/dental implants, Aust. J. Biol. Sci. 14 (2003)423–423–438.

[28] Y. Yang, Q. He, L. Duan, Y. Cui, J. Li, Assembled alginate/chitosan nanotubes for bi-ological application, Biomaterials 28 (2007) 3083–3090.

[29] H. Lv, Z. Chen, X. Yang, L. Cen, X. Zhang, P. Gao, Layer-by-layer self-assembly ofminocycline-loaded chitosan/alginate multilayer on titanium substrates to inhibitbiofilm formation, J. Dent. 42 (2014) 1464–1472.

[30] H. Ao, S. Yang, B.E. Nie, Q. Fan, Q. Zhang, J. Zong, S. Guo, X. Zheng, T. Tang, Improvedantibacterial properties of collagen I/hyaluronic acid/quaternized chitosan multi-layer modified titanium coatingswith both contact-killing and release-killing func-tions, J. Mater. Chem. B 7 (2019) 1951–1961.

[31] J. Wei, J. Shi, Q. Wu, L. Yang, S. Cao, Hollow hydroxyapatite/polyelectrolyte hybridmicroparticles with controllable size, wall thickness and drug delivery properties, J.Mater. Chem. B 3 (2015) 8162–8169.

[32] X. Zhu, J. Shi, H. Ma, R. Chen, J. Li, S. Cao, Hierarchical hydroxyapatite/polyelectro-lyte microcapsules capped with AuNRs for remotely triggered drug delivery,Mater. Sci. Eng. C 99 (2019) 1236–1245.

[33] L. Richert, P. Lavalle, E. Payan, X.Z. Shu, G.D. Prestwich, J.F. Stoltz, P. Schaaf, J.C.Voegel, C. Picart, Layer by layer buildup of polysaccharide films: physical chemistryand cellular adhesion aspects, Langmuir 20 (2004) 448–458.

[34] J. Fu, J. Ji, W. Yuan, J. Shen, Construction of anti-adhesive and antibacterial multi-layer films via layer-by-layer assembly of heparin and chitosan, Biomaterials 26(2005) 6684–6692.

[35] C.-H. Jou, L. Yuan, S.-M. Lin, M.-C. Hwang,W.-L. Chou, D.-G. Yu, M.-C. Yang, Biocom-patibility and antibacterial activity of chitosan and hyaluronic acid immobilizedpolyester fibers, J. Appl. Polym. Sci. 104 (2007) 220–225.

[36] Z. Shi, K.G. Neoh, E.T. Kang, C. Poh, W. Wang, Bacterial adhesion and osteoblastfunction on titaniumwith surface-grafted chitosan and immobilized RGD peptide,J. Biomed. Mater. Res. Part A 86 (2008) 865–872.

[37] J.H.H. Bongaerts, J.J. Cooper-White, J.R. Stokes, Low biofouling chitosan-hyaluronicacid multilayers with ultra-low friction coefficients, Biomacromolecules 10 (2009)1287–1294.

[38] X. Zhang, Z. Li, X. Yuan, Z. Cui, X. Yang, Fabrication of dopamine-modifiedhyaluronic acid/chitosan multilayers on titanium alloy by layer-by-layerself-assembly for promoting osteoblast growth, Appl. Surf. Sci. 284 (2013)732–737.

[39] Y. Li, X. Wang, J. Sun, Layer-by-layer assembly for rapid fabrication of thick poly-meric films, Chem. Soc. Rev. 41 (2012) 5998–6009.

[40] J.J. Richardson, J. Cui, M. Björnmalm, J.A. Braunger, H. Ejima, F. Caruso, Innovation inlayer-by-layer assembly, Chem. Rev. 116 (2016) 14828–14867.

[41] D. Wang, J. Zhang, Y. Zhong, M. Chu, W. Chang, Z. Yao, Mussel-inspired bio-compatible free-standing adhesive films assembled layer-by-layer with water-resistance, RSC Adv. 8 (2018) 18904–18912.

[42] J. Borges, J.F. Mano, Molecular interactions driving the layer-by-layer assembly ofmultilayers, Chem. Rev. 114 (2014) 8883–8942.

[43] K. Halasz, G. Grozdits, L. Csóka, 10 - Functional nanostructured coatings via layer-by-layer self-assembly, in: M. Aliofkhazraei (Ed.), Anti-abrasive Nanocoatings,Woodhead Publishing 2015, pp. 249–281.

[44] N.M. Salwowska, K.A. Bebenek, D.A. Zadlo, D.L. Wcislo-Dziadecka, Physiochemicalproperties and application of hyaluronic acid: a systematic review, J. CosmeticDermatol. 15 (2016) 520–526.

[45] N.M. Alves, I.B. Leonor, H.S. Azevedo, R.L. Reis, J.F. Mano, Designing biomaterialsbased on biomineralization of bone, J. Mater. Chem. 20 (2010) 2911–2921.

[46] N.M. Alves, I. Pashkuleva, R.L. Reis, J.F. Mano, Controlling cell behavior through thedesign of polymer surfaces, Small 6 (2010) 2208–2220.

[47] B. Lu, D. Luo, A. Zhao, H. Wang, Y. Zhao, M.F. Maitz, P. Yang, N. Huang, pH respon-sive chitosan and hyaluronic acid layer by layer film for drug delivery applications,Prog. Org. Coat. 135 (2019) 240–247.

[48] G.M. Luz, J.F. Mano, Preparation and characterization of bioactive glass nanoparti-cles prepared by sol-gel for biomedical applications, Nanotechnology 22 (2011),494014. .

[49] D.S. Couto, N.M. Alves, J.F. Mano, Nanostructured multilayer coatings combiningchitosan with bioactive glass nanoparticles, J. Nanosci. Nanotechnol. 9 (2009)1741–1748.

[50] Z. Hong, R.L. Reis, J.F. Mano, Preparation and in vitro characterization of scaffolds ofpoly(L-lactic acid) containing bioactive glass ceramic nanoparticles, Acta Biomater.4 (2008) 1297–1306.

[51] Y. Abe, T. Kokubo, T. Yamamuro, Apatite coating on ceramics, metals and polymersutilizing a biological process, J. Mater. Sci. Mater. Med. 1 (1990) 233–238.

[52] M. Tanahashi, T. Yao, T. Kokubo, M. Minoda, T. Miyamoto, T. Nakamura, T.Yamamuro, Apatite coating on organic polymers by a biomimetic process, J. Am.Ceram. Soc. 77 (1994) 2805–2808.

[53] A.I. Neto, A.C. Cibrao, C.R. Correia, R.R. Carvalho, G.M. Luz, G.G. Ferrer, G. Botelho, C.Picart, N.M. Alves, J.F. Mano, Nanostructured polymeric coatings based on chitosanand dopamine-modified hyaluronic acid for biomedical applications, Small (12)(2014) 2459–2469.

[54] A.L. Carvalho, A.C. Vale, M.P. Sousa, A.M. Barbosa, E. Torrado, J.F. Mano, N.M. Alves,Antibacterial bioadhesive layer-by-layer coatings for orthopedic applications, J.Mater. Chem. B 4 (2016) 5385–5393.

[55] S.J. Rego, A.C. Vale, G.M. Luz, J.F. Mano, N.M. Alves, Adhesive bioactive coatings in-spired by sea life, Langmuir 32 (2016) 560–568.

[56] K. Kim, K. Kim, J.H. Ryu, H. Lee, Chitosan-catechol: a polymer with long-lastingmucoadhesive properties, Biomaterials 52 (2015) 161–170.

[57] J. Xu, S. Strandman, J.X.X. Zhu, J. Barralet, M. Cerruti, Genipin-crosslinked catechol-chitosan mucoadhesive hydrogels for buccal drug delivery, Biomaterials 37 (2015)395–404.

[58] A. Ghadban, A.S. Ahmed, Y. Ping, R. Ramos, N. Arfin, B. Cantaert, R.V. Ramanujan, A.Miserez, Bioinspired pH andmagnetic responsive catechol-functionalized chitosanhydrogels with tunable elastic properties, Chem. Commun. 52 (2016) 697–700.

[59] A.C. Almeida, A.C. Vale, R.A. Pires, R.L. Reis, N.M. Alves, Layer-by-layer films basedon catechol-modified polysaccharides produced by dip- and spin-coating onto dif-ferent substrates, J. Biomed. Mater. Res. B Appl. Biomater. 108 (2020) 1412–1427In this issue.

[60] S. Burks, W. Spotnitz, Safety and usability of hemostats, sealants, and adhesives,AORN J. 100 (2014) 160–176.

[61] Z. Hong, R.L. Reis, J.F. Mano, Preparation and in vitro characterization of novelbioactive glass ceramic nanoparticles, J. Biomed. Mater. Res. A 88 (2009)304–313.

[62] M.C. Dixon, Quartz crystal microbalance with dissipation monitoring: enablingreal-time characterization of biological materials and their interactions, J. Biomol.Technol. 19 (2008) 151–158.

[63] N.M. Alves, C. Picart, J.F. Mano, Self assembling and crosslinking of polyelectrolytemultilayer films of chitosan and alginate studied by QCM and IR spectroscopy,Macromol. Biosci. 9 (2009) 776–785.

[64] M.V. Voinova, M. Rodahl, M. Jonson, B. Kasemo, Viscoelastic acoustic response oflayered polymer films at fluid-solid interfaces: continuum mechanics approach,Phys. Script 59 (1999) 391–396.

[65] S.X. Liu, J.-T. Kim, Application of Kevin–Voigt model in quantifying whey proteinadsorption on polyethersulfone using QCM-D, JALA 14 (2009) 213–220.

[66] T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity?Biomaterials 27 (2006) 2907–2915.

[67] S. Hong, K. Yang, B. Kang, C. Lee, I.T. Song, E. Byun, K.I. Park, S.-W. Cho, H. Lee,Hyaluronic acid catechol: a biopolymer exhibiting a pH-dependent adhesive or co-hesive property for human neural stem cell engineering, Adv. Funct. Mater. 23(2013) 1774–1780.

[68] K. Kim, J.H. Ryu, D.Y. Lee, H. Lee, Bio-inspired catechol conjugation converts water-insoluble chitosan into a highly water-soluble, adhesive chitosan derivative forhydrogels and LbL assembly, Biomater. Sci. 1 (2013) 783–790.

[69] T.C. Laurent, U.B. Laurent, J.R. Fraser, The structure and function of hyaluronan: anoverview, Immun. Cell. Biol. 74 (1996) A1–A7.

[70] M.K. Cowman, T.A. Schmidt, P. Raghavan, A. Stecco, Viscoelastic properties ofhyaluronan in physiological conditions, F1000Research 4 (2015) 622.

[71] H.Y. Zhou, X.G. Chen, M. Kong, C.S. Liu, D.S. Cha, J.F. Kennedy, Effect of molecularweight and degree of chitosan deacetylation on the preparation and characteristics

http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0050http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0050http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0055http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0055http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0060http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0060http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0060http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0065http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0065http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0065http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0065http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0070http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0070http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0070http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0075http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0075http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0080http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0080http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0080http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0080http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0085http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0085http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0085http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0090http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0090http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0095http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0095http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0095http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0100http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0100http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0100http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0100http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0105http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0105http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0110http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0110http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0110http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0115http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0115http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0115http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0120http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0120http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0125http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0125http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0125http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0130http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0130http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0130http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0135http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0135http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0135http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0135http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0140http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0140http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0145http://refhub.elsevier.com/S0141-8130(20)32953-6/rf0145http://refhu

Documents

Bioactive and adhesive properties of multilayered coatings ... · Bioactive and adhesive properties of multilayered coatings based on catechol-functionalized chitosan/hyaluronic acid