Thermodynamics of multi-walled carbon nanotube biofunctionalization using nisin: The effect of peptide structure

Thermodynamics of multi-walled carbon nanotube biofunctionalization using nisin: The effect of peptide structure

Colloids and Surfaces A 578 (2019) 123611 Contents lists available at ScienceDirect Colloids and Surfaces A journal homepage: www.elsevier.com/locat...

3MB Sizes 0 Downloads 2 Views

Colloids and Surfaces A 578 (2019) 123611

Contents lists available at ScienceDirect

Colloids and Surfaces A journal homepage: www.elsevier.com/locate/colsurfa

Thermodynamics of multi-walled carbon nanotube biofunctionalization using nisin: The effect of peptide structure

T

Guilherme M.D. Ferreiraa,b, Gabriel M.D. Ferreiraa,c, Germanna W.R. Almeidad, ⁎ Nilda F.F. Soarese, Ana Clarissa S. Pirese, Luis H.M. Silvaa, a

Grupo de Química Verde Coloidal e Macromolecular, Department of Chemistry, Federal University of Viçosa, Av. P. H. Rolfs s/n, CEP 36570-900, Viçosa, MG, Brazil Departament of Chemistry, Federal University of Lavras, Campus Universitário, 3037, Lavras, MG, Brazil c Departament of Chemistry, Institute of Exact and Biological Sciences, Federal University of Ouro Preto (UFOP), Campus Morro do Cruzeiro, CEP 35400-000, Ouro Preto, MG, Brazil d Federal Institute of Rondônia, BR 435, Km 63, Zona Rural 76993000, Colorado do Oeste, RO, Brazil e Department of Food Technology, Federal University of Viçosa, Av. P. H. Rolfs s/n, 36570900, Viçosa, MG, Brazil b

GRAPHICAL ABSTRACT

ARTICLE INFO

ABSTRACT

Keywords: Nisin Adsorption Carbon nanotube Thermodynamics Biocomplex

A detailed adsorption thermodynamic study of two nisin variants (nisA and nisZ) on multi-walled carbon nanotubes (MWCNT) was conducted, and the antimicrobial properties of biofunctionalized MWCNT were evaluated. The MWCNT adsorption capacity for peptides (qp ) increased with pH increase, with greater values for nisZ (qp up to 250 mg g−1) than for nisA (qp up to 180 mg g−1) for all studied conditions. Zeta potential measurements (ζ) showed that the electrostatic repulsion between the MWCNT-peptide complexes determined the dispersion features that were stable at pH 2.0 and 3.0, with ζ reaching 45.0 mV at the lowest pH. Despite the similar ζ values for both peptides, slightly greater stabilization of MWCNT dispersions was exhibited in presence of nisZ at pH 3. At pH 4 and 5, peptide adsorption was not able to promote MWCNT dispersibility. Isothermal titration calorimetry revealed that the adsorption process was driven by enthalpy for both peptides, as the adsorption enthalpy changes ( Hads ) were less negative than −99.7 kJ mol−1. Despite the large dependence of Hads on the pH andqp values, indicating the important role of electrostatic interactions on the adsorption process, a change in only one amino acid residue in the nisin structure promoted intense changes in the adsorption thermodynamic parameters. We have suggested that the more hydrophobic character of nisZ at the lower pH values caused this peptide to interact with the MWCNT surface through its two domains. Interesting, the antimicrobial properties of



Corresponding author. E-mail address: [email protected] (L.H.M. Silva).

https://doi.org/10.1016/j.colsurfa.2019.123611 Received 22 April 2019; Received in revised form 11 June 2019; Accepted 24 June 2019 Available online 25 June 2019 0927-7757/ © 2019 Elsevier B.V. All rights reserved.

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

both peptides were not damaged, and the functionalized MWCNT showed antibacterial activity against the indicator micro-organism Lactococcus lactis in the agar diffusion and solution assays.

1. Introduction

to prevent biofilm formation. Dong and Yang [22] developed filters based on MWCNT recovery with nisin. The filters presented dual functions in capturing and inactivating bacteria, and the filter potentiality was attributed to the high ability of nisin bound to the MWCNT. Despite these studies, to the best of our knowledge, no adsorption study of nisin on MWCNT has been carefully conducted. In this work, multi-walled carbon nanotubes were non-covalently functionalized with two variants of nisin peptide (nisin A and nisin Z) at different pH values, and the biofunctionalized MWCNT were characterized using transmission electron microscopy (TEM) and zeta potential. The effect of the change in the peptide structure on the adsorption process was evaluated by combining adsorption isotherms with isothermal titration calorimetry (ITC). Finally, antimicrobial activity of the obtained nanocomposites was evaluated.

Since the discovery of carbon nanotubes (CNT) by Sumio Iijima [1], obtaining stable dispersions of these structures to promote their potential applications, for instance, in electronics, packaging, and biomedicine has been a challenge [2–4]. In this context, several strategies, including the use of surfactants, [5] mineral oils [6], clay platelets [7], polymers [8], and polar aromatic organic molecules [9], have been evaluated to change the surface of CNT, favoring the dispersion of these nanomaterials. Among all strategies to disperse CNT, the biofunctionalization of CNT, i.e., the introduction of biomolecules on the surface, has emerged as a methodology to enhance the dispersion stability of CNT [10], providing a strategy for the development of new bioactive nanomaterials with amazing properties. Carbon nanotubes functionalized with proteins, carbohydrates, and nucleic acids, for example, are highly biocompatible materials that have been applied for the design of bionanosensors, bionanoreactors, and drug carriers for cancer treatment [11–13]. However, studies in this area are still scarce concerning the understanding of the energetics associated with functionalization. Among the biomolecules used to modify CNT, peptides have attracted attention because of their variety of properties [14,15] that can be potentialized when these molecules are attached to the CNT surface [16]. Furthermore, peptides contain different functional groups because of the presence of many types of amino acid residues that can interact with the CNT surface through Van der Waals, electrostatic, hydrophobic, and π-π stacking interactions, favoring non-covalent CNT functionalization [17]. Despite this, the complex balance of intermolecular interactions that determines the success of CNT functionalization by peptides has not been adequately evaluated through thermodynamic studies, which would be invaluable to investigating the relationship among peptide structures, the driven forces that determine peptide adsorption on the CNT surface, and the effect of peptide adsorption on the obtained biocomplex properties. An important class of peptides that can attribute interesting properties to the CNT are antimicrobial peptides, which have been extensively used in biomedical devices, equipment for food processing, and food preservation [18]. In this class, nisin, a natural cationic antimicrobial peptide commonly used in the world food industry, has the ability to inhibit the growth of several gram-positive bacteria. Two natural variants of this peptide are known: nisin A (nisA) and nisin Z (nisZ). These molecules differ from each other by the substitution of only one amino acid residue, in which the histidine in nisA is replaced with asparagine, an uncharged and more polar amino acid residue, in nisZ in the position 27 [19]. Despite this singular difference, nisA and nisZ has some distinct properties, such as solubility and stability [20], which are caused by the structural difference between those two amino acids residues as well as the conformational changes that they can cause in the structure of the peptide. Thus, we believe that calorimetric studies combined with adsorption isotherms of both nisA and nisZ on CNT could supply thermodynamic data that would strategically reveal important details about how nisin variants structure determines the driven forces that govern the adsorption of these biomolecules on those nanostructures. These studies will provide insights to understand the adsorption of other peptides (or proteins) based on their hydrophilic/ hydrophobic balance. Qi et al. [21] have successfully immobilized nisin on multi-walled carbon nanotubes (MWCNT) using a covalent approach in which poly (ethylene glycol) was used as a linker. The MWCNT-nisin composite retained the antimicrobial activities of nisin, improving the capability

2. Materials and methods 2.1. Chemical The antimicrobial peptides, ultrapure nisin A (≥ 95.0%) and ultrapure nisin Z (≥ 95.0%), were purchased from Handary S.A. (Bruxelles, Belgium). The multi-walled carbon nanotubes (CTUBE 100 MWCNT) were obtained from CNT Co. Ltd. (Korea), and their technical specifications were supplied by the manufacturer: average diameter, 10–40 nm; size, 1–25 μm; purity, ≥ 93% w/w; density, 0.03–0.06 g cm−3; specific surface area, 188 m2 g−1. Citric acid and sodium hydrogen phosphate dihydrate were supplied by Vetec/Sigma Aldrich (Brazil). All chemicals were used as received, without further purification. Deionized water was used to prepare all solutions. 2.2. Adsorption isotherms To obtain the adsorption isotherms of nisin on MWCNT, 10.0 mg of adsorbent was dispersed in 10.0 mL of the peptide solution (nisA or nisZ) into 15 mL glass flasks. The peptide concentration in the solutions ranged from 0 to 0.5 mg mL−1. The systems were manually stirred for 10 min and then placed in a temperature-controlled bath for 24 h, where they remained at rest. Afterwards, as the systems had enough time to reach thermodynamic equilibrium, the supernatants were collected for analysis of the peptide concentration, which was determined by high performance liquid chromatography (HPLC) (section 3.3). The amount of peptide adsorbed on the MWCNT (qP ), in mg g−1, was calculated according to Eq. (1):

qP =

( Co

Ce )·V m

(1) −1

where Co and Ce (mg mL ) are the initial and equilibrium concentrations, respectively, of the peptide in the supernatant, V (L) is the volume of the supernatant, and m (g) is the mass of MWCNT in the system. Adsorption studies were performed in buffer solutions of pH 2, 3, 4, and 5. Buffer solutions were prepared by mixing amounts of 0.1 mol L−1 citric acid and 0.2 mol L−1 sodium hydrogen phosphate in adequate proportions. 2.3. HPLC analysis HPLC analysis for nisA and nisZ quantification was performed in a Shimadzu liquid chromatograph (10 AVP), equipped with a UV–vis detector (Shimadzu, SPD-M10Avp). A C18 VP-ODS reversed-phase column (15 cm × 4.6 mm) was used throughout the analysis. The method for peptide quantification used a binary gradient with mobile 2

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

phases containing acetonitrile (20–70%) and 0.1% trifluoroacetic acid (30–80%) at a flow rate of 1.0 mL min−1. The injection volume of sample was 20 μL, and the column effluents were monitored at 233 nm. The retention times of nisA and nisZ were 8.5 and 8.8 min, respectively. Quantification was made according to the analytical calibration curves from the standard solutions of the peptides, for the different conditions evaluated.

where qi, int and qi, dil are the energies absorbed or released in the reaction cell in the presence and absence of MWCNT, respectively, at the ith injection, and ni is the amount of peptide, in mol, that adsorbs onto the adsorbent for the same injection. The ni values were obtained from the adsorption isotherms. All experiments were performed at 25.0000 ± 0.0001 °C, and assays were conducted in duplicate. 2.5. Characterization of MWCNT-peptide complexes

2.4. Isothermal titration calorimetry (ITC)

2.5.1. Retrieval of MWCNT-peptide complexes Dispersions of MWCNT in peptide solution (1 mg mL−1) containing different concentrations of nisA or nisZ (0.050–1.0 mg mL−1) were prepared at pH 2, 3, 4, and 5, as described in section 3.2. These dispersions were manually shaken for 10 min and remained at rest in a temperature-controlled bath for 24 h, at 25 °C. Afterwards, the samples were sonicated in a Desruptor Eco-Sonics (Ultronique), using a micropoint of 4 mm in diameter during 5 min at 350 W. The temperature of the dispersion was monitored with a thermometer and maintained at 25 °C using an ice bath. Then, the dispersions were centrifuged in a Sigma® 4K15 centrifuge at 5100 rpm for 1 h at 25 °C to separate the aggregates with the higher mass from the isolated MWCNT tubes covered with the peptides that remained at the supernatant. Samples were characterized using zeta potential and transmission electron microscopy.

Calorimetry experiments were performed in a TAM III Isothermal Titration Nanocalorimeter (TA instruments, EUA) equipped with two 4.00 mL reactions cells (sample and reference) and controlled by dedicated TAM Assistant™ software. To obtain the enthalpy changes associated with the adsorption process, a concentrated solution of peptide (nisA or nisZ), prepared in buffer, titrated a dispersion of 4.00 mg of MWCNT in 2.70 mL of buffer into the reaction cell, while the reference cell was filled with 2.70 mL of buffer. Titrations were performed stepwise by injections of a known quantity of peptide solution into the cell, using a 500 μL Hamilton syringe controlled by a piston pump. The time intervals between two consecutive injections were at least 45 min. Table S1 presents the peptide concentration in the titration solution (Cp ) and the volume of injection (Vinj ) for each pH evaluated. Solutions were degassed for 10 min prior to titrations to avoid bubble formation inside the system, and a helix stirrer at 150 rpm stirred the dispersion inside the reaction cell during all experiments. Blank experiments, with the addition of a peptide solution in the buffer in the absence of MWCNT, were also performed to discount the effect of the peptide dilution during the adsorption process. The heat, in kJ, resulting from each injection i in the experiment was obtained by the integration of the thermograms provided by the equipment, and the adsorption enthalpy changes ( Hads ), in kJ mol–1, associated with the process were obtained using Eq. (2):

Hads =

i

2.5.2. Zeta potential (ζ) The electrokinetic potential (zeta potential, ζ) of MWCNT and the MWCNT-peptide complex was measured on a Zetasizer NanoZS system (Malvern Instruments Ltd., UK), using disposable folded capillary cells. The measurements were performed at 25 °C, and experiments were conducted in duplicate. 2.5.3. Transmission electron microscopy (TEM) TEM images of MWCNT-nisin complexes were collected by a Zeiss transmission electron microscope, model EM 109, at 80 kV. A drop of the dispersions was placed in a holey-carbon grid of 300 mesh, coated with carbon film, which was kept overnight in a desiccator containing

(qi, int qi, dil ) i

ni

(2)

Fig. 1. Dispersions of MWCNT + nisA at (a) pH 2, and (b) pH 3, at 25 °C, and TEM pictures obtained from the deposition of MWCNT dispersions, prepared in (c) buffer pH 2 + 1.0 mg·mL−1 nisA and (d) buffer pH 2, on the holey-carbon grids. For the dispersions, peptide concentrations for each experiment increase from 0.15 to 0.50 mg mL−1 from left to right (0.15, 0.20, 0.25, 0.30, 0.35, 0.40, 0.45, and 0.50 mg mL−1 were the evaluated concentrations). Dispersions were obtained after sonication for 5 min and centrifugation (5100 rpm) for 1 h.

3

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

of the mixtures. Concentrations of nisA higher than 0.45 mg mL−1 at this pH were needed to promote the same visual effect observed in dispersions obtained at pH 2. Additionally, at pH values equal to 4 and 5, the dispersions obtained at all nisin concentration ranges did not remain stable after centrifugation (Figure S1), showing the important role of the pH on the MWCNT stabilization in the presence of the peptides. TEM analyses of the MWCNT, in the absence and presence of 1.00 mg mL−1 nisA, were conducted to access the effect of the peptide on the MWCNT structure, yielding a better understanding of the dispersion of MWCNT in solution. Fig. 1c presents a representative image of the nisA-stabilized MWCNT dispersion. The TEM micrograph showed the well isolated state of MWCNT with the aid of nisA compared with the MWCNT without any peptide (Fig. 1d), confirming that nisA has the ability to separate the bundled MWCNT that were aggregated on account of the van der Waals interactions occurring between the walls of the tubes. Furthermore, the translucid image observed in the TEM micrograph when nisA was present suggests that this peptide is near the nanotubes, interacting with them. This behavior has been reported for other solutes, including peptides, bound to MWCNT [23].

silica gel for complete drying. The negative contrast was made using 0.5% uranyl acetate. 2.6. Antimicrobial activity of the MWCNT-peptide complexes 2.6.1. Bacterial strain The antimicrobial activity of MWCNT-peptide complexes was tested against Lactococcus lactis ATCC 19435, which was used as an indicator of nisin activity, using decimal dilutions in Lactobacilli MRS Agar (Difco™). 2.6.2. Agar well diffusion An amount of 100 μL of an overnight culture of L. lactis in MRS broth (Difco™) (≈ 108 CFU mL−1) was added to 10 mL of Lactobacilli MRS Agar (Difco™), mixed thoroughly, poured into sterile empty Petri dishes, and left for 1 h at 25 °C until solidification. Then, 5 mm diameter wells were cut into the surface of the agar, and 25 μL of the suspension containing MWCNT-peptide complexes were added to each well. After 24 h incubation at 35 °C, the inhibition zones were measured using a paquimeter (Mitutoyo®). The buffer citrate-phosphate at the evaluated pH values and deionized water were used as control. Each assay was performed in triplicate.

3.2. Effect of nisin structure and pH on the MWCNT stability in aqueous solutions

2.6.3. Suspension test Tubes of 9 mL sterile MRS broth (Difco™) containing 1 mL of MWCNT-peptide complexes suspension (or 1 mL of citrate-phosphate buffer or water, in the control experiments) were inoculated with 20 μL of L. lactis culture (≈ 108 CFU·mL−1). The tubes were incubated at 35 °C under agitation during 3 h. Then, using decimal dilutions, the number of viable cells was determined by plate count in Lactobacilli MRS Agar (Difco™), incubated at 37 °C for 24 h.

Fig. 2 shows the MWCNT dispersions obtained in the presence of nisZ, under the same experimental conditions used to obtain MWCNT dispersions in the presence of nisA. At pH 2, the same visual aspect observed in the MWCNT dispersions in the presence of nisA was observed for nisZ. However, at pH 3, the effect of the change in the nisin structure on the stability of the dispersions became evident, and 0.30 mg mL−1 nisZ was already sufficient to generate dispersions as stable as those observed at pH 2. At pH 4 and 5 (Figure S1), for all concentrations, dispersions were not stable to centrifugation. In general, the stability of colloidal dispersions results from electrostatic repulsions between the charged MWCNT particles in the system [24,25], suggesting that nisA and nisZ peptides modulated the surface charge of the MWCNT with an extension that depended on both pH and nisin structure. To check this hypothesis and access information on the distribution of charge at the functionalized MWCNT surface, zeta potential measurements (ζ) of the MWCNT in the presence of different concentrations of peptides at different pH values were conducted. The results are presented in Fig. 3. In the absence of peptides, the MWCNT surface charge was negative at pH 5, becoming less negative as the pH decreased to 3. At pH 2, ζ was positive, suggesting that the change in the MWCNT surface charge was due to the presence of functional groups on the tubes that were protonated/deprotonated by changing the pH and/or the ions of the buffer adsorbed on the nanomaterial. Despite this, the charge at the MWCNT surface was not enough to promote MWCNT stabilization. At pH 2, in the presence of nisA (Fig. 3a), the ζ values increased as the peptide concentration increased to 0.15 mg mL−1, reaching a value

3. Results and discussion 3.1. Stability of MWCNT dispersions in the presence of nisin The ability of a solute to bind to carbon nanotube structures and promote their dispersion in aqueous solution depends strongly on the solute structure, which can determine the MWCNT applications. To check the ability of nisin peptide to stabilize MWCNT in aqueous solutions, dispersions were prepared by the interaction of the peptide with the carbon nanoaggregates in buffered solutions at different pH. The MWCNT dispersions obtained in the presence of nisA after sonication for 5 min, followed by centrifugation at 5100 rpm for 1 h, are shown in Fig. 1 with peptide concentrations ranging from 0.15 to 0.50 mg mL−1, at pH 2 and 3. In the absence of the peptide, MWCNT dispersions were unstable, while concentrations of nisA higher than 0.15 mg mL−1 generated stable MWCNT dispersions at pH 2, indicating that the peptide favored the stabilization mechanisms of the MWCNT. As the pH increased to pH 3, the ability of the peptide to stabilize the MWCNT dispersions was reduced, and concentrations equal to 0.25 mg mL−1 of nisA produced dispersions with low MWCNT content, as observed by the transparency

Fig. 2. Dispersions of MWCNT + nisZ at (a) pH 2 and (b) pH 3, at 25 °C. The peptide concentration for each experiment increases from 0.15 to 0.50 mg mL−1 from left to right (0.15, 0.20, 0.25, 0.30, 0.35, 0.40, 0.45, and 0.50 mg mL−1 were the evaluated concentrations). Dispersions were obtained after 5 min sonication and 1 h centrifugation (5100 rpm).

4

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

Fig. 3. Zeta potential measurements for MWCNT dispersions versus concentration of (a) nisA and (b) nisZ at different pH values.

of +41.7 mV, showing that nisA affected the charge in the double electric layer of the MWCNT. From 0.15 mg mL−1, ζ remained almost constant, and the ζ versus nisA concentration curve displayed a plateau. Similar behaviors were observed for the other pH values, but the increase in pH promoted a marked decrease in the ζ values for all nisA concentrations. At pH 4 and 5, for example, the ζ values were lower than +15 mV for all concentrations. This last result suggests that unstable MWCNT dispersions exhibit low electrostatic stabilization of nanoaggregates in solution at these pH values. However, at pH 2 and 3, the higher ζ values indicated a higher electrostatic repulsion between the charged MWCNT, which promoted the electric stabilization of the dispersions. When nisA was replaced with nisZ (Fig. 3b), the profiles of the ζ versus concentration curves were the same. The plateau observed at higher nisZ concentrations appeared at the same peptide concentration and with the same ζ value as that observed for nisA, for each pH evaluated, showing that the change in the nisin structure did not affect the average distribution of the surface charge of the nanotubes for the same total peptide concentration. Thus, the distribution of charge on the MWCNT surface could not explain the differences between the abilities of nisA and nisZ to stabilize MWCNT dispersions at pH 3. For example, with the same ζ value at 0.30 mg mL−1 for both peptides, +(28 ± 2) mV, different abilities to stabilize the dispersions were observed (Figs. 1b and 2 b). Our results suggested that, although the repulsion electrostatic effect was the most responsible for the stabilization mechanism of the MWCNT dispersions, it was not the only one. To better understand the effect of pH on the ζ values for the MWCNT dispersions for each peptide, it is important to know the charge of both nisA and nisZ as a function of pH. Both nisA and nisZ are characterized by the absence of residues with negatively charged side chains, implying that these peptides are positively charged over a wide pH range, including the pH values evaluated here [20]. This suggests that the increase in the ζ values of the MWCNT aggregates with increase in the peptide concentration resulted from the transfer of the positively charged peptides from the solution to the MWCNT surface by means of an adsorption process. This process led to the formation of MWCNTnisin biocomplexes, as suggested by TEM analysis. As the pH increased, the positive charge of the peptides decreased because of the deprotonation of the amino acids residues containing the ionizable functional groups, corroborating with the smaller ζ values of the biocomplexes at the highest pH values. Despite the fact that both peptides were positively charged, the change in the histidine in nisA by the asparagine in the nisZ structure provides an extra ionizable side chain (pKa = 6.04 for histidine residue) to nisA, causing nisA to be a more positively charged peptide [20]. This charge difference, however, was not enough to promote the significant

differences between the charge distribution of the MWCNT-nisin biocomplexes (Fig. 3), showing that nisA and nisZ had different abilities to bind to the MWCNT surface. To access this information, adsorption studies were conducted. 3.3. Adsorption studies of the peptides on MWCNT Detailed adsorption studies of the peptides nisA and nisZ on the MWCNT surface can provide valuable information on the MWCNTpeptide complexes, showing the real effect of the change in only one amino acid residue in the nisin structure on the ability of the peptide to stabilize the MWCNT dispersions. Fig. 4 shows the adsorption isotherms of nisA and nisZ on MWCNT, at pH 2 and 25.0 °C. The isotherms of both peptides presented a very high inclination in the lower equilibrium concentration (Ce ) values, characterizing “H” isotherms that indicate that the peptides are bound strongly to the adsorption sites on the adsorbent [26]. After the initial abrupt growth, the amount of peptide adsorbed (qP ) reached a maximum value (qP, max ) that remained almost constant, suggesting that the MWCNT surface became saturated by the peptides. For nisA, qP, max was attained at approximately 175 mg g−1, as Ce increased to 0.050 mg mL−1. For nisZ, the Ce value to attain qP, max was 0.150 mg mL−1, and the amount of nisZ

Fig. 4. Adsorption isotherms of nisA and nisZ on the MWCNT surface, at 25.0 °C and pH 2. 5

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

adsorbed in the plateau was higher than 223 mg g−1, showing the higher efficiency of nisZ in recovering the MWCNT surface under saturation conditions. Interestingly, the increase in the ζ values as the total peptide concentration increased (Fig. 3) followed the increase in qP , confirming that the increase in the surface charge of the MWCNT was due to the adsorbed nisin amount. However, the first concentration in the plateau of the ζ versus nisin concentration curve was smaller than the total concentration of nisin needed to promote the saturation of the MWCNT surface (Figure S2), showing that beyond a specific peptide concentration, nisin adsorption did not change the media distribution charge on the surface, even at the pH in which the peptide charge was higher (pH = 2). The buffer ions were most likely strongly attracted to the double electric shell of the biocomplex, shielding the extra charge acquired by the adsorption process. The difference in the adsorption capacity of the MWCNT, observed for distinct types of solutes, at fixed temperature and pressure, depends on several factors, such as the delicate balance of the intermolecular interactions that take place when the adsorbate goes from the solution to the adsorbent surface, and the molecular size of the adsorbate [27,28]. Regarding the molecular size, larger molecules tend to occupy higher surface area per molecule on the surface, and they can be prevented from accessing some narrow porous regions in the adsorbent, leading to a smaller adsorbed amount. However, nisA and nisZ have very similar molecular sizes and weights, which cannot explain the large difference observed in the isotherms. Thus, our results revealed that the slight change in the nisin structure, changing only one amino acid residue, affected the balance of the interactions involved during the adsorption processes of the peptides on the MWCNT. These interactions can be electrostatic, van der Waals, hydrogen-bond, and hydrophobic interactions. To obtain more information about the role of electrostatic interactions on nisin adsorption, the effect of the pH on the adsorption isotherms was evaluated (Figure S3). The isotherm profiles were similar for all pH values evaluated, but the effect of the pH change on the amount of the adsorbed peptide was large. Fig. 5 shows the experimental qP, max values for nisA and nisZ as a function of pH. The qP, max values increased with the increase in pH for both peptides, showing that the electrostatic interactions played an important role during the nisin adsorption. Furthermore, for all pH values, the adsorption capacity of MWCNT for nisZ was higher than that for nisA, corroborating with the higher ability of the latter to stabilize the MWCNT dispersions. However, the higher the pH was, the smaller the difference was between the qP, max for both peptides. As the pH increased from 2 to 5, the MWCNT charge changed from positive to negative (Fig. 3), favoring attractive electrostatic interactions with the peptides (positively charged for the entire pH range). However, at low Ce values, when the surface concentration of peptide on the nanotubes was low, the effect of the pH on the qP values was negligible (Figure S3), suggesting that MWCNT-peptide electrostatic interactions were not the main forces that determined the qP, max values. In addition to MWCNT-nisin electrostatic interactions, as nisin adsorption occurred and qP increased, positively charged peptides approximated each other on the surface, increasing the electrostatic repulsion among them. This nisin-nisin repulsion was higher as higher was the charge of the peptide, i.e., at pH 2, making qP, max at this pH smaller for both nisA and nisZ in comparison with the other values. Furthermore, because of the extra charge in nisA, the nisin-nisin electrostatic repulsion was higher, contributing to smaller qP, max values for this peptide. However, as the pH increased, the positive charge of the peptides decreased, reducing the nisin-nisin electrostatic repulsion for both peptides, causing the qP, max values for nisA and nisZ to approach each other. For Ce values smaller than those under saturation conditions of the MWCNT surface, in which the nisin-nisin electrostatic repulsion was reduced, and MWCNT-nisin electrostatic interactions were not very

important in promoting large differences in qP values, the adsorption amounts cannot be explained by just the charge in the species involved during the adsorption process. This suggests that non-electrostatic interactions determined the abilities of the peptides to bind to the MWCNT, promoting the stabilization of the biocomplexes. Among all the interactions involved in the adsorption of solutes on MWCNT, the hydrophobic ones are very important because of the hydrophobic character of the nanotube walls [29]. Therefore, amphiphilic molecules can interact with the MWCNT, forming hydrophobic interactions through their non-polar part, maintaining the interaction of the polar region with the solvent in the solution. This adsorption process decreases the interfacial tension on the biocomplex-solution interface and promotes a steric effect that contributes to the stabilization of the biocomplex in the solution. The structure of nisA in the aqueous solution, determined by van de Ven et al. [30] using NMR, shows that this peptide contains two domains (N-terminal and C-terminal). The Nterminal domain presents an amphiphilic character, while the C-terminal contains the positively charged amino acid residues, with a less clear amphiphilic character. The hydrophobic part in the N-terminal domain of the nisA most likely interacted with the MWCNT surface through hydrophobic interactions. When the residue histidine present in the nisA C-terminal domains was replaced with asparagine in the nisZ structure, causing this domain to be less hydrophilic in the evaluated pH range, the hydrophobic interactions of nisZ with the MWCNT surface were favored though their two domains, which can explain the higher stability of the MWCNT dispersions, formed with this peptide at pH 3. This structural difference has also been used to explain the lower solubility of the nisZ at low pH values [20]. These findings reveal the important role of the hydrophilic/hydrophobic balance of peptides to determine the adsorption process, modulating the biocomplex stability. Understanding the energetics associated with nisin variants adsorption will provide the real effect of this balance on the adsorption process, important to modulate the adsorption of other biomolecules on CNT. 3.4. Adsorption thermodynamic analysis To obtain more information on the effect of the nisin structure on its ability to bind to MWCNT surface, we accessed the magnitude of the energy involved in the adsorption of the peptides using isothermal titration calorimetry (ITC). ITC can determine precisely the enthalpy change associated with the adsorption of solutes at different extensions

Fig. 5. The effect of pH on the experimental maximum adsorbed amount of nisA and nisZ on the MWCNT surface at 25.0 °C. 6

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

of surface coverage [28]. Fig. 6 presents the adsorption molar enthalpy change ( Hads ) versus qP for nisA and nisZ adsorption on MWCNT, at pH 2 and 25.0 °C. Adsorption of nisA on the MWCNT surface was an exothermic process for all evaluated qP values. However, as the qP increased, the Hads values became less negative, indicating that the adsorption process involved sites on the MWCNT with which the nisA interacted, releasing different molar enthalpic energies. To better understand this result, the Hads values were observed can be considered, for each qP value, as the sum of the following independent terms in Eq. (3): Hads =

Hads des +

Hads P

S

+

Hads P

P , sur

+

Hads P

P , bulk

+

Hads P

cc

changed during the recovery of the surface. As qP increased, the peptide most likely interacted with sites whose exothermic contribution of P S and ads H des terms were less negative. This result agreed with ads H the heterogeneity of the MWCNT adsorbent used in this work, as discussed by Ferreira et al., [28] which suggested that Ponceau red and Allura red adsorbed preferentially on sites with which the dye-MWCNT site interaction was more exothermic. Our efforts to identify the individual contributions for the Hads values of nisA provided a direction for understanding the effect of the peptide structure on the adsorption process at the MWCNT surface. The Hads versus qP curve for nisZ at pH 2 (Fig. 6) displayed some similar behaviors to that obtained for nisA. However, the Hads versus qP curve for nisZ was almost linear, with Hads values less negative than Hads values for nisA at qP values smaller than 135 mg g−1. This difference revealed that the replacement of only one amino acid residue in the nisin structure affected the intermolecular interactions between the peptide and the MWCNT sites. At pH 2, nisZ, as a more hydrophobic peptide, presents a lower positive charge than nisA. Because of the reduced charge, nisZ has lower solubility in water than nisA, suggesting that this late nisin variant interacted more intensely with the solvent molecules through its Cterminal region [20]. According to our previous hypothesis, the charged C-terminal region of nisA preferentially interacted with the solvent molecules at the surface, while the C-terminal region of nisZ, with a more hydrophobic character, interacted with the hydrophobic surface of the MWCNT. At the same time, the negative charge on the MWCNT surface promoted a higher electrostatic repulsion against nisA than against nisZ, causing the ads HP S term in Eq. (3) to be more negative for the latter, which could not explain the more negative Hads values for nisA. However, because of the establishment of the hydrophobic interactions between nisZ and the MWCNT, involving a higher number of amino acid residues than the interaction with nisA, there should be a greater amount of water molecules to be released from the solvation shell. Water molecules solvating the hydrophobic regions are more structured than water molecules in the bulk solution, through more intense interactions with hydrogen bonds. As a consequence, the desolvation process of hydrophobic molecules generally absorb energy, [31] and the term ads H des in Eq. (3) should be more positive (or less negative) for nisZ, i.e., more energy is needed to desolvate nisZ and the MWCNT surface, contributing to less negativity in Hads for nisZ. To evaluate more precisely the effect of this structural change, the Hads versus qP curves for nisA and nisZ were compared at the other

(3)

is the enthalpy change due the desolvation process where ads when 1 mol of peptide adsorbs onto the MWCNT, including (i) the breaking of interactions to remove the water molecules and buffer ions from the peptide and adsorbent solvation shells, (ii) the formation of interactions involving water molecules and buffer ions in the bulk, and (iii) the change in the solvation shell structure of the MWCNT due to the adsorption process (depending on the energetic balance between the interactions formed and disrupted during the process, ads H des can be positive or negative); ads HP S is the molar enthalpy change for the formation of the interaction between the peptide and the MWCNT adsorption sites and is always negative; and the ads HP P , sur and P P , bulk terms are the molar enthalpy changes resulting from the ads H nisin-nisin interactions on the adsorbent surface (formed interactions) and in the solution (disrupted interactions), respectively. As the adsorption process takes place, the term ads HP P , sur should be positive because it involves the repulsive steric and electrostatic interactions between the positively charged peptides on the adsorbent surface. On the other hand, the sign of the term ads HP P , bulk should be negative because the adsorption removes the positively charged peptides from the bulk, breaking the repulsive electrostatic interactions between them (the negative values of the dilution molar enthalpy change in the nisA concentrated solution corroborates this hypothesis). Finally, ads HP cc is the molar enthalpy change associated with the conformational change in the peptide that comes from the adsorption process and is most likely positive. Although we cannot attribute individual values for each term in Eq. (3), the analysis from the Hads versus qP curve for nisA can provide valuable information on the adsorption process. The contribution of the P cc term for Hads is difficult to establish. However, both rigid ads H domains (N- and C-terminal) of nisA are connected by a flexible hinge region, causing nisA to be a flexible molecule in aqueous solution [30]. Then, ads HP cc should have a magnitude on the order of kT (less than 3.0 kJ mol−1), contributing very little to the Hads magnitude. Additionally, as observed in the thermogram of the dilution of the nisA concentrated solution in the solvent (data not presented), the energy associated with the nisA dilution process was low and almost constant (approximately −3.0 kJ mol−1), indicating that the energy associated with the interactions between nisA molecules in the bulk ( ads HP P , bulk ) did not determined the highly exothermic values of Hads . Thus, the negative values of Hads observed in the adsorption process of nisA showed that the global exothermic contributions in the terms ads H des and ads HP S overcame the endothermic ones, and the exothermicity of the adsorption process was due to the specific interactions between the peptide and the MWCNT surface, and the desolvation processes. The increase in the Hads values as qP increased showed that the balance of the intermolecular interactions during adsorption was altered by the recovery of the MWCNT surface. As the surface coverage increased, the surface area available per adsorbate molecule was reduced, increasing the electrostatic repulsion between the molecules of adsorbed nisA, with a consequential increase in the contribution of the term ads HP P , Sur , making Hads less exothermic. Greater increases in the nisA surface concentration are expected to intensify this effect. However, in the Hads versus qP curve, the slope was slightly higher for the lower qP values, suggesting that the nature of the adsorption sites

H des

Fig. 6. Adsorption molar enthalpy change ( Hads ) versus qP curves for adsorption of nisA and nisZ on MWCNT, at 25.0 °C and pH 2. 7

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

evaluated pH values (Figure S4). At pH 3, the Hads versus qP curve profiles for nisA and nisZ are very similar to those obtained at pH 2. However, when the pH increased to 4 and 5, the opposite tendency was observed, and the Hads values were more negative for nisZ than for nisA. As the pH increased to 4 and 5, the positive charge of the nisin variants decreased, and the non-electrostatic interactions began to determine the Hads values. In this pH range, with the reduction in the charge of the C-terminal domain of nisA, the hydrophobic interactions between this domain of the peptide and the hydrophobic surface of the MWCNT are favored, and the ads H des of both peptides approaches each other. However, the structural difference between the peptides modified the magnitude of the intermolecular interactions of their C-terminal domain with the nanotube surface. While the lateral chain of histidine in nisA is formed by an imidazole ring, the asparagine residue lateral chain in nisZ contains an amide group that can actuate as a donor and acceptor of hydrogen bonds, binding to polar groups in the MWCNT. This strong intermolecular interaction contributed to make P S more negative for nisZ, explaining why the Hads values beads H comes more negative for this peptide at higher pH values. These results have shown how the change in only one amino acid residue can directly affect the adsorption process of the nisin peptide due to the specific interactions that each amino acid residue can perform with the surface of the CNT. This is a valuable information in the context of the adsorption of biomolecules (proteins and peptides) on how the replace of an amino acid residue or a small change in the amino acid sequence in the biomolecule chain can affect the enthalpy change associated with the process and, consequently, the properties of the formed complex. To access the thermodynamic driven forces for the nisin variant adsorptions, corroborating our previous hypothesis, adsorption thermodynamic parameters were obtained in the condition of infinite dilution ( ads G , ads H , and ads S ). The ads G values were obtained from the Langmuir constant (KL ) according Eq. (4). ads

G =

RTlnKL

pH equal to 2 and 3, nisZ, which exhibited a more hydrophobic character with a small positive charge, released a larger number of water molecules from the non-polar regions, resulting in a ads S that was less negative than that for nisA. This difference almost disappeared at the higher pH values, for which the hydrophobic interactions are important for both nisin variants. 3.5. Antimicrobial effect of MWCNT-peptide complexes The adsorption of nisin on a surface involved a series of interactions, in which the peptides can adopt different orientation and conformation. Such changes may impact the antimicrobial properties and/or performance of the biomaterial [32]. For this reason, it is essential to study the antimicrobial activity of MWCNT-peptide complexes. Once adsorbed on MWCNT, nisin variants maintained its antimicrobial activity, and inhibition zones were visible in the surrounding wells (p < 0.05). The MWCNT-adsorbed peptides showed similar antimicrobial activity against L. lactis, as opposed to the isolated peptides (Figure S5), suggesting that possible changes in the nisin conformation did not affect its activity when nisin was adsorbed onto MWCNT. Other researchers have also observed that bacteriocins, such as nisin and pediocin, have exerted their antibacterial activity, even when they were absorbed in different surfaces, such as rubber, stainless steel, polyethyleneterephthalate, [33] and low-density and acrylic-coated polyethylene [32]. On the other hand, the controls used in our experiment (buffers, deionized water, and pure MWCNTs) did not show antimicrobial action (Figure S5c). Despite the fact that the adsorption capacity of nisZ on MWCNT was higher than that of nisA, there was no difference in the inhibition zone diameter caused by both nisin variants (p > 0.05). The average diameters of inhibition zones for L. lactis growth are shown in Table 2. There are two possibilities suggested for the antimicrobial action of nisin variants. The first one is the nisin desorption from MWCNT, stimulated by the incubation temperature (37 °C), followed by its penetration through the bacterial membrane, leading to pore formation and cell collapse [18]. The second proposal is that the MWCNT-nisin complex crossed the bacterial membrane delivering the bacteriocin. Some studies from the food and medical fields have reported the conjugation of antimicrobial molecular activity with carbon nanotubes [34,35]. To diminish the restriction of diffusion observed in the well diffusion test, the antimicrobial activity of nisin was also evaluated in a suspension test. After 3 h of bacterial suspension (8.39 log CFU mL−1) treated with nisA, nisZ, or with functionalized MWCNT (MWCNT-nisA and MWCNT-nisZ), there was an exceptional reduction in the count of L. lactis (> 8 log cycles) (p < 0.05). On the other hand, a reduced count was not observed for the control bacterial suspension and that treated with only MWCNT. The lack of antimicrobial activity of pure MWCNT was attributed to its aggregation and, thus, low diffusion in the medium. Some studies have demonstrated the improvement in the antimicrobial activity of MWCNT by functionalization with peptides and other molecules [36,37].

(4)

Adjust parameters for the Langmuir model are in Table S2. To ob0 , and tain ads H , ads H versus qP curves were extrapolated to qP T ads S were obtained from the thermodynamic fundamental equation T ads S ). Table 1 shows the thermodynamic ( ads G = ads H parameters obtained. The ads G values for nisA and nisZ were negative, indicating that the peptides predominantly adsorbed on the surface in the equilibrium state. The adsorption was enthalpically driven, and as the pH increased, ads H became more negative for both nisA and nisZ. Furthermore, the process was entropically unfavorable, indicating that the peptides adsorbed with decrease in the configurational entropy of the system. The pH increase also caused ads S to be more negative, as ads S for nisA was more negative than for nisZ when the pH was smaller than 4. At pH 4 and 5, smaller differences between the ads S values for nisA and nisZ were prevalent, as the process was more entropically unfavorable for nisZ. Under infinite dilution conditions, the nisin molecules do not interact with one another. Thus, the ads H parameter can be considered as a sum of the same terms in Eq. (3), except those associated with nisin-nisin interactions ( ads HP P , sur + ads HP P, bulk ). The more negative ads H values at the higher pH values for each nisin indicated that the nisin-MWCNT electrostatic interactions had a large contribution to the term ads HP S . The difference between the ads H values for nisA and nisZ was due to the same reason discussed about the difference between the Hads versus qP curves for both peptides. Regarding the entropic contribution for the adsorption of each peptide, the increase in the positive charge of the nisin variants as the pH decreased likely increased the number of buffer ions that were released together with the water molecules from the electric double shell of the MWCNT surface, partially counterbalancing the configurational entropy loss and causing ads S to be less negative at the lower pH. At

Table 1 Adsorption thermodynamic parameters of nisin peptides on MWCNT for infinite dilution condition at different pH values. System nisA

nisZ

8

ads

pH pH pH pH pH pH pH pH

2 3 4 5 2 3 4 5

G (kJ mol−1)

−42.60 −41.14 −39.63 −37.35 −40.40 −40.13 −37.57 −35.01

ads

H (kJ mol−1)

−67.9 −75.9 −78.1 −98.4 −40.5 −45.1 −83.5 −99.7

T

ads

S (kJ mol−1)

−25.30 −34.76 −38.47 −61.05 −0.10 −4.97 −45.93 −64.69

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

References

Table 2 The average diameter of the inhibition zone for L. lactis. Treatments

Inhibition zone diameter / cm

Deionized water Citrate-phosphate buffer MWCNT MWCNT – nisA MWCNT – nisZ nisA nisZ

0.00 0.00 0.00 2.02 2.11 2.01 2.04

± ± ± ± ± ± ±

[1] S. Iijima, Helica microtubules of graphitic carbon, Nature 354 (1991) 56–58. [2] Y. Huang, L. Mei, X. Chen, Q. Wang, Recent developments in food packaging based on nanomaterials, Nanomaterials 8 (2018) 1–29. [3] M. Mohajeri, B. Behnam, A. Sahebkar, Biomedical applications of carbon nanomaterials: drug and gene delivery potentials, J. Cell. Physiol. 234 (2018) 298–319. [4] X. Chen, Z. Zhao, Y. Zhou, Y. Shu, M. Sajjad, Q. Bi, Y. Ren, X. Wang, X. Zhou, Z. Liu, Mwcnts modified α-fe2o3 nanoparticles as anode active materials and carbon nanofiber paper as a flexible current collector for lithium-ion batteries application, J. Alloys. Compd. 776 (2019) 974–983. [5] S. Javadian, A. Motaee, M. Sharifi, H. Aghdastinat, F. Taghavi, Dispersion stability of multi-walled carbon nanotubes in catanionic surfactant mixtures, Colloids Surf. A Physicochem. Eng. Asp. 531 (2017) 141–149. [6] A. Alasli, T. Evgin, A. Turgut, Re-dispersion ability of multi wall carbon nanotubes within low viscous mineral oil, Colloids Surf. A Physicochem. Eng. Asp. 538 (2018) 219–228. [7] M. Manilo, K. Bohacs, N. Lebovka, S. Barany, Impact of surfactant and clay platelets on electrokinetic potential and size distribution in carbon nanotubes aqueous suspensions, Colloids Surf. A Physicochem. Eng. Asp. 544 (2018) 205–212. [8] K. Zhan, H. Liu, H. Zhang, Y. Chen, H. Ni, M. Wu, D. Sun, Y. Chen, A facile method for the immobilization of myoglobin on multi-walled carbon nanotubes: poly(methacrylic acid-co-acrylamide) nanocomposite and its application for direct bio-detection of H2O2, J. Electroanal. Chem. Lausanne (Lausanne) 724 (2014) 80–86. [9] Q. Zhao, K. Yang, S. Zhang, B. Chefetz, J. Zhao, H. Mashayekhi, B. Xing, Dispersant selection for nanomaterials: insight into dispersing functionalized carbon nanotubes by small polar aromatic organic molecules, Carbon 91 (2015) 494–505. [10] L. Meng, C. Fu, Q. Lu, Advanced technology for functionalization of carbon nanotubes, Prog. Nat. Sci. 19 (2009) 801–810. [11] H. Tan, W. Feng, P. Ji, Lipase immobilized on magnetic multi-walled carbon nanotubes, Bioresour. Technol. 115 (2012) 172–176. [12] Z. Chen, B. Mai, H. Tan, X. Chen, Nucleic acid based nanocomposites and their applications in biomedicine, Compos. Commun. 10 (2018) 194–204. [13] S. Mahajan, A. Patharkar, K. Kuche, R. Maheshwari, P.K. Deb, K. Kalia, R.K. Tekade, Functionalized carbon nanotubes as emerging delivery system for the treatment of cancer, Int. J. Pharm. 548 (2018) 540–558. [14] M. Sheikholeslam, M. Pritzker, P. Chen, Dispersion of multiwalled carbon nanotubes in water using ionic-complementary peptides, Langmuir 28 (2012) 12550–12556. [15] A. Antonucci, J. Kupis-Rozmyslowicz, A.A. Boghossian, Noncovalent protein and peptide functionalization of single-walled carbon nanotubes for biodelivery and optical sensing applications, ACS Appl. Mater. Interfaces 9 (2017) 11321–11331. [16] S.L. Yoong, W.L. Lau, A.Y. Liu, D.A. Prendergast, H.K. Ho, V.C.K. Yu, C. Lee, W.H. Ang, G. Pastorin, Mitochondria-acting hexokinase ii peptide carried by shortlength carbon nanotubes with increased cellular uptake, endosomal evasion, and enhanced bioactivity against cancer cells, Nanoscale 7 (2015) 13907–13917. [17] D. Umadevi, S. Panigrahi, G.N. Sastry, Noncovalent interaction of carbon nanostructures, Acc. Chem. Res. 47 (2014) 2574–2581. [18] J.C.P. Santos, R.C.S. Sousa, C.G. Otoni, A.R.F. Moraes, V.G.L. Souza, E.A.A. Medeiros, P.J.P. Espitia, A.C.S. Pires, J.S.R. Coimbra, N.F.F. Soares, Nisin and other antimicrobial peptides: production, mechanisms of action, and application in active food packaging, Innov. Food Sci. Emerg. Technol. 48 (2018) 179–194. [19] J.W.M. Mulders, I.J. Boerrtgter, H.S. Rollema, R.J. Siezen, W.M. Vos, Identification and characterization of the lantibiotic nisin z, a natural nisin variant, Eur. J. Biochem. 201 (1991) 581–584. [20] H.S. Rollema, O.P. Kuipers, P. Both, W.M. De Vos, R.J. Siezen, Improvement of solubility and stability of the antimicrobial peptide nisin by protein engineering, Appl. Environ. Microbiol. 61 (1995) 2873–2878. [21] X. Qi, G. Poernomo, K. Wang, Y. Chen, M.B. Chan-Park, R. Xu, M.W. Chang, Covalent immobilization of nisin on multi-walled carbon nanotubes: superior antimicrobial and anti-biofilm properties, Nanoscale 3 (2011) 1874–1880. [22] X. Dong, L. Yang, Dual functional nisin-multi-walled carbon nanotubes coated filters for bacterial capture and inactivation, J. Biol. Eng. 9 (2015) 20. [23] M.S. Deshpande, S. Mazumdar, Sequence specific association of tryptic peptides with multiwalled carbon nanotubes: effect of localization of hydrophobic residues, Biomacromolecules 13 (2012) 1410–1419. [24] R.C. Murdock, L. Braydich-Stolle, A.M. Schrand, J.J. Schlager, S.M. Hussain, Characterization of nanomaterial dispersion in solution prior to in vitro exposure using dynamic light scattering technique, Toxicol. Sci. 101 (2008) 239–253. [25] J. Jiang, G. Oberdörster, P. Biswas, Characterization of size, surface charge, and agglomeration state of nanoparticle dispersions for toxicological studies, J. Nanopart. Res. 11 (2008) 77–89. [26] G. Limousin, J.P. Gaudet, L. Charlet, S. Szenknect, V. Barthès, M. Krimissa, Sorption isotherms: a review on physical bases, modeling and measurement, Appl. Geochem. 22 (2007) 249–275. [27] Z. Chen, D. Pierre, H. He, S. Tan, C. Pham-Huy, H. Hong, J. Huang, Adsorption behavior of epirubicin hydrochloride on carboxylated carbon nanotubes, Int. J. Pharm. 405 (2011) 153–161. [28] G.M.D. Ferreira, G.M.D. Ferreira, M.C. Hespanhol, J. de Paula Rezende, A.C. dos Santos Pires, L.V.A. Gurgel, L.H.M. da Silva, Adsorption of red azo dyes on multiwalled carbon nanotubes and activated carbon: a thermodynamic study, Colloids Surf. A Physicochem. Eng. Asp. 529 (2017) 531–540. [29] N. Saifuddin, A.Z. Raziah, A.R. Junizah, Carbon nanotubes: a review on structure and their interaction with proteins, J. Chem. 2013 (2013) 1–18. [30] F.J.M.V.D. Ven, H.W.V.D. Hooven, R.N.H. Konings, C.W. Hilbers, Nmr studies of

0.00a 0.00a 0.00a 0.04b 0.09b 0.07b 0.05b

a,b

Averages followed by the same letter in the column did not differ between them using Tukey test (p > 0.05). The diameter values include the well diameter (0.5 cm).

4. Conclusion The adsorption of two nisin variants, nisin A and nisin Z, on MWCNT was studied, using the analysis of adsorption thermodynamic parameters to highlight the structural features of the peptides. The stability of the dispersions formed by the nisin-MWCNT complex was evaluated, and the bionanocomplexes were characterized, with the evaluation of their antimicrobial activity. Driven by electrostatic forces, the nisin-MWCNT dispersion stability was dependent on the concentration and structure of the peptide, and on the pH of the solution. The adsorption capacity of MWCNT was higher for nisin Z than for nisin A for all evaluated conditions of pH. The isothermal titration microcalorimetry technique was demonstrated to be suitable for the direct determination of the adsorption enthalpy changes associated with the nisin adsorption. Enthalpy was the force governing the adsorption of the peptides on MWCNT, with nisin A interacting more exothermically with the surface at the lowest pH values. The results suggested that at low amounts of adsorbed peptides and low pH values, the more hydrophobic character of nisin Z favored interactions with the most hydrophobic sites on the surfaces of adsorbents. Interestingly, the adsorbed nisin retained its antimicrobial activity in the solid and liquid medium, which is important for developing nisin-based biomaterials. Our results have provided insights to design new biocomplex CNTpeptides by modulating the hydrophilic/hydrophobic balance of peptides from the small changes in the amino acids residues in the peptide structure. Conflict of interest Nothing declared. Acknowledgments The authors are grateful for the financial support provided by Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Pesquisa e Desenvolvimento Tecnológico (CNPq), Instituto Nacional de Ciências e Tecnologias Analíticas Avançadas (INCTAA), and Financiadora de Estudos e Projetos (FINEP). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES, Finance Code 001). We are also grateful to Núcleo de Microscopia e Microanálise from Federal University of Viçosa, specially Karla and Gilmar, for the TEM analisys, and to Clascídia Aparecida Furtado, for the relevant discussions. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.colsurfa.2019.123611. 9

Colloids and Surfaces A 578 (2019) 123611

G.M.D. Ferreira, et al.

[31] [32] [33] [34]

lantibiotics: the structure of nisin in aqueous solution, Eur. J. Biochem. 202 (1991) 1181–1188. W. Blokzijl, J.B.F.N. Engberts, Hydrophobic effects. Opinions and facts, Angew. Chem. Int. Ed. Engl. 32 (1993) 1545–1579. L. Karam, C. Jama, N. Nuns, A.S. Mamede, P. Dhulster, N.E. Chihib, Nisin adsorption on hydrophilic and hydrophobic surfaces: evidence of its interactions and antibacterial activity, J. Pept. Sci. 19 (2013) 377–385. N.P. Guerra, A.B. Araujo, A.M. Barrera, A.T. Agrasar, C.L. Macías, J. Carballo, L. Pastrana, Antimicrobial activity of nisin adsorbed to surfaces commonly used in the food industry, J. Food Prot. 68 (2005) 1012–1019. M.E. Foo, P. Anbu, S.C.B. Gopinath, T. Lakshmipriya, C.-G. Lee, H.S. Yun, M.N.A. Uda, A.R.W. Yaakub, Antimicrobial activity of functionalized single-walled

carbon nanotube with herbal extract of hempedu bumi, Surf. Interface Anal. 50 (2018) 354–361. [35] P. Parasuraman, V.T. Anju, S.B. Sruthil Lal, A. Sharan, S. Busi, K. Kaviyarasu, M. Arshad, T.M.S. Dawoud, A. Syed, Synthesis and antimicrobial photodynamic effect of methylene blue conjugated carbon nanotubes on e. Coli and s. Aureus, Photochem. Photobiol. Sci. 18 (2019) 563–576. [36] A.J. Haider, T.R. Marzoog, I.H. Hadi, Z.N. Jameel, A new method of functionalized multi walled carbon nanotubes by natural oil for microorganism cells detection, AIP Conf. Proc. 1968 (2018) 030013. [37] S. Lohan, K. Raza, S. Singla, S. Chhibber, S. Wadhwa, O.P. Katare, P. Kumar, B. Singh, Studies on enhancement of anti-microbial activity of pristine mwcnts against pathogens, AAPS PharmSciTech 17 (2016) 1042–1048.

10