Vesicles from pH-regulated reversible gemini amino-acid surfactants as nanocapsules for delivery

Vesicles from pH-regulated reversible gemini amino-acid surfactants as nanocapsules for delivery

Accepted Manuscript Title: Vesicles from pH-regulated reversible gemini amino-acid surfactants as nanocapsules for delivery Author: Jing Lv Weihong Qi...

1MB Sizes 0 Downloads 12 Views

Accepted Manuscript Title: Vesicles from pH-regulated reversible gemini amino-acid surfactants as nanocapsules for delivery Author: Jing Lv Weihong Qiao Zongshi Li PII: DOI: Reference:

S0927-7765(16)30484-2 http://dx.doi.org/doi:10.1016/j.colsurfb.2016.06.054 COLSUB 7997

To appear in:

Colloids and Surfaces B: Biointerfaces

Received date: Revised date: Accepted date:

9-3-2016 23-6-2016 27-6-2016

Please cite this article as: Jing Lv, Weihong Qiao, Zongshi Li, Vesicles from pH-regulated reversible gemini amino-acid surfactants as nanocapsules for delivery, Colloids and Surfaces B: Biointerfaces http://dx.doi.org/10.1016/j.colsurfb.2016.06.054 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Vesicles from pH-regulated reversible gemini amino-acid surfactants as nanocapsules for delivery Jing Lv a, Weihong Qiao a,* and Zongshi Li a a

State Key Laboratory of Fine Chemicals, School of Chemical Engineering, Dalian University of Technology, Dalian, 116024, P. R. China

1

Graphical abstract

2

2Na+

~

H

ëõõ.- :õõë -:

KJ KJi

2m 叫 C ;;; - - - - - - -

-C H2n 叫

n

Highlights

Reversible transition from micelle to vesicle is accomplished by regulating pH. The key to pH-regulated gemini amino-acid surfactants is the protonation between H+ and -N-CH2COO-. Vesicles fabricated as nanocapsules can be effectively transported into cells.

3

ABSTRACT

Reversible transition from micelles to vesicles by regulating pH were realized by gemini aminoacid surfactants N,N’-dialkyl-N,N’-diacetate ethylenediamine. Measurement results of ζ-potential at different pH and DLS at varying solvents revealed that the protonation between H+ and double -N-CH2COO- groups (generating -NH+-CH2COO-), expressed as pKa1 and pKa2, is the key driving force to control the aggregation behaviors of gemini surfactant molecule. Effect of pH on the bilayer structure was studied in detail by using steady-state fluorescence spectroscopy of hydrophobic pyrene and Coumarin 153 (C153) respectively and fluorescence resonance energy transfer (FRET) from C153 to Rhodamine 6G (R6G). Various pH-regulated and pH-reversible self-assemblies were obtained in one surfactant system. Vitamin D3 was encapsulated in vesicle bilayers to form nano-VD3-capsules as VD3 supplement agent for health care products. By using the electrostatic attraction between Ca2+ and double -COO- groups, nano-VD3-capsules with Ca2+ coated outermost layer were prepared as a formulation for VD3 and calcium co-supplement agent. DLS and TEM were performed to check stability and morphology of the nano-capsules. It is concluded that the pH-regulated gemini amino-acid surfactants can be used to construct colloidal systems for delivering hydrophobic drugs or nutritions without lipids at human physiological pH level. Keywords: gemini amino-acid surfactant; pH-regulated; vesicle; mechanism; nanocapsule

4

1. Introduction Surfactant based self-assemblies, particularly in the form of micelles and vesicles, have gained widespread application in biosciences [1], functional soft materials [2,3], drug [4] and gene deliveries [5], etc. The precise control over the self-assembling process, including assembling time, condition of vesiculation, vesicle’s size and stability [6], is crucial for enabling these applications. Various methods have been introduced to control the transition between micelles and vesicles [7-9]. Usually, the micelles are formed when the concentrate of the surfactant is above the critical micelle concentrate (“cmc”). The aggregation behaviors are controlled by electrostatic, van der Waals, hydrophobic, or steric interactions as well as their delicate balance [10]. According to different surfactants and conditions, micelles can be finely transformed to vesicles to achieve the most appropriate properties of the aggregates (such as the size, charge, stability, and surface activity). Usually, vesicles are preferably formed by the surfactants with double hydrophobic chains because of the steric interactions among hydrophobic groups. Furthermore, vesicles can also be obtained by mixing cationic and anionic surfactants due to the electrostatic attraction. It is significant to find the guidance to design and control the self-assembled structures of mixed surfactants so as to gain the desired characteristics in applications. Catanionic mixed surfactants have been considered as a promising system to construct complex self-assembled nanostructures [11-15]. To enable controllable or regulated self-assembly systems with special characteristics, new functional surfactants or mixtures of different types of surfactants are constantly being developed and formulated. Zhang et al. [16] recently developed CO2-switchable viscoelastic wormlike micellar fluids by utilizing commercially available low-cost anionic

5

sodium dodecyl sulfate (SDS) and small-molecule tertiary diamine N,N,N’,N’-tetramethyl-1,3propanediamine (TMPDA) with a stoichiometric ratio of 2:1. Due to the CO2-stimuli-responsive TMPDA, the macroscopic rheological properties and self-assembled microstructures can be reversibly tuned by cyclically bubbling and removing CO2. Wang et al. [17] established photocontrolled reversible supramolecular assemblies by synthesizing light-sensitive surfactant 1-[10(4-phenylazophenoxy)decyl]-pyridinium bromide (AzoC10) and compounding with αcyclodextrin (α-CD). The host-guest assembly and disassembly between azobenzene and α-CD by external photostimuli can be used as a smart strategy to build up molecular shuttles, motors, and machines [18,19]. Recently, Lin [20] synthesized a pH-responsive single chain surfactant ndecylphosphoric acid (DPA), which can be switched between two states, sodium decylphosphoric or disodium decylphosphoric, by adjusting pH value. Direct transformation of a “1-2” surfactant pair to a “1-1” pair was realized in the catanionic surfactant mixture of DPA and cetyltrimethylammonium bromide (CTAB) and consequently, pH-regulated self-assembled structures, such as spherical micelle, wormlike micelle, vesicle, and lamellar structure, were fabricated. It is praiseworthy for a single low-molecular-weight surfactant to have complex or subtle selfassembly behaviors in aqueous solution [21-23]. As surfactants are widely used in commercial and industrial fields, it is of great significance to endow surfactants with environmental friendly and biocompatible characteristics. Based on the above considerations, a series of pH-regulated amino-acid surfactants N,N’-dialkyl-N,N’-diacetate ethylenediamine was designed and synthesized (Scheme 1a). A gemini structure was chosen as the molecule skeleton in order to obtain unique self-assembly properties superior to the conventional single-chain surfactants [24]. The low irritating amino-acid type was selected in order to obtain better biocompatibility and

6

environmental safety. A -CH2COOH group was attached to a nitrogen atom belonging to the spacer -N-CH2CH2-N- to endow the surfactant with pH-switchable surface activity. More importantly, double pH-stimuli -N-CH2COO- groups could regulate the gemini surfactant as bivalent, monovalent anionic or zwitterionic by adjusting pH value (Scheme 1b). One attractive approach is to mix different types of surfactants (such as anionic and zwitterionic surfactants, etc.) in the aqueous solution and thus utilize electrostatic effects to regulate self-assemblies at specified pH values. In our previous work, a series of fundamental studies have been made to investigate the surface and colloidal properties of the symmetric Ace(n)-2-Ace(n) and asymmetric Ace(m)-2-Ace(n) solution under varying pH conditions according to pKa values [25,26] (listed in Table 1). It was concluded that N,N’-dialkyl-N,N’-diacetate ethylenediamine had high surface activity at alkaline pH, strong ability to form vesicles at isoelectric point pH range, and pH-recyclability at acidic pH. The vesicles formed in asymmetric Ace(m)-2-Ace(n) solution were observed to undergo a clear size change from large to small and finally swell slightly in a wider pH range from alkaline to weak acid. Both pH-reversibility and pH-regulation of the self-assembly process can be accomplished by this single gemini surfactant.

7

Scheme 1 Design schematic of the gemini pH-sensitive surfactant. a) design principle of N,N’dialkyl-N,N’-diacetate ethylenediamine; b) surfactant type transformation adjusted by pH Table 1 pKa values of N,N’-dialkyl-N,N’-diacetate ethylenediamine Gemini surfactants pKa1 pKa2 Ace(8)-2-Ace(8) 7.59 7.02 Ace(10)-2-Ace(10) 7.83 6.59 Ace(12)-2-Ace(12) 8.41 6.55 Ace(8)-2-Ace(10) 8.44 4.73 Ace(10)-2-Ace(12) 8.40 4.44 Ace(10)-2-Ace(14) 8.31 4.34 Ace(8)-2-Ace(12) 8.51 4.33 Ace(12)-2-Ace(14) 8.10 4.67 Ace(10)-2-Ace(16) 8.33 4.58 Many kinds of lipophilic active components (bioactive lipids, nutraceuticals and drugs) need to be incorporated into aqueous formulations to become suitable for commercial consumption as foods, health care products, and pharmaceuticals [27]. Lipophilic active compounds are usually dissolved in fat- or lipid-soluble medium and emulsified with H2O to achieve a considerable drug loading capacity [28,29]. Therefore, suppliers prefer to prepare oil-in-water emulsions, microemulsions, or nanoemulsions as delivery system for practical applications, which is in conflict with current health recommendations [30]. From a consumer’s perspective, it is generally not desirable to supplement the target nutrition while intaking oils or lipids. In this study, we used detailed physical measurements to investigate the formation and regulation of pH-reversible vesicles and aimed to create an aqueous colloidal system without any lipidic co-solvents to encapsulate Vitamin D in vesicle bilayers formed by single gemini aminoacid surfactant. Vitamin D (VD) plays a critical role in controlling calcium transport, bone metabolism, and renal calcium reabsorption [31], but cannot be synthesized endogenously [32]. This substance typically comes in two different molecular forms: VD2 (ergocalciferol) and VD3 (cholecalciferol). In humans, VD2 potency is less than VD3, which can be synthesized in human

8

skin after exposure to sunlight [33]. Here we shall present results of encapsulating VD3 in vesicle bilayers formed by single gemini amino-acid surfactant. Considering that the main physiological function of VD3 is to promote the adsorption of intestinal mucosa to calcium and that the double carboxylic acid can chelate with bivalent Ca2+, we will show schemes to construct Ca2+-coatednano-VD3-capsules with the Ca2+ surrounding the outermost carboxylic acid ions. The resultant capsules can be taken as a combined calcium and VD3 cosupplementation agent [34,35]. 2. Experimental methods 2.1 Materials Gemini surfactants Ace(n)-2-Ace(n) and Ace(m)-2-Ace(n) were previously synthesized in our laboratory [25,26]. Pyrene (purity 98%), SDS and cetyl trimethyl ammonium bromide (1631, purity 98%) were purchased from Sigma-Aldrich. Other purchased chemicals include Coumarin 153 (C153), Rhodamine 6G (R6G), and Vitamin D3 (J&K Chemical Ltd., purity 98%) and CaCl2 (Sinopharm Chemical Reagent Co., Ltd, AR). Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), and phosphate buffered solution (PBS) were purchased from PAA Laboratories GmbH. The water used for preparing the solution in all experiments was generated from XYJ-250-H Pine-Tree DI water system, with resistivity of 18.25 ΩM·cm. 2.2 Measurements 2.2.1 Turbidity measurements The turbidity of the surfactant solution as a function of various pH or solvents was measured using a Lambda 750 S ultraviolet spectrophotometer. The wavelength for the measurement of optical density was selected to be 500 nm where there is no absorption of the surfactant component [36]. 2.2.2 Colloid property measurements

9

DLS and ζ-potential measurements were carried out with a Malvern Autosizer (ZETASIZER nano series Nano ZS-90, Malvern, UK) at 298 K. All static light scattering measurements were carried out at 90º scattering angle. The freshly prepared samples were dispersed with ultrasound for 45 min and stored for 12 hours before testing. The obtained scattering data were fitted using an intensity-weighted cumulative analysis to estimate the diffusion coefficient of the aggregates in aqueous solution. ζ-potential was measured using 600 μL of sample solution in a standard cuvette. The average hydrodynamic diameter, the poly-dispersion index (PDI), counter rate, and characteristics of the surface charge of the aggregates were investigated. 2.2.3 Fluorescence measurements Steady-state fluorescence spectra were recorded by HITACHI F-7000 Fluorescence Spectrophotometer. The surfactant containing fluorescence probe solutions were dispersed with ultrasound for 45 min and kept for 12 hours before testing. For pyrene-containing solution, the emission spectra wavelength ranged from 350 nm to 550 nm, the excitation wavelength was focused at 335 nm with slits being fixed at 2.5 nm and 2.5 nm respectively, and intensities of peaks labelled as I1 and I3 were taken from the emission intensities at 374 and 383 nm respectively [37]. Samples for FRET were prepared by dissolving 3 μM C153 and R6G in 1 mM Ace(n)-2-Ace(n) or Ace(m)-2-Ace(n) solutions. FRET measurements were done by scanning the wavelength from 420 to 725 nm while maintaining constant excitation wavelength at 405 nm with slits being fixed at 5 nm and 5 nm respectively. For double blank control the spectra of surfactant solutions containing 3 μM C153 or R6G were also prepared. 2.2.4 Preparation of Vitamin D3-loaded nanocapsules (nano-VD3-capsules) N,N’-dialkyl-N,N’-diacetate ethylenediamine was dissolved in excess of NaOH solution at the concentration of 2 mM. Then the pH was adjusted to be within the isoelectric point pH interval

10

of gemini surfactant, approximately weak alkaline for Ace(n)-2-Ace(n) and neural for Ace(m)-2Ace(n). VD3 solid, surfactant mother solution and isometric H2O were mixed and magnetic stirred for 12 hours, finally dispersed with ultrasound at 80 Hz for 45 min to obtain vesicular solutions. The sample solutions were stored at 298 K for 12 hours before further testing. 2.2.5 Transmission Electron Microscopy TEM micrographs were obtained with a JOEL JEM-2000EX system. Negative staining samples were prepared with phosphotungstic acid acetate solution (2%) as the staining agent. One drop of the solution was placed onto a carbon-coated copper TEM grid (300 mesh). Filter paper was employed to suck away the excess liquid. Then one drop of the staining agent was placed onto the copper grid. Excess liquid was also sucked away by filter paper. 2.2.6 Cellular internalization observation of C153-loaded vesicles The cellular internalization observation of C153-loaded vesicles formed by Ace(10)-2-Ace(16) were performed by confocal microscopy measurement. MCF-7 cell (human breast cancer cell) were seeded into the special dish for confocal image (Mattek, 35 mm/1 dish) at 2 × 105 cell per well in 2 mL medium (DMEM, pencillin, and 10% FBS) and cultured at 37 ºC in a humidified atmosphere containing 5% CO2 for 24 hours, followed by removing 1 mL culture medium and adding 1 mL 10-3 M aqueous Ace(10)-2-Ace(16) solution (pH=7.50) containing 15 μM C153. The cells were incubated for another 4 hours. Then, the culture medium was removed, and the cells were washed with PBS for three times [38]. Finally, the dish was observed with a Olympus Fv1000-IX81 confocal microscope. 3 Results and discussion 3.1 Characterization of vesicles in aqueous surfactant solutions

11

Turbidity measurement provides a good indication of vesicle formation and aggregation. To understand the extent of turbidity, the optical density of the solutions was measured at a fixed wavelength of 500 nm with decreasing pH. The results are displayed in Fig. S1 (in the Supporting Information). The optical density value of the solution was near 0 at strong alkaline pH, but gradually increased with the decrease of pH. With the aid of TEM, spherical vesicles were observed in the solution with strong bluish color. The results confirmed unambiguously that pH can effectively regulate the gemini surfactant’s ability to assemble into vesicles at specific pH range. Moreover, this pH-mediated structure change is reversible, which is demonstrated by pH-cycling turbidity and DLS measurements (Fig. S2 and S3). At acidic pH, solid precipitated out from the colloid system. It is remarkable that both pH-responsiveness and reversibility have been realized in this sole surfactant system. 3.2 Mechanism studies about pH-driven vesicle formation 3.2.1 Deduced mechanism from ζ-potential at varying pH Collision always inevitably happens during the aggregates’ Brownian motion in a colloidal solution. If the aggregates have high positive or negative zeta potentials, they tend to repel each other due to electrostatics; otherwise, they are ready to flocculate or aggregate [39]. Thus, ζpotential not only characterizes the surface charge distribution of aggregates, but also gives an indication of the colloidal system stability. Usually, aggregates with ζ-potential values either larger than +30 mV or smaller than -30 mV are considered to be stable. From Table S1 and S2, it can be seen that the ζ-potential decreases with the decrease of pH. This can be understood by considering the protonation between H+ and -N-CH2COO-: at strong alkaline pH, completely dissociated double -CH2COO- groups presented the surfactant as gemini anionic type and constituted vesicle’s outermost hydrophilic membrane, so the aggregates have

12

the smallest negative ζ-potential. As the H+ was gradually combined with nitrogen atom, the original anionic -N-CH2COO- converted into electro-neutral -NH+-CH2COO-, so the negative ζpotential became higher (described in Scheme 2). Because of the bilayer configuration that orients -CH2COO- groups toward water phase and hydrophobic long carbon chains inside, ζpotential of the colloidal system always shows a electronegativity smaller than -30 mV, reflecting the stable interaction between aggregates.

Scheme 2. Schematic mechanism about pH-regulation deduced from ζ-potential measurements

3.2.2 Confirmed mechanism based on DLS at varying solvents From the ζ-potential data, we hypothesized that the protonation between H+ and -N-CH2COO- is the driving force to promote the transition from micelles to vesicles within isoelectric point range in aqueous gemini amino-acid surfactant solutions. The protonation is thought to be adequately carried out in H2O with strong polarity and high dielectric constant, and thus can be adjusted by changing the property of solvents.

13

Ace(12)-2-Ace(12) and Ace(8)-2-Ace(12) were taken as the representative of symmetric and asymmetric gemini surfactant respectively, and dissolved in H2O and CH3OH mixtures. With the proportion of CH3OH increasing from 0 to 90%, gradual disappearance of bluish color was clearly seen (Fig. S4), which is consistent with decreasing turbidity (Fig. S5). It implied that the addition of CH3OH had destructed the vesicle structure to some extent. DLS measurement results of 10-4 M Ace(12)-2-Ace(12) and Ace(8)-2-Ace(12) at pKa1
14

3.3.1 Steady-stated fluorescence spectra of pyrene-surfactant solutions Large quantities of vesicles dramatically materialized by regulating pH from strong to weak alkaline in aqueous pyrene-surfactant solution, as shown qualitatively from the fluorescence intensity of pyrene-surfactant at different pH values. It is mainly because the vesicles dissolved more pyrene molecules in the hydrophobic bilayers that improved the solubility of pyrene in water. The polarity index I1/I3 describes the polarity degree of the micro-environment where pyrene locates. The I1/I3 value in strong polar H2O is high (up to 1.8) and decreases to near 0.8 with the help of surfactant aggregates. For instance, the I1/I3 in the strong hydrophobic center of the micelle is very low and shows more subtle distinctions when pyrene locates at different sites within the aggregates (e.g., aggregate core, palisade layer, headgroup region) [41]. As shown in Table S1 and S2, the polarity index I1/I3 of 10-3 M Ace(n)-2-Ace(n) and Ace(m)-2-Ace(n) decreases monotonically with decreasing pH, which indicates that the micro-environment around pyrene is becoming more and more non-polar. The differences in polarity index are ascribed to the disparity in the degree of water penetration of the vesicles depending on the variations in the compactness of the surfactant headgroups: less water penetration and more compactness in the headgroups result in smaller I1/I3 values [42]. The I1/I3 data reflect that pH directs the surfactant molecules to pack more closely when forming vesicle bilayers within the pKa range. This phenomenon can be attributed to the fact that the formation of -NH+-CH2-COO-: -NH+generates attraction between NH+ and -COO-, which compresses the spacial extent of CH2COO-. Furthermore, -NH+- can attract the surrounding nitrogen atom and the negative charge of -CH2COO- belonging to adjacent surfactants, which is similar to the interaction

15

between anionic and zwitterionic surfactants. The above two synergies can weaken the repulsion between adjacent CH2COO- groups and make the bilayers pack more tightly, which leads to lower I1/I3 value. 3.3.2 Steady-stated fluorescence spectra of C153 Besides pyrene, C153 can act as another kind of micro-environment-sensitive hydrophobic fluorescence probe to explore the formation process and properties of self-assemblies. The emission spectra of C153 suggest a strong correlation with solvent characteristics: different degrees of blue shift would happen at one excitation wavelength in different solvents; the less polar the solvent, the more prominent blue shift, as shown in Fig. 1a. Fig. 1b and 1c showed the fluorescence spectra of 3 μM C153 in 1 mM aqueous Ace(12)-2-Ace(12) and Ace(8)-2-Ace(12) solutions, respectively (the other seven surfactants can be seen in Fig. S6). As pH decreases, the gradual blue-shift of the emission maxima indicates the gradual incorporation of the probe molecules into the hydrophobic bilayer region of the vesicles. The emission maxima of C153 show a relatively small shift ≈ 25 nm at weak alkaline pH and a more dramatic shift ≈ 50 nm (corresponding to the blue-shift in ethyl acetate) at neutral or weak acidic pH conditions toward the blue region. The more rigid and hydrophobic micro-environment surrounding the probe molecules results in the significant spectral changes.

16

Fig. 1. The fluorescence spectra of 3 μM C153 in: a) different solvents; b) 1 mM Ace(12)-2Ace(12) solution; c) 1 mM Ace(8)-2-Ace(12) solution; d) Ace(12)-2-Ace(12) at pH = 11.90; e) Ace(8)-2-Ace(10) at pH = 11.70 Besides the blue-shift phenomenon, the spectra of C153 excited at 375 nm in surfactant solutions at strong alkaline pH is found to be obviously broader, which can be interpreted as a superposition of two peaks at ≈ 525 nm and ≈ 480 nm, respectively. This double-peak shoulder is particularly obvious in Ace(12)-2-Ace(12) and Ace(8)-2-Ace(10) (as shown in Fig. 1d and 1e). And the shoulder shape disappeared with the excitation wavelength increasing from 375 to 435 nm. The broadness of the emission spectra and the λex-dependence suggest that C153 molecules are located at varied environments [43], which is probably caused by the coexistence of dissociative surfactant ions, micelles, and vesicles in the solution at strong alkaline pH. 3.3.3 pH-regulated FRET from C153 to R6G To further probe the distribution of model hydrophobic and hydrophilic molecules in the vesicles, fluorescence resonance energy transfer (FRET) measurements were conducted using the donorreceptor fluophore pairs C153 and R6G incorporated in the surfactant bilayers. According to

17

Förster model, the rate of FRET is inversely proportional to the sixth power of distance (R DA) between the donor (D) and the acceptor (A) [44]. Thus only when the pair of donor and acceptor is close enough will FRET effectively happen, as shown in Fig. 2a. Hydrophilic donor C153 and hydrophobic acceptor R6G is a classical pair of donor and acceptor catering to the FRET principle - if located in appropriate environment, the emission wavelength of C153 could exactly excite R6G; as a result, the emission of R6G would show up at the excitation wavelength of C153, as shown in the experimental data in Fig. 2c. In theory, FRET can be realized in gemini amino-acid surfactants via the following mechanism. While the vesicles formed at near neutral pH can encapsulate C153 in the hydrophobic bilayer, the anionic -CH2COO- groups on both sides of the bilayer can attract cationic R6G close to the membrane surface (as shown in Fig. 2b), which enables FRET to happen from C153 to R6G. FRET will be enhanced by increasing the number of vesicles with more tightly bound bilayers. The fluorescence spectra of C153, R6G, and C153+R6G mixtures in Ace(10)-2-Ace(10) and Ace(8)-2-Ace(12) were displayed in Fig. 2d and 2e, respectively (FRET data for other seven surfactants can be seen in Fig. S7-13). At strong alkaline pH, the spectra basically represent the spectral signature of C153. As pH was decreased to weak alkaline, vesicles started to form, and the intensity of C153 increased due to the improved solubility within the bilayer. The spectra show a prominent shoulder from the combination of C153 (≈ 500 nm) and R6G (≈ 550 nm). At neutral and weak acidic pH, a large quantity of assembled vesicles dissolved more C153 and greatly promoted FRET: the spectra of C153 further blue-shifted and the intensity of R6G further increased, so the emission maxima of C153 and R6G became more and more apparent with the decrease of pH.

18

All the above fluorescence spectra fully demonstrate the effect of pH on the vesicle bilayer. The generated -NH+-CH2COO- tuned by pH regulates the surfactant transformation from bivalent anionic to monovalent anionic to zwitterionic. As different types of surfactants interact each other in the pH-responsive gemini surfactant solution and pH decreases from pKa1 to pKa2, vesicles self-assemble and the bilayers become more and more tightly bound.

Fig. 2. a) The mechanism of FRET; b) FRET occurring in vesicles; the fluorescence spectra of 3 μM C153/R6G in: c) different solvents, d) 1 mM Ace(10)-2-Ace(10) solution and e) 1 mM Ace(8)-2-Ace(12) solution 3.4 Application of pH-regulated gemini surfactant vesicles as nanocapsules 3.4.1 Property of nano-VD3-capsules A method has been developed to use pH-regulated gemini surfactant vesicles as nanocapsules for VD3. The recommended daily intake (RDI) of VD3 is 5 μg/day [45] and dosage of VD3 for

19

additional supplement is controlled approximately 3 μg/day to avoid hypervitaminosis. VD3 is stable in alkaline condition but sensitive to acid. To protect VD3 from being inactivated by H+, a 2 mM surfactant solution with neutral pH had been carefully prepared before adding isometric H2O and pre-calculated amount of VD3 solid according to recommended dosage. The assembled vesicles in the concentrated surfactant solution suffered disintegration and reformation under the dilution effect of isometric H2O. During the vesicle reforming process, VD3 can be integrated within the bilayer. The commercial anionic SDS and cationic 1631 were respectively prepared as surfactant/VD3 solutions following the same procedures for comparison.

Fig. 3. TEM photographs of nano-VD3-capsules formed by Gemini surfactants (the bar in the photograph represents 100 nm) VD3 loading capacity in 1 mM gemini surfactant solutions is 40 μg/mL, namely a 1 mL nanoVD3-capsule stock solution can meet 8 days’ RDI of VD3 for human body. The DLS results in Table S5 reveal the favorable dispersibility and strong light scattering intensity of nano-VD3capsule. Except for Ace(8)-2-Ace(8), Ace(8)-2-Ace(12), and Ace(12)-2-Ace(14), the particle

20

size in other six Gemini surfactant solutions can be controlled to be less than 200 nm. Even after been diluted 10 & 20 times (see in Table S6), the solution is still stable enough to sustain its colloidal characteristics. TEM was employed to confirm the morphology of nano-VD3-capsules (Fig. 3). It is worth mentioning that the capsules were found to be stable without any significant change in size, PDI and counter rate after being stored at 298 K for 60 days & 120 days. PDI of VD3/1631 sample is higher than that of nano-VD3-capsule. After 90 days, VD3/1631 solution appeared yellowish, possibly due to VD3 being oxidized by air. In the aqueous SDS solution, VD3 can hardly be dissolved and no further measurements were conducted. 3.4.2 Property of Ca2+-coated-nano-VD3-capsules Development was carried out to further enchance the functionality of VD3 loaded nano-capsules by coating them with Ca2+, a widely taken important nutritional mineral ion. Although the theoretical highest molar ratio α (defined as Ca2+: gemini surfactant) is 1:1, the higher ratio may lower the solubility of surfactant in water. Before making the Ca2+-coated-nano-VD3-capsules, it was necessary to pre-prepare Ca2+-coated-vesicles (“cal-vesicles” for short) to explore the most appropriate α. Cal-vesicles were prepared at four different α values (0.25, 0.5, 0.75, and 1:1). The solution appearance and DLS results are displayed in Fig. S14 and Table S7. It can be seen that particle size decreases with α increasing from 0 to 0.75, suggesting that Ca2+ has condensing effect on the vesicles. At α = 1, the size increase could be caused by the poor solubility when Ca2+ matched up with all the -CH2COO-. And the size of cal-vesicles at neutral pH is generally larger than weak alkaline pH, implying that the Ca2+-coated-nano-VD3-capsule may be more stable at weak alkaline pH. During the preparation of Ca2+-coated-nano-VD3-capsules, VD3 should be preferentially integrated into the bilayer before the particles are coated with Ca2+ (as shown in Fig. 4a). Based

21

on data in Table S7, α should be selected as high as possible on the promise of guaranteeing the stability of colloidal system. DLS measurement results were summarized in Table S8. Furthermore, Ace(12)-2-Ace(14) samples were taken as a representative to measure TEM to ensure the formation of capsules (see Fig. 4b). Except for several sample conditions, Ca2+coated-nano-VD3-capsules can be prepared with size < 200 nm and PDI < 0.5 . The impetus is to achieve a small particle size and uniform size distribution to enhance the capsules’ ability to cross the cell membrane.

Fig. 4. a) Schematic formation process of nano-Ca2+-capsule; b) TEM photographs of Ca2+coated-nano-VD3-capsule formed by Ace(12)-2-Ace(14) It is found that the size of capsules formed by asymmetric Ace(m)-2-Ace(n) is apparently smaller than symmetric Ace(n)-2-Ace(n). Even after being stored for 30 days at 298 K, the capsules were still stable in six Ace(m)-2-Ace(n)s while precipitated in three Ace(n)-2-Ace(n)s. This is possibly due to the fact that the asymmetry impels double alkyl chains to inlay each other during bilayer formation. The inlay figuration makes the vesicles more stable.

22

3.5 Cell internalization studies It is an important factor for nanocapsules that they can be effectively absorbed into cells. To study cell internalization of the nanocapsules, C153-loaded vesicles (1 mM Ace(10)-2-Ace(16) with 15 μM C153) were prepared and observed under confocal microscope (in Fig. 5a); then added into the culture medium, where MCF-7 cells were incubated at 310 K for 4 hours. Cell internalization was evaluated by both inversion fluorescence microscope and fluorescence confocal microscope. With the aid of fluorescence from C153, the photogenic cells can be observed under fluorescence microscope (Fig. 5b and 5c). The dyed cytoplasm can be more clearly visualized under confocal fluorescence microscope (Fig. 5d). These results confirm the effective internalization of nanocapsules formed by this series of gemini surfactants.

Fig. 5. Fluorescence images: a) C153-loaded vesicles formed by Ace(10)-2-Ace(16) under confocal microscope; MCF-7 cells cultured by C153-loaded vesicles for 4 hours under b) bright field and c) dark field by normal fluorescence microscope, d) confocal microscope; e) cytotoxicity of Ace(n)-2-Ace(n) & Ace(m)-2-Ace(n) against MCF-7 cells

3.6 In vitro cytotoxicity of vesicles

23

The biocompatibility or cytotoxicity is a key index for nanocapsule’s biomedicine application. The cytotoxicity of Ace(n)-2-Ace(n) and Ace(m)-2-Ace(n) vesicles was evaluated by MTT method against MCF-7 cells. Fig. 5e shows the cell viability after 24 hours incubation with 10 μM surfactant solutions at pH≈7.50. In addition, surfactant assemblies with 10 μM 1631 was also investigated for comparison. The cytotoxicity of N,N’-dialkyl-N,N’-di(sodium acetate) ethylenediamine is lowest in Ace(8)-2-Ace(8) (88%) and highest in Ace(12)-2-Ace(12) (32%), both of which are considerably lower than that in 1631 (15%). Although it has been demonstrated so far that the nanocapsule formation pH range in the surfactant solution is consistent with the pH of human physiology, it remains to be seen whether the nanocapsules can stably exist in the human circulations system. If the nano-VD3-capsules or Ca2+-coated-nano-VD3-capsules are aimed to be prepared for oral consumption, additional antiacidic encapsulation for protecting vesicles from gastric acid is definitely needed to help the nano-capsules travel through stomach and reach the intestines where the pH is more “friendly”. Among the new stimuli methods being explored to tune the molecular amphiphilicity, pH stimuli is one of the most important methods, due to the different pH conditions in normal organs and tumors: the pH value drops markedly from that in the normal extracellular physiological environment (pH 7.4) to pH 6-5, and to around pH 4-5 in primary and secondary lysosomes. This provides great opportunities for us to develop pH-responsive systems for drug delivery [42]. The nanocapsules formed by Ace(n)-2-Ace(n) & Ace(m)-2-Ace(n) are stable at near neutral pH, but easily suffer disaggregation at acidic pH. Thus, this nano-delivery system holds promise for steadily participating human physiological system and releasing drug when meeting particular cells located at specific acidic pH tissue. Further work will include: protective encapsulation around outer bilayers for protecting the nanocapsules from potential damage in hemodynamic

24

environment, stability in the blood circulatory system and release profile of the encapsulated substance. Moreover, if the nanocapsules are aimed to be prepared as oral agents, anti-acid protection against the gastric acid is absolutely necessary. 4. Conclusion Due to the protonation between H+ and -N-CH2COO- groups, reversible transformation among bivalent anionic, monovalent anionic, or zwitterionic surfactant assemblies were realized in N,N’-dialkyl-N,N’-di(sodium acetate) ethylenediamine solutions by adjusting pH values. The pH-regulated self-assembled structures-micelles and vesicles with various specific sizes were fabricated based on the interaction between -N-CH2COO- and -NH+-CH2COO-. This is a new route for creating pH-regulated nanostructures. It is interesting to find that surfactant molecules arrange more tightly during constituting process of the vesicles’ bilayers either within pH range of isoelectric point or in asymmetric Ace(m)-2-Ace(n) solutions. The characteristic that pH range of vesicles formation is near human physiological environment is utilized to encapsulate lipidsoluble VD3 in hydrophobic vesicle bilayers as nano-VD3-capsules to prepare VD3 tablet or liquid without oil or fat. Compared with the macroscopic insolubility VD3 in SDS, nano-VD3capsules formed by Ace(n)-2-Ace(n) or Ace(m)-2-Ace(n) have better stability; the size of Ca2+coated-nano-VD3-capsules can be basically controlled within 200 nm. The fluorescence microscope measurements indicate that these nanocapsules can be effectively internalized by MCF-7 cells to facilitate an efficient nano-delivery across the cell membrane. The results of MTT against MCF-7 cell confirmed the better biocompatibility of Ace(n)-2-Ace(n) or Ace(m)2-Ace(n) than 1631. In conclusion, we have demonstrated that this new type of pH-regulated gemini amino-acid surfactants holds great promise for formulating nano-carriers to deliver drugs and nutritional supplements.

25

Acknowledgments This work was financially supported by International Science & Technology Cooperation Program of China No.2015DFA41670. Dalian University of Technology Chemistry Analysis & Research Center is acknowledged for providing measurement instruments. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://?. Corresponding Author *Weihong Qiao, e-mail: [email protected]; fax: +86-0411-84986232.

26

References [1] R. Langer, D. A. Tirrell, Nature 428 (2004) 487. [2] C. Wang, Q. Chen, H. Xu, Z. Wang, X. Zhang, Adv. Mater. 22 (2010) 2553 [3] Y. Chen, B. Zhu, F. Zhang, Y. Han, Z. Bo, Angew. Chem. Int. Ed. 47 (2008) 6015. [4] W. Wang, D. Cheng, F. Gong, X. Miao, X. Shuai, Adv. Mater. 24 (2012) 115. [5] L. Xu, L. Feng, R. Dong, J. Hao, S. Dong, Biomacromolecules 14 (2013) 2781. [6] X. Huang, M. Cao, J. Wang, Y. Wang, J. Phys. Chem. B 110 (2006) 19479. [7] J. Huang, G. Zhao, Colloid Polym. Sci. 273 (1995) 156. [8] M. Miyahara, H. Kawasaki, V. M. Garamus, N. Nemoto, R. Kakehashi, S. Tanaka, M. Annaka, H. Maeda, Colloids and Surfaces B: Biointerfaces 38 (2004) 131. [9] K. Takedaa, H. Okunoa, T. Hatab, M. Nishimotoa, H. Matsukia, S. Kaneshina, Colloids and Surfaces B: Biointerfaces 72 (2009) 135. [10] L. Jiang, K. Wang, M. Deng, Y.Wang, J. Huang, Langmuir 24 (2008) 4600. [11] R. O. Brito, E. F. Marques, S. G. Silva, M. L. Vale, P. Gomes, M. J. Araújo, J. E. Rodriguez-Borges, M. R. Infante, M. T. Garcia, I. Ribosa, M. P. Vinardell, M. Mitjans, Colloids and Surfaces B: Biointerfaces 72 (2009) 80. [12] C. Aiello, P. Andreozzi, C. L. Mesa, G. Risuleo, Colloids and Surfaces B: Biointerfaces 78 (2010) 149. [13] S. Song, Q. Zheng, A. Song, J. Hao, Langmuir 28 (2012) 219. [14] C. Zhou, X. Cheng, Y. Yan, J. Wang, J. Huang, Langmuir 30 (2014) 3381. [15] J. Lee, C. Chang, Colloids and Surfaces B: Biointerfaces 121 (2014) 171. [16] Y. Zhang, Y. Feng, Y. Wang, X. Li, Langmuir 29 (2013) 4187. [17] Y. Wang, N. Ma, Z. Wang, X. Zhang, Angew. Chem. Int. Ed. 46 (2007) 2823.

27

[18] A. Harada, Accounts Chem. Res. 34 (2001) 456. [19] I.A. Banerjee, L. Yu, H. Matsui, J. Am. Chem. Soc. 125 (2003) 9542. [20] Y. Lin, X. Han, X. Cheng, J. Huang, D. Liang, C. Yu, Langmuir 24 (2008) 13918. [21] T. Patra, S. Ghosh, J. Dey, Soft Matter 6 (2010) 3669. [22] W. Wang, W. Lu, L. Jiang, J. Phys. Chem. B 112 (2008) 1409. [23] H. Lu, L. Wang, Z. Huang, RSC Adv. 4 (2014) 51519. [24] K. Sakai, M. Kaji, Y. Takamatsu, K. Tsuchiya, K. Torigoe, K. Tsubone, T. Yoshimura, K. Esumi, H. Sakai, M. Abe, Colloids and Surfaces A: Physicochem. Eng. Aspects 333 (2009) 26. [25] J. Lv, W. Qiao, C. Xiong, Langmuir 30 (2014) 8258. [26] J. Lv, W. Qiao, Soft Matter 11 (2015) 2577. [27] K. Ziani, Y. Fang, D.J. McClements, Food Chem. 134 (2012) 1106. [28] C. Solans, I. Solé, Curr. Opin. Colloid Interface Sci. 17 (2012) 246. [29] R. Deutch-Kolevzon, A. Aserin, N. Garti, Chem. Phys. Lipids 164 (2011) 654. [30] S. Xia, S. Xu, J. Sci. Food Agric. 86 (2006) 2119. [31] A. Cranney, H. A. Weiler, S. O’Donnell, L. Puil, Am. J. Clin. Nutr. 88(suppl) (2008), 513S. [32] M. Guttoff, A.H. Saberi, D.J. McClements, Food Chem. 171 (2015) 117. [33] L.A.G. Armas, B.W. Hollis, R.P. Heaney, J. Clin. Endocrinol. Metab. 89 (2004) 5387. [34] C. Meier, M. Kränzlin, Swiss Med. Wkly. 141 (2011) w13260. [35] O. Rey, S.H. Young, J. Yuan, L. Slice, E. Rozengurt, J. Biol. Chem. 280 (2005) 22875. [36] S. Mandal, J. Kuchlyan, S. Ghosh, C. Banerjee, N. Kundu, D. Banik, N. Sarkar, J. Phys. Chem. B 118 (2014) 5913. [37] X. Wang, J. Liu, L. Yu, J. Jiao, R. Wang, L. Sun, J. Colloid Interface Sci. 391 (2013) 103.

28

[38] L. Meng, W. Huang, D. Wang, X. Huang, X. Zhu, D. Yan, Biomacromolecules 14 (2013) 2601. [39] Y. Jiang, F. Li, Y. Luan, W. Cao, X. Ji, L. Zhao, L. Zhang, Z. Li, Int. J. Pharm. 436 (2012) 806. [40] S.C.D. Santos, L.M. Moreira, D.L.R. Novo, L.R.R. Santin, D. Bianchini, J.A. Bonacin, A.P. Romani, A.U. Fernandes, M.S. Baptista, H.P.M. de Oliveira, J. Porphyrins Phthalocyanines 19 (2015) 920. [41] J. Kim, M.M. Domach, R.D. Tilton, J. Phys. Chem. B 103 (1999) 10582. [42] D. Mitra, I. Chakraborty, S.C. Bhattacharya, S.P. Moulik, S. Roy, D. Das, P.K. Das, J. Phys. Chem. B 110 (2006) 11314. [43] U. Mandal, S. Ghosh, D. K. DAS, A. Adhikari, S. Dey, K. Bhattacharyya, J. Chem. Sci. 120 (2008) 15. [44] S. K. Mondal, S. Ghosh, K. Sahu, U. Mandal, K. Bhattacharyya, J. Chem. Phys. 125 (2006) 224710. [45] K. Rajakumar, MD, Pediatrics 112 (2003) e132. [46] Y. Wang, H. Xu, X. Zhang, Adv. Mater. 21 (2009) 2849.

29