Stearylated cycloarginine nanosystems for intracellular delivery – simulations, formulation and proof of concept

Abstract

Cationization of nanocarriers has been sought after as a viable strategy to surmount cellular barriers that impede intracellular drug and gene delivery. A plethora of cationic compounds including both lipids and polymers have been designed and developed for efficient cellular penetration. The toxicity of these cationic agents, however, precludes their use in drug delivery. Conjugation with biomolecules like sugars and amino acids to produce novel cationic agents is being investigated to generate biocompatible molecules to achieve safe drug delivery. In this study, we propose to mitigate the toxicity of stearylamine, a cationic lipid, by tethering it to arginine, to yield a novel cationic ligand. Ligand loaded liposomes and nanoparticles were fabricated and evaluated for their surface charge and cellular uptake. Furthermore, molecular dynamics simulations were utilized as a tool to understand the accessibility of the novel ligand and stearylamine loaded liposomal systems. This paper presents the one pot synthesis of a novel stearylated arginine dipeptide and its incorporation in delivery systems along with its in vitro and in vivo toxicity evaluation.

---

1. Introduction

Cationic nanosystems have been widely and successfully explored for intracellular delivery of both drugs¹ and genes.² Tumor cells express an excess of sulfated proteoglycans as a part of their cell membrane. This renders the surface negatively charged.^3,4 Cationization of macromolecules and particulate drug carriers enhances their cellular uptake as moieties bearing a positive charge are electrostatically attracted and adsorbed to negatively charged proteoglycans, thus triggering endocytosis. There is abundant evidence in literature regarding the application of cationic systems to overcome biological barriers and targeting drugs to body organs, including the brain, eyes, nose, skin and inflamed intestinal epithelium as well as to tumor vasculature and in cancer therapy.^5–17 A myriad of cationic compounds viz. cationic polymers like chitosan and polyethylenimine (PEI) and lipids such 2,3-dioleyloxy-N-[2-(sperminecarboxamido)ethyl]-N,N-dimethyl-1-propanaminium trifluoroacetate (DOSPA) and N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl-sulfate (DOTAP) have been incorporated in drug delivery systems.¹⁸ Toxicity of these cationic agents however, precludes their use in drug delivery.

Stearylamine is a well established cationic agent used in drug delivery systems. It is, however, toxic; it exhibits LD₅₀ values of 2395 mg kg⁻¹ and 250 mg kg⁻¹ in rats and mice respectively when taken via the oral route. It is considered to be both, a poison when administered via the intraperitoneal route and a skin irritant.¹⁹ Stearylamine liposomes are reported to be haemotoxic, the toxicity ascribed to the charge on the stearylamine.²⁰ Yoshihara E. et al. have reported the cytolytic activity of liposomes containing stearylamine on rabbit erythrocytes. They report that liposomes with less than 5% (w/w) of stearylamine are devoid of haemolytic activity. In contrast, liposomes containing 7.5–12.5% of stearylamine show an abrupt increase in their haemolytic activity. Liposomes with a critical density of stearylamine possess haemolytic activity.²⁰ Adams D. H. et al. have shown that stearylamine containing liposomes are toxic to the mouse brain²¹ when injected intracerebrally; toxicity has been attributed to the stearylamine component of the liposomes. The toxic effect is mainly determined by the cationic nature of the agent, a property resulting from the structure of the hydrophilic head group. Quaternary amines are more toxic than their tertiary amine counterparts.²² Efforts are now being focused towards the design of newer cationic agents with lower toxicity and enhanced efficacy through the alteration of their molecular weight, nature and/or the size of their head group, thereby altering the charge and improving their biodegradability.²²

The conjugation of cationic agents with non-toxic molecules such as amino acids and sugars is thus currently being explored as a promising approach to produce biocompatible agents that can be incorporated into liposomal bilayers.^23,24 Arginine (AG), a naturally occurring cationic amino acid present in the physiological environment, has emerged as an immensely useful molecule for intracellular delivery. Studies have demonstrated that arginine content can play a pivotal role in cellular uptake of Tat (trans-activating transcriptional activator).²⁵ Short oligomers of arginine exhibit better cellular penetration in comparison to those of lysine, histidine and ornithine.²⁶ Further, incorporation of stearyl group in arginine peptides has resulted in a doubling of the transfection activity in comparison to the octaarginine peptide, attributed to hydrophobic interactions of a stearyl chain with the cell membrane.²⁷ Stearylated octaarginine has been shown to exhibit toxicity comparable to that of Lipofectamine®, a commercial transfection agent.^28,29 Recently, arginine rich cell penetrating peptides (CPPs) have been the focus of research to deliver cargo molecules like drugs and genes intracellularly as they traverse the membrane easily with little toxicity.^30,31 Furthermore, cyclization of the guanidine moiety of arginine in CPPs increases it transduction activity across the cell membrane.³² In this study, we propose to amalgamate the advantages of low toxicity and the efficient membrane penetrating ability of CPPs with stearylamine. This was accomplished through tethering the arginine to a stearylamine group and fabricating novel ligand targeted liposomes for intracellular delivery of encapsulated drugs. Recently all atom molecular dynamics has been demonstrated to be a very useful tool for the design of liposome based drug delivery.^33–35 It provides vital mechanistic understanding which is not available experimentally regarding the incorporation of novel cationic agents, through providing a picture with all atom resolution of the liposome structure assembly and its surface properties. Here, we observed the orientation and exposure to the solvent of a novel ligand embedded in the liposome and the results were compared to the previously reported SA and arginine stearylamine ligands.³⁶ Stearylamine (SA) liposomes, liposomes containing stearylamine mono arginine conjugate (SAAG) and stearylated arginine cyclic dipeptide (SACD) were evaluated for their surface charge and cellular uptake. This paper presents one pot synthesis of a novel stearylated arginine dipeptide, its incorporation in delivery systems along with its in vitro and in vivo toxicity evaluation.

2. Materials and methods

Stearylamine (Sigma Aldrich, 99%, India), L-arginine (Himedia, 99%, India), 1,1′-carbonyldiimidazole (CDI) (98%, Himedia, India), chloroform ExcelaR (Qualigens, India), Tetrahydrofuran AR (SD Fine Chem. Ltd, India). 20 × 20 cm glass slides, TLC chamber, silica gel G for thin layer chromatography (Qualigens, India), iodine (Loba Chemie, 99.5%, India). Hydrochloric acid 5 M was freshly prepared from concentrated hydrochloric acid ExcelaR (Qualigens, India), double distilled water was prepared and used when required. Phospholipon 90 (PL 90, soya phosphatidylcholine) and Phospholipon 90 H (PL 90H, hydrogenated soya phosphatidylcholine) were obtained as gift samples from Lipoid® GmbH Germany. Cholesterol (95% pure) was obtained from Fischer Scientific. Tween 80 AR was obtained from S. D. Fine Chem. Ltd. Stearylamine (99% pure) was obtained from Sigma Aldrich.

2.1 Molecular dynamic simulations

Molecular Dynamics (MD) simulations were carried out on three solvated lipid bilayer systems. In each of the three systems, the lipid bilayers were composed of 186 distearoylphosphatidylcholine (DSPC) and 90 cholesterol (CHOL) molecules (ratio ∼ 2 : 1). In each of the three systems, 12 SA molecules were inserted into the membrane, six in each leaf. In one system the SA molecules were functionalized with our new novel targeting ligand (SACD). For comparison, 2 other systems were built; one with SA molecules conjugated with arginine (SAAG) and a second with SA molecules alone. 20 036 water molecules were added in all the three simulation systems, to solvate the membrane and achieve full hydration.³⁷ Physiological salt concentration of NaCl was added to all systems and additional counter ions were added to neutralize the net charge of the system. The parameterization of all molecules and ions was carried out using the OPLS-AA force field,³⁸ with extensions that allow for a compatible lipid model³⁹ water molecules were modelled using the TIP3P model⁴⁰ that is compatible with the OPLS-AA (Optimized Potential for Liquid Simulations – All Atom) force field. All simulations were carried out using the GROMACS 4.6.5 simulation package.⁴¹ In all directions periodic boundary conditions with minimum image convention were used. All simulations were carried out at physiological pressure and temperature (300 K and 1 bar). Constant temperature conditions were maintained using the Nosé–Hoover thermostat^42,43 with the membrane and solvent coupled to separate heat baths. The isobaric conditions were maintained using a semi-isotropic Parinello–Rahman barostat.⁴⁴ For both thermostat and barostat, the time constant was set to 1 fs. The length of all covalent bonds to hydrogen atoms was held fixed using the LINCS algorithm.⁴⁵ The time step for all simulations was set to 2 fs. The Particle Mesh Ewald method (PME) was used for the long range electrostatic interactions, with a real space cut-off of 1 nm.⁴⁶ Visual Molecular Dynamics (VMD) was used for visualization of the trajectories.⁴⁷ For all targeting ligands, partial charges were derived in accordance with the OPLS-AA methodology by fitting to the electrostatic potential using the RESP procedure implemented in ANTECHAMBER.^48,49 All systems were first equilibrated using the following procedure: (1) the steepest descent algorithm was used to minimize the energy of the initial configurations, then (2) 10 ns of molecular dynamics simulation under normal pressure temperature (NPT) conditions was run with the lipid tails constrained to stay in the bilayer center and the SA head group positions restrained not to enter the lipid core of the bilayer. Once this was completed (3) all systems were then run for 150 ns without restraint, during which time the area per lipid in all systems reached a stable value. Now that all systems had reached equilibrium concerning the properties that we measured, a 150 ns production run was performed for all systems and this trajectory was used for all analysis. Analysis of the simulation data was performed using tools implemented in Gromacs software and in house developed tcl scripts in VMD.⁴⁰ The solvent accessible surface area is calculated as described in the previous work by Lehtinen et al.³⁴

2.2 Synthesis of SACD

To a stirred solution of arginine (0.57 mmol) in 0.2 mL water, 1,1′-carbonyldiimidazole (CDI) was rapidly added (0.57 mmol) and arginine was activated for 30 minutes. SA (0.57 mmol) dissolved in tetrahydrofuran was added drop wise to the amino acid solution. The mixture was heated to 50 °C with drop wise addition of double distilled water to obtain clear solution. The reaction mixture was refluxed for a period of 72 hours at 50 °C while monitoring the reaction progress using TLC. The reaction kinetics was monitored by taking aliquots at 24 hours, 48 hours and 72 hours and analyzing by TLC and mass spectroscopy. After completion of reaction at 72 hours, solvent was evaporated under vacuum at a temperature not exceeding 40 °C and the residue obtained was washed with double distilled water, filtered under vacuum conditions, and air dried. The crude product obtained was further purified by preparative thin layer chromatography. The plates were developed by ascending technique using ethyl acetate : formic acid : glacial acetic acid : water in the ratio 40 : 2 : 2 : 1 as the mobile phase in a pre saturated chamber. Each plate was run twice in the same solvent system and dried before visualization by iodine. Plates were air dried overnight and silica was scraped off from the plate to yield two fractions: a mixture of SA and SAAG and pure SACD. The products were isolated by chloroform extraction and characterized by FT-IR, ¹H NMR and mass spectroscopy.

2.3 Fabrication of cationic nanosystems of cationic nanoparticles

2.3.1 Cationic nanoparticles. SA and SACD (5 mg) dissolved in chloroform was added to a 0.1% w/v solution of Tween 80 with cyclomixing to form an emulsion. The emulsion was subjected to size reduction using a probe sonicator at 30 W for 4 minutes. Chloroform was evaporated under vacuum at a temperature not exceeding 40 °C from the emulsion resulting in precipitation of cationic lipid in the form of nanoparticles. The nanoparticles were characterized for particle size and zeta potential. All the batches were taken in triplicate and data expressed as mean size ± std. dev.

2.3.2 Cationic liposomes. Liposomes were prepared by lipid film hydration method. Briefly, phospholipid and cholesterol were dissolved in chloroform and chloroform was evaporated under vacuum on a rotary evaporator (Buchi, Germany) at a temperature not exceeding 40 °C in a water bath to yield a uniform thin lipid film on the surface of the flask. The film was hydrated with 5 mL MilliQ water at 60 °C and the flask was maintained at 60 °C to anneal the liposomes to yield a 20 mg mL⁻¹ lipid concentration. The liposome dispersion was subjected to probe sonication (Branson Sonifier 250, Danbury) at 30 watts for 4 minutes and stored at 4 °C till further use. Liposomes containing SA and SACD were prepared in the same manner as the conventional liposomes wherein cationic agents were incorporated into the chloroform solution of phospholipids. All cationic lipids were incorporated at a concentration of 1.5 mol% and 5 mol% in liposomes.

2.4 Characterization of cationic formulations

The blank as well as cationic nanoparticles and liposomes prepared were characterized as follows.

2.4.1 Particle size determination. The particle size of the liposomes was determined using the particle size counter Malvern ZetaSizer after suitable dilution with MilliQ water to obtain particle counts in the range specified by the instrument software (PCS control Software Version 3.02.) at 25 °C with a 25 mW helium neon laser (wavelength: 632.8 nm) incident on the sample at an angle of 90. Particle size was also determined using TEM technique. Briefly an aliquot of 5 μL of the dispersion was mixed with uranyl acetate (1%) and placed on the TEM grids for 4 minutes. Excess of liquid was removed with a lint-free filter paper and the grid was air dried. TEM grids were investigated using a Philips CM 200 TEM microscope operated at 80 kV.

2.4.2 Zeta potential. The zeta potential of the system was measured on a ZetaSizer Nano ZS Malvern instruments at a temperature of 25 °C with a 4 mW 25 mW helium neon laser (wavelength: 632.8 nm).

2.4.3 In vitro haemolysis. In vitro haemolysis was performed as reported for SA liposomes with slight modifications in the reported procedure.^50,51 Rat erythrocytes were isolated by collecting their blood in heparinized tubes. The tubes were centrifuged at 4000 rpm for 10 minutes and the plasma was discarded. Pellet was washed thrice with two times the quantity of ice cold phosphate buffered saline (PBS). The tubes were centrifuged at 4000 rpm for 10 minutes and the supernatant was discarded. The final pellet in the tube contained the unlysed rat erythrocytes. The cells were resuspended in the PBS to obtain a concentration of 20% v/v. From this stock, 5% cell suspension was prepared in the PBS and the cells were incubated with various concentrations of nanoparticulate formulations at 37 °C for one hour. Incubation with PBS served as a negative control, whereas complete lysis was obtained with 1% Triton X which served as a positive control. After incubation, the tubes were centrifuged at 4000 rpm for 10 minutes. Haemolysis was assessed by measuring the absorbance of supernatant at 540 nm. The results were expressed in terms of % hemolysis which was calculated through the following formula.
where, A_s is the absorbance of the sample, A₁ is the absorbance of the negative control and A₂ is the absorbance of the positive control.

2.4.4 In vitro cell uptake. In vitro cell uptake was evaluated using two cell lines LN229 (glioblastoma cell line) and HaCaT (Human keratinocyte cell line). Cells were cultured in 6 well plates at a density of 2 × 10⁶ cells per well for 24 hours and incubated at 37 °C, 5% CO₂, 95% air and 100% relative humidity for 24 hours prior to addition of various liposomal formulations for adhesion. Cells were then incubated with various concentrations of SA and SACD liposomes for 4 hours. Liposomes were loaded with daunorubicin, an auto-fluorescent molecule at a concentration of 1.6 mg mL⁻¹ as optimized from previous experiments⁵² for the determination of cell uptake. After 4 hours, the cells were washed with DMEM (Dulbecco's Modified Eagle's Medium) and imaged using a confocal microscope (Ziess LSM 780, Carl Ziess Germany) and the images were captured and analyzed using ZEN software. Cells were trypsinized and suspended in phosphate buffer and acquired on flow cytometer (FACS Calibur BD) to quantify the red fluorescence (FL2) and the results were analyzed using CellQuestPro software.

2.4.5 Acute toxicity study of SACD liposomes. The acute toxicity of SACD liposomes in female mice was determined as per the OECD Guidelines (No. 423), once the protocol had been approved by the Institutional Animal Ethics Committee. A sterile batch of liposomes was formulated, as described earlier. The liposomal formulation ≅ 0.8543 mg kg⁻¹ SACD (calculated on the basis of previous experiments) and vehicle saline was administered in a single dose by intravenous route via the tail vein. Animals were observed individually after dosing at least once during the first 30 minutes, periodically during the first 24 hours, with special attention given during the first 4 hours and daily thereafter, for a total of 14 days. Observations included changes in skin and fur, eyes and mucous membranes, and also respiratory, circulatory, autonomic and central nervous systems, and somatomotor activity and behaviour pattern. Body weights of animals were determined shortly before the administration of the formulation and weekly thereafter. Food consumption was determined every day by considering food given and food left in the group. On the 15^th day animals were sacrificed, dissected, opened and subjected to gross necropsy. Major organs (liver, lung, kidney, heart, brain, ovaries, and spleen) were removed and wet weight recorded. Relative weight of the organs (organ weight/weight of the animal in grams) was calculated and compared with control. Histopathological examination of major organs such as the brain, liver, heart, spleen and lungs of one representative animal from each group was carried out. Blood samples were withdrawn from one animal of each group before their sacrifice and subjected to hematological investigation.

2.4.6 Freeze drying and stability. SACD liposomes were converted to solid state for stability evaluation by freeze drying process using 10% trehalose as the cryo-protectant. 1 mL formulations was filled in clear glass vials and pre frozen at −76 °C with cooling at the rate of 1 °C min⁻¹ followed by primary drying at −43 °C for 12 hours at a ramping rate of 1 °C min⁻¹ and secondary drying at 35 °C for 4 hours in a freeze drier (LabConco USA) with a ramping rate of 0.3 °C min⁻¹. The pressure was maintained at 0.08 mbar throughout the process. Samples were stored at 4 °C and 25 °C for a month and evaluated for changes in particle size using DLS technique after reconstitution with MilliQ water. All the experiments were performed in triplicate and results reported as data ± SD.

3. Results

3.1 Molecular dynamics simulations

We performed MD simulations of lipid bilayers with three different molecules inserted: (1) pure SA (with NH₂ head group), (2) SA with an arginine ligand and (3) SACD. From the simulation trajectories we calculated the solvent accessible surface area (SASA)⁵³ for each of the three ligands to determine the extent to which they are exposed to the solvent, thus accessible to the target. As shown in Fig. 1A, the head group of SA has a solvent accessible surface area (SASA) of 31% of its area exposed to the solvent while arginine stearylamine conjugate has a SASA of 35%. A significantly larger SASA was noted for the SACD ligand where around 66% of the head group was found to be solvent accessible. SACD thus has a greatly enhanced exposure to possible target receptors leading to enhanced targeting efficacy. Visualization of the system, also shown in Fig. 1, indicated that the SACD ligand sat higher in the membrane, the cause of the increased SASA. The mass density profile, shown in Fig. 2, indicates that SACD collocates with the phosphate group of the membrane lipids as compared to the arginine ligand and SA head group, thus again indicating that it sits higher in the membrane, also indicating enhanced exposure to the solvent, thus ability to target receptors. The stearylamine ligand preferred to align parallel to the membrane bilayer indicating that the SA head group is very rigid in comparison to the other 2 ligands. Both the arginine ligand and SACD have a broad distribution of angles, indicating enhanced mobility, however, the distribution for the SACD covers a broader range of angles, thus indicating yet greater mobility. Finally, we measured the charge density at the lipid membrane surface. The effective surface charge of the membrane bilayer results from the charge distribution of the lipid head group, embedded ligands and their interactions with the ions present in the simulated systems. This calculated effective surface charge at the membrane bilayer can be correlated with in vitro zeta-potential measurements.⁵⁴ We found that the charge density at the membrane bilayer surface is decreased for the case of SACD in comparison to that of the membrane bilayer with pure SA molecules. A similar qualitative trend was observed in the in vitro ζ-potential measurements.

Fig. 1 (A) Solvent accessible surface area of ligands when inserted in liposomal bilayers. (B) Visualization of ligands in bilayer.

Fig. 2 Mass density profile results: position of ligands within the bilayers and interaction with sodium ions.

3.2 Synthesis of SACD

Synthesis was carried out using a one pot scheme. Reaction time was optimized based on TLC. After 48 hours, three spots were visualized on the TLC plate, one corresponding to SA, one with a slightly lower R_f value than SA, corresponding to SA arginine monoconjugate and a third spot at the base indicating greater hydrophilicity of the dipeptide. The presence of a band at 1629 cm⁻¹ in the IR spectrum indicated the presence of an amide bond. The two N–H stretching bands in SA at 3333 cm⁻¹ and 3256 cm⁻¹ disappeared and one broad band at 3333 cm⁻¹ appeared. The IR spectra overlay of reactants and the isolated crude product is as shown in Fig. 3.

Fig. 3 IR spectra of stearylamine, arginine and the conjugate. Appearance of amide bond in product spectrum at 1640 cm⁻¹.

Mass spectra of the crude product revealed the following peaks.

Mass spectrum: 426.4 (M + 1), 676.6 (base peak).

MS/MS of 426.4: 426.4: (M + 1), 406.5: (M − 18), 326.4 (M − 99), 270.5: (M − 155).

MS/MS of 676: 676.5: (M + 1), 659.6: (M − 16), 617.5 (M − 58), 407.4 (M − 268), 365.5 (M − 310).

The peak at m/z 426.4 (M + 1) was attributed to the presence of SA monoarginine conjugate (mol. wt 452.4). Loss of ammonia resulted in the peak at m/z 406.5 via four membered cyclic rearrangement. The peak at m/z 326.4 was ascribed to the loss of arginine side chain from the conjugate and peak at m/z 365.5 due to loss of arginine side chain along with loss of carbon monoxide. The peak at m/z 676.6 was attributed to the presence of SA diarginine conjugate with a carbonyl imidazole function. At end of 72 hours, the reaction was stopped and product isolated. Since the TLC indicated the presence of SA in the product, it was thought to isolate the desired product using preparative TLC. From preparative TLC, it was possible to separate two fractions. SA and arginine monoconjugate merged together as a single spot near the solvent front whereas the product appeared at the base and was labelled as a novel cationic lipid (SACD). The SACD was an off white coloured free flowing solid. The IR spectra indicated the presence of a broad band around 3400 cm⁻¹ due to N–H stretch and another broad band near 1640 cm⁻¹ due to C=O stretch of secondary amides thus confirming the presence of conjugate. The purity of the conjugate was confirmed through mass spectroscopy. A single base peak at 608.6 was observed in the spectra indicative of the purity of the fraction. (Data: 608.6 (M + 1), 313.5 (M − 294.3)). The ¹H NMR spectra of SA and SACD is as shown in Fig. 4 and integration values in Table 1a and b respectively.

Fig. 4 ¹H NMR of stearylamine and SACD in CDCl₃.

Table 1 ¹H NMR values for (a) stearylamine (b) SACD

Chemical shift	Protons	Intensity	Multiplicity
a
0.85–0.90	Methyl	3H	Triplet
1.068	Amine	2H	Singlet
1.15–1.32	Methylene (CH2)n−3	30H	Broad multiplet
1.40–1.45	Methylene CH2b	2H	Multiplet
2.65–2.69	Methylene CH2a	2H	Triplet

---

b
0.85–0.90	Methyl	3H	Triplet
1.1–1.45	Methylene (CH2)15	30H	Broad multiplet
1.55–1.7	Methylene CH2b arginine side chain 2^nd ch2	10H	Multiplet
1.95–2.1	Methylene CH2 arginine side chain	4H	Triplet
2.3–2.4	C=NH	2H	Singlet
2.5–2.6	CH2–NH	4H	Multiplet
3.1–3.4	H–C–(NH–C=O)–C=O–NH	2H	Multiplet
3.6–3.7	CH2–NH–C=O	2H	Triplet
4.3–4.4	NH of ring	2H	Bs
4.4–5	NH of guanidine	4H	Bs
5.3–5.4	NH–C=O	1H	Bs
8.3–8.4	NH cyclic	1H	Bs

---

The peak at m/z 608.6 is attributed to the presence of a SA arginine cyclic dipeptide conjugate (mol. wt 607.87) (Fig. 5). The peak at m/z 313.5 is attributed to the loss of a stearylamido side chain (Fig. 5).

Fig. 5 Mass spectrum of the product after purification by preparative TLC.

3.3 Formulation and characterization of cationic formulations

In order to obtain a true estimate of the cationic nature of the agent, nanoparticles of the pure cationic agent stabilized by a non ionic emulsifier were prepared and evaluated for their particle size and zeta potential. The particle size and zeta potential of all formulations are shown in Tables 2 and 3 respectively. SACD nanoparticles were found to have a lower particle size as compared to SA nanoparticles. SA nanoparticles displayed a greater cationic charge as compared to SACD nanoparticles. Nonetheless, all particles were found to be cationic in nature. Stearylamine liposomes were found to have a greater cationic charge as compared to SACD liposomes, although both formulations displayed significantly greater charge as compared to blank liposomes. SACD liposomes were observed to possess a size of range of 50 nm to 200 nm as determined by TEM technique (Fig. 6). The wide size distribution could be attributed to the greater polydispersity of the sample as seen from DLS data (Table 2).

Table 2 Particle size distribution of cationic formulations

Formulation	Particle size (nm)	Polydispersity index
SA nanoparticles	256.9 ± 34.59	0.505 ± 0.033
SACD nanoparticles	126.9 ± 7.99	0.405 ± 0.062
Blank liposomes	144.6 ± 2.33	0.533 ± 0.003
SA liposomes (1.5 mol% SA)	231.2 ± 15.45	0.513 ± 0.067
SA liposomes (5 mol% SA)	207.1 ± 1.510	0.458 ± 0.006
SACD liposomes (1.5 mol% SACD)	157.1 ± 1.229	0.448 ± 0.049
SA liposomes (5 mol% SACD)	140.3 ± 5.138	0.343 ± 0.020

---

Table 3 Zeta potential of various cationic formulations

Formulation	Zeta potential (mV)
Stearylamine nanoparticles	+60.5 ± 2.85
SACD nanoparticles	+27.9 ± 1.15
Blank liposomes	−17 mV + 2.35
Stearylamine liposomes	+39.3 mV + 1.16
SACD liposomes	−2.09 mV + 1.12

---

Fig. 6 TEM images of SACD liposomes.

3.4 In vitro hemolysis

Percentage haemolysis obtained after incubating the blood cells with various formulation are as shown in the Fig. 7. Values below 10% are considered to be non-hemolytic whereas values above 25% are hemolytic.⁵⁵

Fig. 7 % hemolysed RBC's after incubation at 37 °C with various concentrations of stearylamine and SACD nanoparticles.

It was observed that though an increase in hemolytic activity was observed for both SA and SACD nanoparticles beyond 12.5 μM concentration, the increase was less sharp for the case of SACD with a maximum hemolysis value of 80% at 200 μM concentration compared to 110% hemolysis with SA at equimolar concentration thus confirming the reduced toxicity of the SACD to erythrocytes (Fig. 7).

3.5 In vitro cell uptake

Confocal microscopy images of Ln229 and HaCaT cells after treatment with various formulations are shown in Fig. 8A and B respectively. FACS profile and quantitative analysis are depicted in Fig. 9 and 10 respectively. As compared to blank liposomes, it was observed that both SA as well as SACD liposomes had greater cell uptake in both the cell lines evaluated. The SACD liposomes showed greater uptake in LN 229 cells as compared to SA liposomes. In HaCaT cells however, not much difference was observed between the cellular uptake of two liposomes tested.

Fig. 8 (A): Confocal microscopy images after treating LN229 cells with different concentrations of blank, SA and SACD liposomes. (B) Confocal microscopy images after treating HaCaT cells with different concentrations of blank, SA and SACD liposomes.

Fig. 9 FACS plot of LN229 cells and HaCaT treated with different formulations. Cells were treated with formulations equivalent to 5, 10 and 20 μM of cationic agent and cell fluorescence was quantified using flow cytometry.

Fig. 10 Mean fluorescence intensity of LN229 and HaCaT cells treated with different formulations. Data plotted as mean of triplicate experiments ± SD.

3.6 In vivo toxicity: acute toxicity study of SACD liposomes (OECD Guideline 420)

Acute toxicity was performed as per OECD Guideline 420. The dose of SACD selected was 0.8543 mg kg⁻¹ (based on doxorubicin drug loading in liposomes as reported in our previous work).⁴³ The animals were critically observed for first 4 hours of the dose administration and periodically after 24 hours for 14 days. There were no changes in their behaviour pattern. There were no significant changes in body weight in the formulation treated animals compared to control group (Fig. 11). During the 14 days of the experimental period, food consumption of all formulation treated animals was found to be comparable to the control group. On the 14^th day, animals were sacrificed and subjected to gross necropsy. Gross morphological examination did not reveal any abnormality in treatment group. There were no significant changes in the organ to body weight ratio in the formulation treated and control group (p < 0.05) (Fig. 12). There was no mortality observed during the course of the study. No gross pathology was seen in the formulation treated group in comparison to the control group (Fig. 13). Thus, SACD was found to be safe at the single intravenous dose of 0.8543 mg kg⁻¹. The haematological investigations of blood samples of the control group as well as that of the formulation demonstrated that there was no significant difference in the values of various parameters between the two groups. All the values for both the groups were in the normal range as reported for mice blood parameters (Table 4).

Fig. 11 Average body weight of animals on day 1, 7 and 14.

Fig. 12 Organ to body weight ratios of control and treated group on day 14.

Fig. 13 Histopathology of various organs for control (upper row) and SACD liposome treated group (lower row images). (A) Section of liver (B) section of kidney (C) section of heart (D) section of spleen (E) section of lungs. (CM: cardiac muscles; RP: red pulp; WP: white pulp of spleen; B: bronchiole; H: hepatocytes; CV: central vein; T: renal tubule; G: glomerulus.)

Table 4 Haematological investigation of mice blood samples

Parameter	Control	Formulation
Erythrocytes
Red blood cells	10.01	9.36
Haemoglobin	12.6	13
Haematocrit	49	47
Mean Corpuscular Volume (MCV)	49	50.2
Mean Corpuscular Haemoglobin (MCH)	12.6	13.9
Mean Corpuscular Haemoglobin Concentration (MCHC)	25.7	27.7
Leucocytes
White blood cells	6900	5900
Neutrophils	8	8
Lymphocytes	90	89
Eosinophils	1	1
Monocytes	1	2
Basophils	0	0
Platelets
Platelet count	1440 × 10^3	1248 × 10^3

---

3.7 Freeze drying and stability evaluation

Trehalose was employed as the cryo-protectant for freeze drying and it was found suitable at the concentration of 10% when incorporated in the formulations. The particle size of the formulations after freeze drying and during one month storage at 4 °C and 25 °C are as shown in the table below (Table 5).

Table 5 :Particle size data of freeze dried formulations after reconstitution when stored at various conditions for a month

Temperature	4 °C			25 °C
Time	0	15 days	30 days	0	15 days	30 days
Particle size	127.9 ± 2.687	165.4 ± 1.31	178.8 ± 2.682	127.9 ± 2.687	340.3 ± 11.52	Size in micron
PI	0.381 ± 0.006	0.294 ± 0.016	0.323 ± 0.018	0.381 ± 0.006	0.608 ± 0052	—

---

The formulations possessed 100 nm particle size even after reconstitution thereby validating freeze drying process was not to be detrimental to formulations when trehalose was used as a cryo-protectant. It was observed that there was slight increase in particle size and polydispersity of ligand loaded liposomes over a period of one month when stored at 4 °C though the formulations possessed size less than 200 nm even at the end of one month. However, an increase in both size and PI was observed for the formulations when stored at 25 °C.

4. Discussion

Toxicity associated with cationic agents hampers their application in parenteral delivery. With an endeavour to mitigate the toxicity of SA, we conjugated its head group with a non toxic cationic molecule, arginine. Arginine and arginine containing CPPs have been extensively explored for facilitating intracellular delivery. Moreover, cyclization of guanidine moiety in CPPs has manifested in increased rigidity and transduction across the cell membrane.³² Consequently, it was thought to conjugate SA with two arginine molecules by an intra-molecular cyclization to yield the final conjugate.

Molecular dynamics simulations were employed as a tool to understand the behaviour or the cyclic guanidine ligand in liposomes as compared to SA and SA arginine monoconjugate along with in vitro experiments. It was observed from MD results that SACD liposomes possessed lower surface charge, which was validated with zeta potential measurements as compared to SA liposomes. The cyclic dipeptide ligand, however sat higher in the membrane, with a greater solvent accessible surface area thus making it more accessible to interact with the negatively charged residues of the cell membrane and the cationic amino acid receptor, thereby anticipating in enhanced cell uptake.

Amino acid conjugations using Boc/Fmoc chemistry involve a number of protection/deprotection steps making the process cumbersome. In our study, we developed a one pot scheme for synthesizing the conjugate. Reactions of unprotected amino acids with various amines to give the corresponding amides using 1,1′-carbonyldiimidazole as an activating agent have been reported.⁵⁶ The reagent has also been used in oligomerization of amino acids through intramolecular acyl group transfer. The reaction works well with only small amines but fails with very hydrophobic molecules. The SA bearing a C₁₈ chain is extremely hydrophobic. It was thought, therefore, to explore the possibility of reacting unprotected AG with SA using CDI as an activating agent with some modifications in the reported procedure. Advantage was taken of the fact that SA dissolves in tetrahydrofuran (THF) at 50 °C. Tetrahydrofuran is water miscible and THF/arginine solution yield a single phase. Precipitation was observed, however, on addition of SA. Controlling the temperature and solvent composition critically was thus sought as a solution for obtaining a single phase.

After 48 hours, presence of the dipeptide formation was observed. Its structure was elucidated based on mass spectrometry (Fig. 14e). Since the amino groups of arginine are unprotected, once the mono conjugate is formed (Fig. 14a), the guanidine amino as well as the alpha amino group is free to react with the activated carboxyl group of another molecule to form one more amide bond resulting in a dipeptide (Fig. 14c). The alpha amino group is sterically hindered and the guanidine amino group being more nucleophilic reacts with another arginine molecule to form the dipeptide. Reports suggest that carbonyldiimidazole reacts with the amino groups in a similar way as a carboxyl group when excess CDI is present.⁵⁷ Once the dipeptide is formed, there are three primary amino groups which are free to react with a molecule of carbonyldiimidazole since excess CDI is present – one is the alpha amino group of arginine molecule conjugated to SA, second is alpha amino group of the second arginine molecule and third is the guanidine group of second arginine molecule. Based on mass fragmentation pattern, it is postulated that the alpha amino group of the second arginine molecule reacts with CDI to give the dipeptide imidazole N-carboxamide intermediate along with some mono arginine conjugate (Fig. 14c and b).

Fig. 14 (a) Arginine stearylamine monoconjugate (b) possible nucleophilic sites of attack (c) dipeptide imidazole N-carboxamide (d) isomers proposed on basis of mass spectrometry data (e) structure of SACD.

The reaction was further continued and after 72 hours, it was decided to purify this product so as to separate the dipeptide, monoconjugate and SA from each other. There are reports in the literature that show that when CDI is reacted with two moles of amine, initially an imidazole-N-carboxamide is formed which can dissociate to an isocyanate and is further trapped by another molecule of amine that forms a urea.⁵⁸ Advantage was taken of this fact to result in intramolecular cyclization of guanidine. The imidazole N-carboxamide underwent intramolecular urea formation. In the dipeptide structure, there are free amino groups present which can react with the carboxamide to form a urea and thereby release the imidazole. There are two free primary amino groups present now which can participate in urea formation – one is the α-amino group of the first arginine molecule and second is guanidine amine group of the second arginine molecule. Both yield compounds having the same mass i.e. 607.5 (Fig. 14d). The compound however yields a red colour on the TLC plate with Sakaguchi spray reagent indicating the free nature of the guanidine group. It has been established that di- and tri- substituted guanidines do not give a positive Sakaguchi test result.⁵⁹ Thus, it can be concluded that it is the free alpha amino group that reacts to finally yield the urea. The structure of the urea (Fig. 14e) has also been confirmed by the fragments obtained for 608.6 ion (Fig. 14e).

The synthesized conjugate was then incorporated into liposomes and nanoparticles. Although, the nanoparticles demonstrated a reduced cationic charge at the same weight concentration of SA, it could be attributed to a reduced number of moles of SACD. The charge was observed to be close to +30 mV thus confirming the cationic nature and colloidal stability of the formulation. Blank liposomes as well as liposomes containing SA, and SACD were prepared. SA and SACD were incorporated at concentrations of 1.5 mol% and 5 mol% respectively, in liposomes. Zeta potential greater than 30 mV was observed when ligands were incorporated at a concentration of 1.5 mol% thereby indicating colloidal stability. Therefore this concentration was chosen as the optimum concentration for performing molecular dynamic simulations and in vivo studies. For in vitro cell uptake, a higher concentration of SA and SACD was chosen for evaluation as lower concentrations were not optimum to yield significant uptake.

Phospholipids are known to be negatively charged and therefore the blank liposomes which are composed of phospholipids and cholesterol are negatively charged. The zeta potential of blank liposomes was found to be −17 mV. SA is used as a positive charge imparting agent in liposomes. It has an amine group which has a pK_a of 10 and therefore gets protonated at a pH of 7.4 thus giving a positive charge to the system. Thus, when incorporated at a concentration of 1.5 mol%, the zeta potential of blank liposomes changes from −17 mV to +39 mV. SACD was also postulated to have cationic charge. The zeta potential value was observed to be −2.09 mV. Presence of SACD increased the zeta potential values as compared to blank liposomes but not as much as that obtained with SA. This indicates that though these agents are not as strongly positively charged as SA (reason attributed to its toxicity), they are cationic. This observation is also in sync with the effective lower charge density observed in MD simulations. This result could be due to the fact that the novel ligand does not locate at the membrane–water interface, hence not providing any specific ion binding site as compared to stearylamine and arginine–stearylamine ligand.

SACD and SA nanoparticles were evaluated for haemolysis. Yoshihara E. et al. have reported the cytolytic activity of liposomes containing SA on rabbit erythrocytes. They found that liposomes containing less than 5% (w/w) of SA did not have any haemolytic activity, and the liposomes containing 7.5–12.5% of SA showed an abrupt increase of the haemolytic activity. Further, increase of SA in the liposome membrane resulted in the decrease of the haemolytic activity. The result suggests that only those liposomes with the critical density of SA had the ability to cause the haemolysis. However, in our study we observed that, SACD nanoparticles had significantly reduced toxicity to RBC's even at higher concentrations. This suggests that SACD is a potentially safe agent to be employed at higher concentrations as compared to SA to obtain the necessary charge.

In vitro cell uptake was studied using two cell lines viz. LN229 and HaCaT. LN229 is a glioblastoma cell line and has been reported to express highly sulfated keratan sulphate^60,61 in its cell membrane thereby possessing a negatively charged surface due to the presence of O-sulfate residues. It was thus adopted as one of the cell lines for appraising cellular uptake of cationic delivery systems. HaCaT cells are derived from normal human keratinocytes and in contrast do not possess excessive negatively charged residues on their cell membrane.⁶² They however manifest presence of cationic amino acid transporters and can be therefore of utility to study receptor mediated uptake of SACD liposomes. We observed that SACD liposomes penetrated LN229 cells more than SA liposomes, in spite of bearing a lower cationic charge. This could be indeed explained on the basis of MD results which reveal a significant portion of SACD ligand head group to be protruding outside the membrane. This could have resulted in an increased interaction of the liposomes with negatively charged sulfated residues and fatty acids thereby amounting to increased uptake as compared to SA liposomes. In contrast, evaluation of cellular uptake in HaCaT cells did not reveal any significant difference between the SACD liposomes and SA liposomes. This could be attributed to failure of recognition of the SACD ligand by the cationic amino acid receptor thereby indicating non receptor mediated uptake similar to SA liposomes. Both SA as well as SACD liposomes however, had greater cellular penetration in both cell lines as compared to blank liposomes, thus reasserting the importance of cationic charge in expediting cellular uptake.

Acute toxicity study in mice revealed the absence of any damage to the highly perfused organs with no abnormal hematological parameters, further establishing the safety of the synthesized agent in vivo.

5. Conclusion

A novel stearylated arginine derivative was synthesized using an easy one pot reaction scheme. Molecular dynamics served as a tool for comprehensive understanding and validation of ligand behaviour within delivery systems. Simulation results could be corroborated with experimental values thus providing another dimension in feasibility of applying simulations to delivery systems' design. The ligand was found to be better hemocompatible and biocompatible in vivo. The novel cationic agent was successfully subsumed in liposomes which demonstrated greater propensity to penetrate cellular membrane and effect intra cellular delivery of drugs. Efforts are now being planned towards application of these novel cationic liposomes for gene delivery.

Acknowledgements

Authors are thankful to IIT Mumbai for performing NMR, Mass spectroscopy and TEM studies. Authors are also thankful to Finnish IT Centre for Scientific Computing (CSC) for providing computational resources. Authors also wish to thank Academy of Finland, Magnus Ehrnrooth Foundation, Finland for funding and All India Council for Technical Education and Lady Tata Memorial Trust for providing financial assistance.

References

1. S. Blau, T. T. Jubeh, S. M. Haupt and A. Rubinstein, Drug targeting by surface cationization, Crit. Rev. Ther. Drug. Carrier. Syst., 2000, 17(5), 425–465.
2. S. Simões, A. Filipe, H. Faneca, M. Mano, N. Penacho and N. Düzgünes, et al., Cationic liposomes for gene delivery, Expert Opin. Drug Delivery, 2005, 2(2), 237–254.
3. G. Yogeeswaran and P. Salk, Metastatic potential is positively correlated with cell surface sialylation of cultured murine tumor cell lines, Science, 1981, 212(4502), 1514–1516.
4. M. Abercrombie and E. Ambrose, The surface properties of cancer cells: a review, Cancer Res., 1962, 22(5), 525–548.
5. M. Yamauchi, H. Kusano, E. Saito, T. Iwata, M. Nakakura and Y. Kato, et al., Development of wrapped liposomes: novel liposomes comprised of polyanion drug and cationic lipid complexes wrapped with neutral lipids, Biochim. Biophys. Acta, Biomembr., 2006, 1758(1), 90–97.
6. S. Bai, V. Gupta and F. Ahsan, Cationic liposomes as carriers for aerosolized formulations of an anionic drug: safety and efficacy study, Eur. J. Pharm. Sci., 2009, 38(2), 165–171.
7. M. El-Shabouri, Positively charged nanoparticles for improving the oral bioavailability of cyclosporin-A, Int. J. Pharm., 2002, 249(1), 101–108.
8. H. Natsume, S. Iwata, K. Ohtake, M. Miyamoto, M. Yamaguchi and K.-i. Hosoya, et al., Screening of cationic compounds as an absorption enhancer for nasal drug delivery, Int. J. Pharm., 1999, 185(1), 1–12.
9. K. Shimizu and N. Oku, Cancer anti-angiogenic therapy, Biol. Pharm. Bull., 2004, 27(5), 599–605.
10. C. R. Dass, Improving anti-angiogenic therapy via selective delivery of cationic liposomes to tumour vasculature, Int. J. Pharm., 2003, 267(1), 1–12.
11. R. B. Campbell, D. Fukumura, E. B. Brown, L. M. Mazzola, Y. Izumi and R. K. Jain, et al., Cationic charge determines the distribution of liposomes between the vascular and extravascular compartments of tumors, Cancer Res., 2002, 62(23), 6831–6836.
12. M. S. Al-Dosari and X. Gao, Nonviral gene delivery: principle, limitations, and recent progress, AAPS J., 2009, 11(4), 671–681.
13. V. V. Dhawan, G. V. Joshi, A. S. Jain, Y. P. Nikam, R. P. Gude and R. Mulherkar, et al., Apoptosis induction and anti-cancer activity of LeciPlex formulations, Cell. Oncol., 2014, 37(5), 339–351.
14. A. A. Date, M. S. Nagarsenker, S. Patere, V. Dhawan, R. Gude and P. Hassan, et al., Lecithin-based novel cationic nanocarriers (Leciplex) II: improving therapeutic efficacy of quercetin on oral administration, Mol. Pharm., 2011, 8(3), 716–726.
15. A. S. Jain, D. T. Makhija, P. N. Goel, S. M. Shah, Y. Nikam and R. P. Gude, et al., Docetaxel in cationic lipid nanocapsules for enhanced in vivo activity, Pharm. Dev. Technol., 2014, 1–10.
16. P. Khachane, A. Jain, V. Dhawan, G. Joshi, A. Date and R. Mulherkar, et al., Cationic nanoemulsions as potential carriers for intracellular delivery, Saudi Pharm. J., 2015, 23(2), 188–194.
17. S. M. Shah, M. Ashtikar, A. S. Jain, D. T. Makhija, Y. Nikam and R. P. Gude, et al., LeciPlex, invasomes, and liposomes: a skin penetration study, Int. J. Pharm., 2015, 490(1), 391–403.
18. S. Zhang, Y. Xu, B. Wang, W. Qiao, D. Liu and Z. Li, Cationic compounds used in lipoplexes and polyplexes for gene delivery, J. Controlled Release, 2004, 100(2), 165–180.
19. I. A. Michael Ash, Handbook of Preservatives: Synapse Information Resource, 2004.
20. E. Yoshihara and T. Nakae, Cytolytic activity of liposomes containing stearylamine, Biochim. Biophys. Acta, Biomembr., 1986, 854(1), 93–101.
21. D. Adams, G. Joyce, V. Richardson, B. E. Ryman and H. Wisniewski, Liposome toxicity in the mouse central nervous system, J. Neurol. Sci., 1977, 31(2), 173–179.
22. H. Lv, S. Zhang, B. Wang, S. Cui and J. Yan, Toxicity of cationic lipids and cationic polymers in gene delivery, J. Controlled Release, 2006, 114(1), 100–109.
23. A. Pinazo, R. Pons, L. Pe rez and M. R. Infante, Amino Acids as Raw Material for Biocompatible Surfactants, Ind. Eng. Chem. Res., 2011, 50, 4805–4817.
24. S. A. Ruckman, X. Rocabayera, J. F. Borzelleca and C. B. Sandusky, Toxicological and metabolic investigations of the safety of N-alpha-lauroyl-L-arginine ethyl ester monohydrochloride (LAE), Food Chem. Toxicol., 2004, 42(2), 245–259.
25. E. Vives, Cellular uptake of the Tat peptide: an endocytosis mechanism following ionic interactions, J. Mol. Recognit., 2003, 16(5), 265–271.
26. L. R. Wright, J. B. Rothbard and P. A. Wender, Guanidinium rich peptide transporters and drug delivery, Curr. Protein Pept. Sci., 2003, 4(2), 105–124.
27. T. Hagigit, T. Nassar, F. Behar-Cohen, G. Lambert and S. Benita, The influence of cationic lipid type on in vitro release kinetic profiles of antisense oligonucleotide from cationic nanoemulsions, Eur. J. Pharm. Biopharm., 2008, 70(1), 248–259.
28. S. Futaki, W. Ohashi, T. Suzuki, M. Niwa, S. Tanaka and K. Ueda, et al., Stearylated arginine-rich peptides: a new class of transfection systems, Bioconjugate Chem., 2001, 12(6), 1005–1011.
29. C. Morán, P. Clapés, F. Comelles, T. García, L. Pérez and P. Vinardell, et al., Chemical structure/property relationship in single-chain arginine surfactants, Langmuir, 2001, 17(16), 5071–5075.
30. D. M. Copolovici, K. Langel, E. Eriste and U. Langel, Cell-penetrating peptides: design, synthesis, and applications, ACS Nano, 2014, 8(3), 1972–1994.
31. H. D. Herce, A. E. Garcia and M. C. Cardoso, Fundamental molecular mechanism for the cellular uptake of guanidinium-rich molecules, J. Am. Chem. Soc., 2014, 136(50), 17459–17467.
32. G. Lattig-Tunnemann, M. Prinz, D. Hoffmann, J. Behlke, C. Palm-Apergi and I. Morano, et al., Backbone rigidity and static presentation of guanidinium groups increases cellular uptake of arginine-rich cell-penetrating peptides, Nat. Commun., 2011, 2, 453.
33. A. Bunker, A. Magarkar and T. Viitala, Rational design of liposomal drug delivery systems, a review: combined experimental and computational studies of lipid membranes, liposomes and their PEGylation, Biochim. Biophys. Acta, Biomembr., 2016, 1858(10), 2334–2352.
34. J. Lehtinen, A. Magarkar, M. Stepniewski, S. Hakola, M. Bergman and T. Róg, et al., Analysis of cause of failure of new targeting peptide in PEGylated liposome: molecular modeling as rational design tool for nanomedicine, Eur. J. Pharm. Sci., 2012, 46(3), 121–130.
35. P. Pathak, V. Dhawan, A. Magarkar, R. Danne, S. Govindarajan and S. Ghosh, et al., Design of cholesterol arabinogalactan anchored liposomes for asialoglycoprotein receptor mediated targeting to hepatocellular carcinoma: in silico modeling, in vitro and in vivo evaluation, Int. J. Pharm., 2016, 509(1), 149–158.
36. J. A. Almeida, M. C. Morán, M. R. Infante and A. Pais, Interaction of arginine-based cationic surfactants with lipid membranes. An experimental and molecular simulation study, ARKIVOC, 2010, 34, 34–40.
37. I. Z. Zubrzycki, Y. Xu, M. Madrid and P. Tang, Molecular dynamics simulations of a fully hydrated dimyristoylphosphatidylcholine membrane in liquid-crystalline phase, J. Chem. Phys., 2000, 112(7), 3437–3441.
38. W. L. Jorgensen and J. Tirado-Rives, The OPLS [optimized potentials for liquid simulations] potential functions for proteins, energy minimizations for crystals of cyclic peptides and crambin, J. Am. Chem. Soc., 1988, 110(6), 1657–1666.
39. A. Maciejewski, M. Pasenkiewicz-Gierula, O. Cramariuc, I. Vattulainen and T. Rog, Refined OPLS all-atom force field for saturated phosphatidylcholine bilayers at full hydration, J. Phys. Chem. B, 2014, 118(17), 4571–4581.
40. W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey and M. L. Klein, Comparison of simple potential functions for simulating liquid water, J. Chem. Phys., 1983, 79(2), 926–935.
41. B. Hess, C. Kutzner, D. Van Der Spoel and E. Lindahl, GROMACS 4: algorithms for highly efficient, load-balanced, and scalable molecular simulation, J. Chem. Theory Comput., 2008, 4(3), 435–447.
42. S. Nosé, A unified formulation of the constant temperature molecular dynamics methods, J. Chem. Phys., 1984, 81(1), 511–519.
43. W. G. Hoover, Canonical dynamics: equilibrium phase-space distributions, Phys. Rev. A, 1985, 31(3), 1695.
44. M. Parrinello and A. Rahman, Polymorphic transitions in single crystals: a new molecular dynamics method, J. Appl. Phys., 1981, 52(12), 7182–7190.
45. B. Hess, H. Bekker, H. J. Berendsen and J. G. Fraaije, LINCS: a linear constraint solver for molecular simulations, J. Comput. Chem., 1997, 18(12), 1463–1472.
46. U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee and L. G. Pedersen, A smooth particle mesh Ewald method, J. Chem. Phys., 1995, 103(19), 8577–8593.
47. W. Humphrey, A. Dalke and K. Schulten, VMD: visual molecular dynamics, J. Mol. Graphics, 1996, 14(1), 33–38.
48. C. I. Bayly, P. Cieplak, W. Cornell and P. A. Kollman, A well-behaved electrostatic potential based method using charge restraints for deriving atomic charges: the RESP model, J. Phys. Chem., 1993, 97(40), 10269–10280.
49. J. Wang, W. Wang, P. A. Kollman and D. A. Case, Automatic atom type and bond type perception in molecular mechanical calculations, J. Mol. Graphics Modell., 2006, 25(2), 247–260.
50. M. J. Parnham and H. Wetzig, Toxicity screening of liposomes, Chem. Phys. Lipids, 1993, 64(1–3), 263–274.
51. C. Chen, Y. C. Cheng, C. H. Yu, S. W. Chan, M. K. Cheung and P. H. F. Yu, In vitro cytotoxicity, hemolysis assay, and biodegradation behavior of biodegradable poly(3 hydroxybutyrate)–poly(ethylene glycol)–poly(3 hydroxybutyrate) nanoparticles as potential drug carriers, J. Biomed. Mater. Res., Part A, 2008, 87(2), 290–298.
52. P. O. Pathak, M. S. Nagarsenker, C. R. Barhate, S. G. Padhye, V. V. Dhawan and D. Bhattacharyya, et al., Cholesterol anchored arabinogalactan for asialoglycoprotein receptor targeting: synthesis, characterization, and proof of concept of hepatospecific delivery, Carbohydr. Res., 2015, 408, 33–43.
53. B. Lee and F. M. Richards, The interpretation of protein structures: estimation of static accessibility, J. Mol. Biol., 1971, 55(3), 379.
54. A. Magarkar, V. Dhawan, P. Kallinteri, T. Viitala, M. Elmowafy and T. Róg, et al., Cholesterol level affects surface charge of lipid membranes in saline solution, Sci. Rep., 2014, 4, 1–5.
55. K. Amin and R. M. Dannenfelser, In vitro hemolysis: guidance for the pharmaceutical scientist, J. Pharm. Sci., 2006, 95(6), 1173–1176.
56. R. K. Sharma and R. Jain, Unprecedented 1,-Carbonyldiimidazole-Mediated Amidation of Unprotected-Amino Acids in Water, Synlett, 2007, 18(4), 603–606.
57. A. Armstrong, N,N′-Carbonyldiimidazole, John Wiley & Sons, 1999.
58. R. Paul and G. W. Anderson, N,N′-Carbonyldiimidazole, a New Peptide Forming Reagent1, J. Am. Chem. Soc., 1960, 82(17), 4596–4600.
59. J. D. Mold, J. M. Ladino and E. Schantz, The Sakaguchi and Biacetyl Reactions for the Identification of Alkyl Guanidines, J. Am. Chem. Soc., 1953, 75(24), 6321–6322.
60. Y. Kato, N. Hayatsu, M. K. Kaneko, S. Ogasawara, T. Hamano and S. Takahashi, et al., Increased expression of highly sulfated keratan sulfate synthesized in malignant astrocytic tumors, Biochem. Biophys. Res. Commun., 2008, 369(4), 1041–1046.
61. N. Hayatsu, S. Ogasawara, M. K. Kaneko, Y. Kato and H. Narimatsu, Expression of highly sulfated keratan sulfate synthesized in human glioblastoma cells, Biochem. Biophys. Res. Commun., 2008, 368(2), 217–222.
62. Y. Chen, L. Zhou, L. Yuan, Z.-h. Zhang, X. Liu and Q. Wu, Formulation, characterization, and evaluation of in vitro skin permeation and in vivo pharmacodynamics of surface-charged tripterine-loaded nanostructured lipid carriers, Int. J. Nanomed., 2012, 7, 3023.

---