Introduction

Nowadays there is a great interest in new delivery systems for local anesthetics, especially for topical application because of its easy-to-use, painless and systemic side effect-free nature [1–4]. Liposomes are spherical vesicles (usually 0.05–5 µm in diameter) that are formed with energy input when certain phospholipids are hydrated in aqueous media [5]. The vesicles consist of one or more concentrically ordered phospholipid bilayers: the fatty acid chains are in the core of the bilayer, while the hydrophilic heads are oriented to the aqueous phase [6]. Liposomes improve drug bioavailability, reduce systemic toxicity and increase the half-lives of drugs in vivo [7–11]. Furthermore, these carriers enable a more intense localization of the active agent in the layers of the skin [12].

DMPC is a liposome-forming saturated neutral phospholipid with a smaller head group widely used as a model system of biomembranes since lecithins are a major component of most mammalian cell membranes [13, 14].

Local anesthetics (LA) comprise two major components: a lipophilic aromatic group and a polar region, connected by an intermediate carboxyl group in an amide bond [15]. Therefore, they can interact with the liposome membranes, usually by sitting in the lipid region, but a fraction of molecules is also retained in the aqueous phase [16]. Lidocaine (LID) is a commonly used local anesthetic with fast onset and intermediate duration of action (90–240 min) [17].

While in most studies and in the marketed formulations this local anesthetic is used in the hydrochloride form, we have chosen the base form because of its improved lipophilicity, thus better penetration properties through the lipophilic stratum corneum and the ability to form a depot in the hydrophilic dermis [18]. Furthermore, hydrophobicity is also crucial for drug partitioning into the nerve fibers, thus an appropriate amount of LA molecules remain within that membrane [17]. In addition, it was shown before that uncharged LID preferentially interacts with neutral membranes [19], such as DMPC.

In our studies, a series of samples was prepared showing a constant lipid concentration and increasing concentrations of the LID to examine the effects of lidocaine on the membrane properties.

The liposome preparations are always a mixture of entrapped and unentrapped drug fractions. To determine the encapsulation efficiency (EE %), the first step for the most common methods is the separation between the encapsulated drug (within the liposomes) and the free drug. This separation can be performed with mini-column centrifugation, dialysis membrane and ultracentrifugation [20]. After the separation, most of the published studies focus on measuring the unentrapped drug concentration in the supernatant and subtract this concentration from the total drug concentration (indirect method) [21–25]. The other technique is that when the supernatant is removed after the separation, the lipid bilayer (containing drug-loaded liposomes) is disrupted with organic solvent, and the released material is quantified (direct method). The above-mentioned procedures are obviously very laborious and time-consuming; furthermore, the attained results depend on the separation, which may not be complete [26].

On the other hand, differential scanning calorimetry (DSC) can serve as a powerful tool for the quality control of liposomes without needing to separate them. This method can give information about the drug–lipid interactions, size, partition coefficient and encapsulation efficiency with one measurement [27]. In our work, we focused on the comparison of the generally used ultracentrifugation method and DSC for the evaluation of encapsulation efficiency.

The principle of using DSC for this approach is based on the reduction in temperature of the main phase transition depending on the partitioning between LID and the lipid in the fluid or in the gel phase [28].

If the drug mixes ideally with the fluid phase of the membrane but is completely excluded from the gel, the difference between the actual phase transition temperature (T) and the temperature of the main phase transition of pure lipid (T 0) can be written as (∆T m = T 0–T)

$$\Delta T_{\text{m}} \left( X \right) = - \frac{{RT^{2} }}{\Delta H}X_{\text{d}}^{\text{b}}$$
(1)

where R is the ideal gas constant (1.9858775 cal K−1 mol−1), ΔH is the enthalpy of the main phase transition and \(X_{\text{d}}^{\text{b}}\) is the molar fraction of the drug bounded in the liposome [29]. So, the shift in the melting temperature is independent of the special properties of the drug (as long as the ∆H of melting is not strongly affected) and proportional to its mole fraction in the membrane [30].

Knowing the mass of the liposome measured, units can be converted from mole fraction to mass [31]:

$$c_{\text{d}}^{\text{b}} = c_{\text{lipid}} \frac{{X_{\text{d}}^{\text{b}} }}{{1 - X_{\text{d}}^{\text{b}} }}$$
(2)

where \(c_{\text{d}}^{\text{b}}\) is the concentration of the bounded drug and \(c_{\text{lipid}}\) is the lipid concentration. From these values, the encapsulation efficiency can be calculated:

$${\text{EE }}\% = \frac{{c_{\text{d}}^{\text{b}} }}{{c_{\text{d}}^{\text{total}} }} \approx \frac{{X_{\text{d}}^{\text{b}} }}{{X_{\text{d}}^{\text{total}} }}$$
(3)

To characterize a formulation, it is also useful to determine its membrane–water partitioning properties because lipophilicity plays an important role in biological activity. The direct calculation of solute partitioning into bilayers can also be monitored by DSC [31]. According to previous empirical evidence, the free drug concentration is proportional to the mole ratio of drug to lipid in the membrane (R), so the mole ratio partition coefficient (K R) [30, 32, 33] can be calculated:

$$K_{\text{R}} = \frac{R}{{c_{\text{d}}^{\text{free}} }} = \frac{{c_{\text{d}}^{\text{b}} }}{{c_{\text{d}}^{\text{free }} c_{\text{lipid}} }}$$
(4)

where \(c_{\text{d}}^{\text{total}}\) is the total drug concentration, \(c_{\text{d}}^{\text{free}}\) is the unentrapped drug concentration and \(X_{\text{d}}^{\text{total}}\) is the maximum bounded drug fraction.

Experimental

Materials

Phospholipid (1,2-dimyristoyl-sn-glycero-3-phosphocholine, DMPC) was supplied by Avanti Polar Lipids. Chloroform, ethanol, N-(2-hydroxyethyl) piperazine-N’-(2-ethanesulfonic acid) (HEPES), sodium chloride and lidocaine were obtained from Sigma-Aldrich.

Preparation of liposomes

Liposomes were prepared with the dry film hydration method [6, 34]. Stock solutions were prepared with chloroform of DMPC and LID. Aliquots were added to individual vials to reach 5 mg mL−1 lipid and 1-2-3-4-10 w/w% LID. These solutions were dried under rotation (RVC 2-18 Speed Dry Rotational Vacuum Concentrator, Martin Christ Gefriertrocknungsanlagen GmbH, Germany, 30 °C, 1.5 h, 1500 rpm). The lipid film was then placed in a vacuum desiccator overnight, to ensure the complete removal of the solvent [35]. The hydration of the film was done with 1 mL HEPES buffer (20 mM, containing 154 mM NaCl, pH = 7.4) at room temperature, alternating with vortex agitation for 5 min. Liposomal formulations were stored at 4 °C and used within 1 week.

Particle size and zeta potential measurements

Measurements were taken with a Malvern Nano ZS based on dynamic light scattering. The system works according to the phase analysis light scattering (PALS) principle, and the data are automatically evaluated on the basis of the Smoluchowski equation (the particle size is much larger than the Debye length, ≈1 nm). The sample was thermostated to 25 °C with a built-in Peltier device. Measurements were taken in standard disposable cuvettes using Malvern’s dip cell. Each measurement was taken in triplicate. To ensure the validity of the data, a zeta standard was measured every 30 min. For these measurements, the samples were diluted with their aqueous phase in order to avoid multiscattering phenomena. The polydispersity index was also evaluated as a measurement of the homogeneity of the dispersion.

Ultracentrifugation

The encapsulation efficiency of the drug was determined for several preparations using the ultracentrifugation method combined with spectrophotometry. About 1 g of each preparation was placed into Beckman polycarbonate centrifuge tubes and diluted to 1.5 mL with HEPES buffer. The samples were centrifuged in a Beckman Coulter Optima XE-90 Ultracentrifuge for 3 h at 35,000 rpm at 4 °C. The supernatant was removed, leaving the pellet containing the liposomes at the bottom of the tubes. Then a washing step was performed: 1.5 mL of the HEPES buffer was added to the pellet and the centrifugation process was repeated. After that, the supernatant was removed again. Both the supernatant (“indirect method”) and the pellet (dissolved in 1 mL ethanol—“direct method”) were measured with UV spectrometer at 262 nm to determine the concentration of LID [36].

It is a crucial to carefully adjust the experimental parameters for the complete pelleting, but some studies reported micellization for LA-loaded liposomes [37, 38], which can cause difficulties in determining encapsulation efficiency in many preparations.

To prove that all the lipids are pelleted down, we performed Bartlett assay [39] before the centrifugation process and after redissolving the pellets; thus, we can be sure that the pellets contain all the lipids and there are no micelles or liposomes in the supernatant.

DSC measurements

DSC measurements were performed using a MicroCal VP-DSC device (MicroCal Inc., Northampton, USA). Before the calorimetric experiments, the solutions were degassed and then filled into the sample cell (0.4988 mL). A heating rate of 1 °C min−1 in the 5–80 °C range was applied. Phase transition temperatures and enthalpy (∆H) scales were calculated. As a reference, a 20 mM HEPES buffer solution was used. Three up and down scans were performed for each sample to prove the reproducibility. All curves shown in the figures originate from the first heating scan. The Origin 7.0 software was used to subtract the baselines from the curves and to convert the raw data into data of molar heat capacity. The thermodynamic parameters were obtained by using the non-two-state model provided by the software.

Statistical analysis

One-way ANOVA followed by the Bonferroni test was used to determine the statistical differences between the results by using GraphPad Prism software (GraphPad Software Inc., La Jolla, CA, USA). Differences were regarded as significant if p < 0.05*, p < 0.01** and p < 0.001***.

Results and discussion

Particle size and zeta potential measurements

The light scattering analysis showed that the sizes of liposomes were in the range of 1844 ± 562.9 nm and 4842 ± 275.57 nm. The results presented a decrease in vesicle size and a reduction in homogeneity with increasing the amount of the added LID, as indicated by the growth of the polydispersity index (Table 1). The polydispersity index (PDI) of the investigated vesicles showed values from 0.279 ± 0.253 to 1.000 ± 0.000, representing heterogenous populations (PDI > 0.3) of vesicles [40].

Table 1 Particle size and zeta potential
Full size table

These findings could be explained by considering that the drug will be located within the liposomal bilayer and could alter the microstructure of the vesicular membrane, and reduce the liposomal membrane organization.

On the other hand, measurements of zeta potential values showed that it was barely influenced by the presence of the drug, so a contribution of the drug to the liposomal charge can be excluded (Table 1). Therefore, the nearly zero charge of the vesicles could be attributed to the properties of the DMPC molecules, which is a zwitterionic lipid that, at physiological pH, forms membranes with practically zero surface charge density [41]. The slight difference in the values could be explained by the measurement difficulties around the value zero.

DSC

The effect of LID on the thermotropic behavior of DMPC was investigated by DSC as a function of the anesthetic concentration.

Literature data report that pure DMPC liposomes show a strong and sharp main endothermic transition near 24 °C, with ∆H ≈ 6000 cal mol−1, because of the conversion of the gel phase to the lamellar liquid crystal phase [42]. These bilayer membranes also exhibit a weak pre-transition peak at about 13.5 °C, arising from the conversion of a lamellar gel phase to a rippled gel phase [43]. Figure 1 shows the heating curves of the DMPC multilamellar vesicles in the absence and presence of different LID amounts up to 10 w/w%.

Fig. 1
figure 1

DSC curves of the liposomes showing the pre transitions (a) and main transitions (b)

Full size image

Table 2 gives the values of T m (main transition temperature) and ∆H (enthalpy of transition) measured in all the formulations in the heating cycles.

Table 2 Main transition temperature and enthalpy values of the formulations
Full size table

The values measured for the T m, ∆H in pure DMPC liposomes (23.85 °C, 5293 cal mol−1, respectively) were in good agreement with the literature data [44].

The most outstanding feature of the series of curves is that the main phase transition for DMPC–represented by the peak in Cp—is progressively lowered and broadened with increasing concentrations of LID. The position of the main peak is reduced from T m = 23.85 °C for DMPC down to about 22.66 °C with 10 w/w% LID. This suggests a chain disordering effect in the lipid membrane (Fig. 2).

Fig. 2
figure 2

Phase transition temperatures of DMPC bilayers as a function of LID concentration

Full size image

The analysis of ∆H, given by the peak area, can provide information about the impact (location) of the LID molecules in the phospholipid bilayers. In the investigated formulations, LID does not affect ∆H but causes a decrease in T m. This can be explained by a superficial interaction between the LID and the DMPC molecules and/or the intercalation of LID molecules between the chains of the lipids without reducing the stability of the membrane [45].

The pre-transition peak is broad and nearly symmetrical and, because of the formation of an intermediate metastable phase, slowly turns into the lamellar gel phase [43]. This peak became smaller by the addition of LID and disappeared at the concentration of 10 w/w%. The vanishing nature of the pre-transition is common upon the addition of membrane solutes [46, 47].

Encapsulation efficiency

The encapsulation efficiency values of the liposomal formulations obtained by different experimental methods are represented in Fig. 3 and Table 3.

Fig. 3
figure 3

Drug encapsulation efficiency expressed as % of loading concentration

Full size image
Table 3 Encapsulation efficiency
Full size table

The results are quite ambiguous. With the most widespread indirect method, we measured notably higher EE % values (9.4–32.2%) and the measurements also resulted in a high standard deviation, while the results of the direct method (3.5–10.1%) and the DSC measurements (5.4–13.0%) correlate nicely. These differences could be explained by some possible experimental errors in connection with the indirect method. First of all, indirect method is a derived result, and it is evident that each experimental step and each further calculation will introduce an experimental error and, moreover, may cause a loss of product (for example adsorption on the vial wall or pipette tips). The other probable explanation is the imperfect separation and the presence of liposomes in the supernatant as well.

According to Nernst’s partitioning law (1891) for dilute solutions, the ratio of concentrations in two separated phases is the constant partition coefficient (K R). The calculated average K R values are also similar for the direct and DSC methods (0.00903 ± 0.00455 mM−1 and 0.00895 ± 0.00468 mM−1) and really different for the indirect method (0.0375 ± 0.0306 mM−1).

Considering these outcomes, we can conclude that the direct method and DSC confirmed each other’s results, while the indirect method suffered from an unknown error source.

Conclusions

In conclusion, we successfully developed and characterized a liposomal lidocaine formulation. The effect of LID on the liposome structure is notable even in the presence of very small amounts of foreign substances added. The encapsulation efficiency of the local anesthetic on the lipid bilayer of DMPC multilamellar vesicles was investigated using DSC. The results were compared with those obtained with ultracentrifugation.

We conclude that the DSC method is more convenient compared to the techniques used generally for the determination of encapsulation efficiency in cases when phase transition measurements are taken with the aim of obtaining further information. These findings should be extended for higher lipid concentrations and other formulations.

Although this work is still preliminary, it provides an integrated approach to the study of encapsulation efficiency with a novel method. Moreover, it provides guiding lines for future investigations.