**1. Introduction**

A growing body of evidence indicates dietary polyphenols to be efficient antioxidants counteracting negative effects of oxidative stress accompanying many diseases such as cardiovascular diseases, inflammation, and also cancer [1–6]. In line with previous reports, many of these compounds may act as chemopreventive or even chemotherapeutic agents [5,7–9]. Specifically to malignant cells, cinnamic

acid (CinA) and its hydroxy derivatives have been shown to exhibit both antioxidant a antineoplastic activity, with their cytostatic effect being largely related to their structural characteristics [10–13]. However, the mode of action of CinA and its derivatives is still incompletely determined, but may involve scavenging of reactive oxygen species, modulation of gene expression, activation of xenobiotics metabolism-related enzymes, and regulation of signal transduction pathways essential for tumor cell growth and progression [14].

Despite their natural origin and many promising biological effects, bioavailability of hydroxycinnamic acids presents certain limitations. Although working well in aqueous media, their hydrophilic nature is usually a restriction for lipophilic system protection [15]. Lipophilicity is a major physicochemical property of chemical substances, which affects their biological activities. It is known to be important for describing both pharmacodynamic and pharmacokinetic aspects of drug action. In biological systems, lipophilicity largely determines key properties of potential pharmacological agents, such as solubility of substances in biological fluids, penetration through the biological membranes, rate of absorption, affinity to plasma and tissue proteins, and distribution into the specific body compartments [16]. Lipophilicity is commonly expressed by the logarithm of n-octanol/water partition coefficient (log*P*) for ionizable compounds of a neutral form of compounds. Initially, log*P* was considered important in drug and pesticide discovery and design, but now it is an essential characteristic of all chemicals. This is because log*P* largely determines chemicals fate both inside a living organism and in the environment. Log*P* values are typically between −3 (very hydrophilic) and +10 (extremely hydrophobic) [17]. For ionizable forms of compounds, the distribution coefficient (log*D*) at a specific pH is also often used. As opposed to log*P*, which is only valid for a single electrical species, log*D* represents the pH-dependent mixture of all electrical species occurring at given pH [18].

Based on the observation that most medication drugs are relatively small and lipophilic molecules, Lipinski et al. formulated "the rule of five" (Ro5) [19]. This is a rule of thumb in determining if a pharmacologically/biologically active chemical compound, a candidate for a drug, would be potentially bioavailable via oral administration in humans [20]. According to Ro5, chemicals are less prone to adsorb on the cell membrane and more likely to permeate the bilayer when their calculated n-octanol/water partition coefficient (*c*log*P*) ≥ 5, they have ≥10 H-bond acceptors, ≥5 H-bond donors, and their molecular weight (*M*WT) exceeds 500 g/mol [19]. Because each threshold is a multiple of 5, the rule was called Ro5. Molecules not complying with more than one of these rules may have problems with bioavailability [21].

To obtain lipophilicity parameters of significant pharmacokinetic and pharmacodynamic relevance, partition coefficients between (mostly but not exclusively) artificial membranes and water have been tested recently [22]. Monolayers and bilayers have been used for several decades as models of biomembranes to study solute/biological membranes interactions. A variety of lipids and their mixtures may serve as components to prepare both types of these membrane models. Nevertheless, phospholipids tend to be the most widely exploited as they are easily reproducible and have well standardized systems [23]. These lipids are either negatively charged or zwitterionic (electrically neutral due to an equal number of positive and negative charges). Phospholipids are amphipathic and occur naturally in all living organisms as the major components of cell membranes. Most biological membranes are characterized by the asymmetrical distribution of phospholipids within the inner and outer leaflets. The inner leaflet of the bilayer mainly consists of negatively charged lipids, such as phosphatidylserine (PS), while electrically neutral lipids, such as phosphatidylcholine (PC) and phosphatidylethanolamine (PE), are mostly located in the outer layer [24]. At physiological pH, negative charge of the outer leaflet lipids is due to low values of acid dissociation constant (p*K*a) of the phosphate moieties of the lipid head group [25].

Our understanding of the properties, function and structure of biological membranes has benefitted a lot from the experimental and theoretical studies of the electrical properties of lipid bilayers. Therefore, evaluation of the parameters such as surface charge density (σ) of the membrane is critical

for determination of the electrostatic interactions between membranes and their surrounding solutes. The surface charge of a lipid membrane, which may change with pH, depends also on the lipids present in the outer layer and can be quantified by zeta-potential measurements using electrochemical light scattering (ELS) technique. Zeta potential depends on a lots of parameters, including temperature, pH, conductivity (ionic strength), and solvent (viscosity). Small changes in any of these parameters can potentially dramatically affect the zeta potential values [26].

In order to best mimic electrochemical conditions observed in natural membranes, model systems of artificial bilayers ought to possess similar values of capacitance (*C*m) and resistance (*R*m) to those observed in biological membranes. In this respect, the parallel plate capacitor model may serve as an estimator of the lipid bilayer capacitance. Assuming that typical hydrophobic thickness of a membrane equals 4 nm and a relative permittivity of ranges from 2 to 4, the capacitance of the bilayer is supposed to be 0.5–1.0 µF/cm<sup>2</sup> . In fact, *C*<sup>m</sup> values placing within this range have been already confirmed in experimental research [27–29]. Typical values of bilayer resistance are in the range 104−10<sup>7</sup> Ω cm<sup>2</sup> [30,31]. The measurement of lipid membrane resistance may sometimes be difficult and unreproducible. These variations are most likely a result of a leakage at the bilayer support and/or trapped emulsified droplets [32]. Luckily, membrane capacitance and resistance can be easily measured by electrochemical impedance spectroscopy (EIS). EIS is a technique that enables to extract the information not only about the bulk phase of tested material (e.g., dielectric constant and conductivity) but also about their inner and outer interfaces (e.g., interfacial region capacitance and derived quantities) [33]. μ − Ω

Herein, we report on the modulation of the electrical properties of model cell membranes by cinnamic acid and its two naturally occurring hydroxy derivatives: *p*-coumaric acid (*p*-CoA) and ferulic acid (FA). The selected phenolic acids are widely distributed in plants, fungi and algae and recognized as privileged structures for the development of bioactive compounds with therapeutic potential. These compounds are structurally related (Figure 1) and a correlation between their structures and behavior in the surrounding solution seems to warrant further investigations. Although a considerable amount of data concerning hydroxycinnamic acids interactions with membranes exists, still little is known about the electrical properties of bilayer lipid membranes modified by CinA, *p*-CoA or FA. Therefore, this paper is focused on the effect of cinnamic acid and its derivatives on the resistance, capacitance, and the surface charge density of model membranes (spherical lipid bilayers and liposomes).

**Figure 1.** Chemical structures of compounds under study. **Figure 1.** Chemical structures of compounds under study.

We previously described changes in the electrical parameters caused by e.g., membrane compositions in lipid–lipid [34], lipid–sterol [35], lipid–fatty acid [36], or lipid-carotenoid membranes [37]. The investigations reported in this paper are a continuation of the studies on the interaction of model membranes of increasing complexity with naturally occurring phenolic compounds. The bilayers were formed from 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), 1,2-diacyl-sn-glycero-3-phospho-l-serine (PS) or from DOPC-PS mixture in concentrations of 9:1 and 8:2 mol%, respectively, which corresponds to PS content in the human cerebral cortex [38]. Next, phospholipid membranes were modified by CinA, *p*-CoA or FA in the concentrations determined on the basis of MTT analysis shown in previous studies on human glioblastoma cell lines. Afterwards, we analyzed the influence of the pH of the electrolyte solution and the composition of membranes on their surface charge to better describe interactions in model membranes modified by the phenolic

acids. We also analyzed if CinA, *p*-CoA and FA were capable of altering the electrical resistance and capacitance of the bilayers. The results of our research should give useful indications in understanding the role of chemical structure of hydroxycinnamic acids in determining their interactions with the microenvironment of model lipid bilayer and, thereby, allow us to speculate about the in vivo bioavailability of the investigated compounds.
