Lipid nanodiscs of poly(styrene-alt-maleic acid) to enhance plant antioxidant extraction

1 Introduction

Plants produce a variety of bioactive metabolites for different purposes. Their primary metabolites, including nucleic acids, sugars, lipids, and amino acids, are essential compounds needed for plant growth and development. Secondary metabolites are not essential for the survival of plants, but they play an important role in plant defensive mechanisms and adaptation to fluctuating environments. Based on the chemical structure, plant secondary metabolites can be classified into three groups: terpenes, alkaloids, and phenolic compounds. Plant phenolics are composed of a hydroxyl functional group attached to an aromatic ring. They are produced by plants as a chemical defense against natural enemies. Due to their excellent electron donation, the compounds exhibit impressive antioxidant activity under both in vitro and in vivo conditions (1). Plant flavonoids, the largest group of naturally-occurring phenolic compounds, are one of the best antioxidant food supplements for human health. The global demand for plant flavonoids in 2024 was expected to reach USD 1.33 billion (2).

Extraction of plant antioxidants is generally performed in liquid solvents, either by maceration, soxhlet extraction, or heat reflux. The extraction efficiency depends on the increased mass transfer, solvent type, and energy input. Due to the toxicological concerns for human health and the environment, replacing organic solvents with less toxic media has always received considerable attention in the last few years. Water extraction is one of the most promising candidates for the extraction of plant phenolics. The method is readily available, cheap, recyclable, eco-friendly, and non-toxic. Nonetheless, the technique is quite limited in the case of moderate- or low-polarity compounds. To overcome this drawback, combined use of water and a surfactant, such as Triton-X, sodium dodecyl sulfate (SDS), cetyltrimethylammonium bromide, or poly(ethylene glycol), was suggested (3,4,5,6). One of the most promising models depicting surfactant-assisted extraction of cellular biomolecules was proposed by Kalipatnapu and Chattopadhyay (7) and Lichtenberg et al. (8). In this model, molecules of surfactants were believed to be adsorbed onto the outer layers of biological membranes. Depending on their concentration and hydrophilicity/hydrophobicity balance, they can further flip into the lipid membranes and create micro-pore defects in the bilayer structure. This step is usually accompanied by a slow leakage of cytoplasmic components, such as soluble proteins, mitochondria, chloroplasts, and plant secondary metabolites. Although surfactant-assisted extraction is cheap and convenient, it has been considered to exhibit complex dependency on different parameters, such as pH, energy input, extraction time, ionic strength, surfactant concentration, and lipophilicity values. With a large number of factors to be optimized, the technique becomes less accessible and inefficient (3). Furthermore, traditional micellar extraction is a time-consuming process with several steps to be undertaken. A microwave-assisted micellar extraction technique was thereafter developed to resolve the issues. Previous work revealed a shortened extraction time with improved extraction yields when using a combined technique (9,10). Nonetheless, long-time exposure to microwave radiation may induce structural changes and thermal decomposition in phenolic compounds, causing them to lose their biological functions.

Poly(styrene-alt-maleic anhydride) (PSMAnh) has emerged as a promising surface-active agent in cell membrane disruption. The polymer is normally synthesized by solution radical polymerization using an equimolar ratio of styrene (St) to maleic anhydride (MAnh). Due to the high reactivity of the anhydride ring toward nucleophilic reagents, PSMAnh can be hydrolyzed into poly(styrene-alt-maleic acid) (PSMA) or PSMA (Figure 1). Unlike PSMAnh, PSMA is water-soluble, making it more suitable for a green extraction process. One of the most interesting properties of PSMA is its ability to undergo a conformation transition upon pH adjustment. Above its acid dissociation constant value (pH > pK _a ∼ 4), the polymer chains are mostly ionized. The charged repulsion along the polymer backbone prevails over the styryl–styryl hydrophobic effects, causing PSMA molecules to adapt into an extended chain. As the solution pH is lowered (pH ≤ 4), the carboxylic acid groups of PSMA are fully protonated. The loss of charged repulsion allows the hydrophobic effects to become a dominant factor. As a result, the polymer chains are collapsed progressively into a distinct microdomain and eventually adapt to an α-helix coil configuration (11,12), as illustrated in Figure 1. Here, the hydrophobic styryl groups were believed to be located along the central core axis of a PSMA coil structure, while those of the hydrophilic carboxylic acid groups were allocated around the outer structure surface facing toward water molecules. Due to the amphipathic nature of the PSMA coil, it can act as a surface-active agent to destabilize cellular membranes and then induce a release of cytoplasmic contents. The ability of PSMA to undergo a conformation transition enables its surface activity to be tunable in response to a pH change.

Figure 1

Formation of an extended PSMA chain upon alkaline hydrolysis of poly(styrene-alt-maleic anhydride) precursor and its conformational transition into a surface active α-helix coil.

The biggest advantage of PSMA as a membrane lysis agent is its ability to combine with the reconstituted membrane lipid and then produce membrane-like nanostructures, as presented in Figure 2 (11,12,13,14). These nanostructures, so-called astosomes, nanodiscs, SMALP, or Lipodisq, were reported to have a size diameter of 10–40 nm (11,14,15,16,17). According to the molecular models proposed by Tanaka et al. (18) and Dörr et al. (19), this membrane mimetic system is composed of reconstituted lipid bilayers surrounded by the polymeric chains of PSMA. The central lipoid zone of the nanoassemblies (pointed by the arrows in Figure 2) can facilitate the direct incorporation of the active compounds into their core structures, thereby allowing the compounds to be stabilized and resolved in their full functioning state. The concept was successfully demonstrated for the extraction of membrane proteins from a variety of bacterial and eukaryotic sources, including Staphylococcus aureus (20), Escherichia coli (14,19,21), Rhodobacter sphaeroides (22), Pichia pastoris (23), and Saccharomyces cerevisiae (24). However, there has been a limited focus on plant extraction.

Figure 2

Polymer–lipid nanostructures with amphipathic α-helical PSMA coil arrangement around lipid bilayer.

This study provided new insights into the use of PSMA lysis agents in plant antioxidant extraction. The method was performed in aqueous media with no toxic organic solvents. The leaves of the robusta coffee tree were selected as a model plant due to their potential sources of natural antioxidants, such as caffeic acid, chlorogenic acid, p-coumaric acid, quercetin, kaempferol, catechin, and mangiferin (25,26,27). The extraction efficiency was evaluated on the basis of the total phenolic and flavonoid content, as well as the radical scavenging capacity. The results were compared with those extracted in deionized (DI) water, in both the absence and presence of the most commonly used surfactant, SDS.

2 Materials and methods

3 Results and discussion

Coffee leaf extracts have been used for centuries as an alternative medicine due to their diverse pharmacological effects, including anti-inflammation, antimicrobial, antidiabetic, antioxidant, and neuroprotective properties. The crude extracts have long been used to treat various human diseases, such as gastrointestinal, diabetes, obesity, cardiovascular, and dermatological conditions (30,31). Their phytochemical and therapeutic properties are related to phenolic constituents, which vary with geographic location, seasoning, plant species, and the age of leaves. The main active compounds have been identified as mangiferin, catechins, kaempferol, rutin, caffeic acid, hydroxycinnamic acid, coumaric acid, and chlorogenic acid (31,32). Due to their structural complexity and lipophilicity nature, the recovery yields of these compounds by using water extraction alone are quite low. This is consistent with the results in Figures 3 and 4 (control media). In the absence of SDS or PSMA, the total contents of phenolic acids and flavonoids are found to be as low as 5.8 and 3.2 mg CE·g⁻¹, respectively. On the addition of PSMA, approximately twofold and threefold increases in the recovery yields of phenolic acids and flavonoids are detected, respectively. The effects seem to be more effective than those of SDS, which is one of the most effective membrane lysis agents commonly used in both research and industrial applications. Interesting results may be obtained when comparing the DPPH antioxidant activity in all crude extracts. As noted in Figure 5 (blue bars), the crude extracts obtained by the PSMA extraction show the lowest averaged IC₅₀ values of around 3.7 µg CE·mL⁻¹, as compared to those by DI water extraction (9.5 µg CE·mL⁻¹) and SDS extraction (4.0 µg CE·mL⁻¹). The use of PSMA has thus proved its potential in recovering plant antioxidants with improved radical scavenging activity.

Figure 3

Comparison of the total phenolic acids in robusta coffee leave extracts obtained by different extracting systems. Mean values followed by the same letter in a bar chart plot are not significantly different at p < 0.05, according to the Duncan’s multiple range test.

Figure 4

Comparison of the total flavonoid contents in robusta coffee leave extracts obtained by different extracting systems. Mean values followed by the same letter in a bar chart plot are not significantly different at p < 0.05, according to the Duncan’s multiple range test.

Figure 5

Comparison of the DPPH scavenging assay IC₅₀ values in the coffee leave extracts by different extracting systems.

The excellent extractability of the PSMA lysis agent may be due to the formation of PSMA/lipid assemblies, as presented in Figure 2. The central lipoid zone of these structures is non-polar in nature. This can serve as a binding domain for low-polarity antioxidants to be encapsulated, thus promoting their aqueous solubility and the extraction yields, as illustrated in Figures 3 and 4. Encapsulation within these membrane-like structures protects plant active compounds from the external environments and allows them to be resolved in their fully functional state. The formation of such a biomimetic system was experimentally proved, for the first time, by Qiu and co-workers (33). Their main focus was to extract the AcrB multidrug exporter from the Gram-negative bacterial host. By using PSMA, this compound was proved to be transferred into the central cavity of the renatured lipid bilayer. Led by the cryo-EM results, this bilayer was composed of 24 lipid molecules, occupying the areas of 264 Ų (outer leaflet) and 400 Ų (inner leaflet). The intimate protein–lipid interactions enable these bilayer lipids to sense any change in the AcrB trimers. If there was any conformational change, the membrane lipids would harmonize the peristaltic motion along the peptide chains, thus driving the protein to fold back to its native structure. The process was considered to serve as a protection mechanism to stabilize the AcrB native structure, thereafter preserving its biological function during the extraction process. The concept has already been applied to various transmembrane proteins from different cellular membrane sources (7,11,12,14,15,17,22,23,34). However, there has been no report on PSMA-assisted extraction of plant bioactive compounds. The current study is the first to confirm that this concept can also be applied to plant phenolics. The improved antioxidant activity by the PSMA extraction presented in Figure 5 provides good supporting evidence on this matter.

An important question is whether these stabilization effects exist in the case of SDS lysis agents. The answer to this question is related to the membrane solubilization mechanism by SDS detergent. At room temperature, membrane solubilization by SDS is likely to proceed via the micelle extraction mechanism, while that by PSMA occurs through the transbilayer flip–flop mechanism (8,35). The difference is mainly because of the slower kinetics of flip–flop motion across the membrane bilayer of the former. In the beginning, the micelles of SDS are adsorbed onto the outer monolayer of the membrane bilayer. The nearly lipid molecules are detached (pinched off) and then associated with SDS micelles, producing the mixed SDS–lipid micelles at the end. Due to the favorable electrostatic attraction between the lipid and SDS headgroups (36), the biological functions of phospholipids are expected to be altered. As such, the aforementioned stabilization effects would become less effective. Consequently, direct exposure to the external environment would cause permanent damage to the encapsulated compounds, thus causing them to lose their bioactivities, as observed in Figure 5 (blue bars).

The extraction performance of PSMA to maintain plant antioxidants is considered to be affected by the salt species in the buffer systems (e.g., acetic acid, sodium, and acetate ions). This is convinced by the change in the IC₅₀ value from 3.7 µg CE·mL⁻¹ (DI water) to around 7.3 µg CE·mL⁻¹ (pH 4 buffer), 5.6 µg CE·mL⁻¹ (pH 5.6 buffer), and 6.3 µg CE·mL⁻¹ (pH 7 buffer). Due to the lack of hydrolysable functional groups along the polymer backbone, decomposition of PSMA is unlikely to occur during mild experimental DPPH testing (e.g., room temperature). Even at pH 4, this copolymer still showed excellent surface activity (34,37,38). To clarify the extent of PSMA degradation upon acidification (pH 4), we performed additional dynamic light scattering (DLS) experiments for PSMA itself and its complexation with the synthetic phospholipid (1,2-dimyristoyl-sn-glycero-phosphatidylcholine or DMPC). Under these experimental conditions, PSMA molecules display the ability to self-assemble into very fine and stable nanoparticles. These particles have an average size diameter of only around 18 nm with a polydispersity index (PDI) of 0.98 (data not shown). As previously mentioned, the particle formation is primarily driven by a conformational transition of PSMA from an extended chain to a compact coil. At pH 4, complexation with DMPC produces the end-state particles with a slightly increased size diameter of around 33 nm (data not shown). A unimodal size distribution (PDI = 0.22) suggests that the formation of these PSMA/DMPC nanoparticles is quite uniform with a low possibility of fusion or degradation. The effect of pH neutralization has already been demonstrated on a similar PSMA/lipid system (39). As the pH was increased from 4 to 7, the nanoparticles of PSMA and dilauroyl phosphorylcholine (DLPC) displayed an increased particle size, changing from 16 nm (PDI = 0.32) to 53 nm (PDI = 0.25). The narrower size distribution and the absence of a new size population suggested little sign of vesicular fusion or degradation. Led by these DLS results, the reduced antioxidant profile in the buffered systems compared to the DI water (Figure 5) is perhaps due to the greater molecular changes of the plant antioxidants inside the loosely packed nanoassemblies. Some buffer species may have competitively interacted with PSMA and/or the reconstituted lipids, creating void defects in the PSMA/lipid assemblies (Figure 2). In the presence of these defects, the encapsulated antioxidants would have been exposed to the surrounding environment (e.g., UV light, oxygen gas, chemicals). As such, they would have a higher risk of molecular decomposition and/or structural conversion into inactive forms. To avoid this, PSMA extraction should be performed in a buffer-free medium, such as DI water. Among the three acetate buffers, the crude extracts from the pH 4 system exhibit the weakest radical scavenging ability, as demonstrated by the highest IC₅₀ value of around 7.3 µg CE·mL⁻¹ (Figure 5). In the pH 4 solution (pH = pK _a), the molecules of PSMA are fully protonated, driving the polymer chains to adopt into a helical coil (Figure 1). In the absence of charged repulsion, a hydrophobically driven association with DMPC is supposed to be highly effective. As such, encapsulation of water-insoluble compounds into the deeper core region of the mixed assemblies can become most favorable. Chemical reactions with the DPPH reagent would thereafter be shielded to the greatest extent, resulting in the weakest inhibitory effect and so the highest IC₅₀ value, as demonstrated in Figure 5 (pH 4). The balanced shielding effects are therefore a crucial element to the successful design of these lipid analogues in practice.

4 Conclusion

The underexploited application of PSMA lysis agent for the recovery of coffee antioxidants directly into an aqueous formulation was demonstrated. In the absence of the charge shielding effects (e.g., buffer-free water), the crude extracts displayed the highest total contents of phenolic acids (11.6 mg GAE·g⁻¹) and flavonoids (9.6 mg CE·g⁻¹), compared to those obtained by traditional SDS and water extraction techniques. The PSMA-derived samples also showed enhanced DPPH antioxidant activity (IC₅₀ = 3.7 µg·mL⁻¹). The improved extractability was believed to be due to the formation of the membrane-like reservoirs that could facilitate a protective mechanism for those phytochemical antioxidants to be stabilized within the inner core structures, thus avoiding undesirable chemical reactions and the loss of their biological functions. The proposed biomimetic approach has shown the great potential to recover plant antioxidants for the benefits of human health, pharmaceutics, and biomedical uses in the future.