Peptide and Protein Emulsifiers

Abstract

There has recently been a growing attention towards peptide and protein molecules as potential bioemulsifiers for the stabilization of foams and emulsions, thanks to their innate tendency towards interfacial adsorption. Additionally, peptides and proteins are biodegradable and biocompatible, making them less toxic if compared to traditional emulsifiers. This chapter provides a comprehensive overview of the different classes of peptide, protein and mixed protein–polysaccharide emulsifiers and discusses the emulsification mechanisms of these systems. In essence, peptide-mediated emulsification can occur either via traditional surfactant-like mechanism, where amphiphilic molecular peptide chains adsorb at the biphasic interface forming ‘spherical micelles’, or through peptide self-assembly into higher secondary structure (α-helices or β-sheets) with the formation of amphiphilic nanofibrous structures adsorbing at the interface. Moreover, peptides can self-assemble in the continuous aqueous phase forming nanofibrous network of viscous hydrogels that enhance system stability. On the other hand, emulsion stabilization by proteins is mainly achieved through either electrostatic repulsion or steric stabilization. The various characterization techniques for emulsification and interfacial stabilization will be visited throughout this chapter, focusing on structural, mesoscopic and macroscopic characterization of these systems.

13.1 Introduction

Emulsifiers have been widely developed and exploited in many industries, such as wastewater treatment, improved oil recovery, food and beverages, detergents, cosmetics as well as coating industries, in addition to their utilization in chemical catalysis and cell assays (Moreira et al. 2017; Dexter et al. 2006; Lv et al. 2019; Dexter 2010; Bai et al. 2014; Platzman et al. 2013). Importantly, they have been used in various pharmaceutical formulations for drug encapsulation and controlled drug delivery (Dexter et al. 2006; Nishida et al. 2017; Windbergs et al. 2013; Naseef et al. 2018). Emulsifiers are required for controlling the characteristics of fluid–fluid interface, which is highly crucial for stabilization of foam and emulsion-based formulations (Dexter et al. 2006; Dexter 2010; Wilde 2000). Long-term stabilization remains as a pivotal challenge for these industries when developing such colloidal systems, which are instinctively thermodynamically unstable and vulnerable to phase separation (Li et al. 2016). Therefore, there is a crucial need for the development of emulsifiers that can enhance long-term stabilization while being biocompatible, non-toxic, sustainable and environment friendly (Li et al. 2016).

In essence, emulsifiers are a distinctive group of surfactants that can efficiently adsorb at different interfaces, particularly air–liquid and liquid–liquid, reducing interfacial tension and thus allowing for the dispersion of immiscible phases and formation of kinetically stable colloidal systems (Bouyer et al. 2012; Lee 2008; Mondal et al. 2017). According to Myers, emulsifiers can be classified into four main classes, based on their interfacial adsorption and emulsification mechanisms (Myers 1999). The first class includes non-surfactant ionic compounds capable of interfacial adsorption at drop surface resulting in electrostatic repulsions between close droplets, thereby stabilizing emulsions against coalescence. However, these ionic compounds neither impact the interfacial tension nor contribute to the emulsification steps. The second category comprises small non-surfactant colloidal solids, such as clay or silica, forming Pickering emulsions, as they adhere to droplet surface generating a physical barrier between globules that delays or prohibits coalescence. The third group encompasses the typical monomeric surfactant amphiphiles. They are composed of a charged/hydrophilic head conjugated to a hydrophobic tail. According to the inherent charge of the head part, they can be subdivided into anionic, cationic, zwitterionic or non-ionic molecules (Lee 2008; Caballero et al. 2003). At the interface, they adsorb and self-assemble, directing the hydrophilic head to the aqueous phase, whereas the hydrophobic tail towards the oil or gas phase forming protective layer preventing breakdown and lowering interfacial tension between phases, and consequently stabilizing the system (Petsev 2004). Finally, the fourth class contains polymer surfactants that impart steric and/or electrostatic stabilization to the system besides their interfacial tension diminishing capacity. They can also form cohesive interfacial films which inhibit the rupture of thin layers of emulsifier surrounding the droplets or bubbles, which is the last stage in coalescence (Dexter et al. 2006; van Aken 2004). Moreover, they improve emulsion stability through enhancing elasticity and viscosity of the interface or by changing the bulk viscosity of the systems.

It can be noticed that interfacial adsorption and self-assembly is a common characteristic among the four classes of emulsifiers. Interfacial self-assembly is governed by mutual interactions between the emulsifier units and the interface itself. These interactions can also affect intermolecular forces between assembled emulsifier units and thus make the interfacial assembly significantly different from bulk self-assembly (Lee 2008). Better understanding of interfacial self-assembly processes can inform the rational design of emulsifier units that are capable of self-assembly at interfaces creating interfacial nanoarchitectures which can inhibit coalescence and coarsening (Scott et al. 2016).

Besides the molecular nature of emulsifiers, emulsions can be classified, according to the proportion and distribution of aqueous to oil phase in the system, into either simple or double emulsions. Simple emulsion may be either oil in water (O/W) when the oil is the dispersed phase and water constitutes the continuous phase or water in oil (W/O) which is the reverse state. Double emulsions are W/O/W or O/W/O systems offering merits over simple emulsions for encapsulation of therapeutic molecules and their controlled release due to the enveloped internal framework (Okochi and Nakano 2000; Hanson et al. 2008). The type of emulsion created, either O/W or W/O, depends on various intrinsic parameters including physicochemical properties of the emulsifier used, temperature and composition of the system. These parameters could affect the competition between coalescence tendency and emulsified droplet stabilization. In general, the highly stabilized phase will constitute the dispersed droplets, while the other phase coalesces forming the continuous phase (Lee 2008). Extrinsic manufacturing parameters can also affect emulsion order, such as the type of energy used (stirring), the way and time of its application as well as the order of ingredients mixing. Based on globular size and thermodynamic behaviour, emulsions can be classified into microemulsions (10–100 nm, thermodynamically stable), nanoemulsions (100–400 nm, kinetically stable) and macroemulsions (>400 nm, kinetically stable) (McClements 2012; Callender et al. 2017). The miniaturized size of nanoemulsions/microemulsions droplets shows several privileges over macroemulsions involving improved optical clarity, enhancement of drug cargo bioavailability and long-term formulation stability, which are all key attributes for drug delivery vehicle (Table 13.1) (Yang et al. 2017).

Table 13.1 Emulsion types and their properties

Although traditional emulsifiers are well suited for formation and stabilization of foams and emulsions, many of these emulsifiers are neither biodegradable nor biocompatible (Scott et al. 2016). Additionally, some have potential toxicity towards human health and the environment (Bouyer et al. 2012; Mondal et al. 2017; Rebello et al. 2014). Other limitations include impaired stabilization under environmental stresses, such as elevated temperatures, extreme pH ranges and high salt concentrations, as well as the insufficient long-term stability and not being amenable for structural modifications, all hindering their practical use in various industries (Adjonu et al. 2014; Fowler et al. 2011; Minkenberg et al. 2009; Lotfallah et al. 2015). Peptides and proteins came in the forefront as potential alternative emulsifiers that can overcome these limitations, thanks to the innate surface activity, biocompatibility and physicochemical tunability.

In this chapter, we will shed the light on recent advances in formulation of emulsions and foams using peptide, protein and mixed protein/polysaccharide emulsifiers. The various mechanisms of emulsification and interfacial stabilization will be discussed in terms of their structural and mesoscopic characteristics and their stabilization efficiency for longer term and under drastic conditions.

13.2 Peptide Emulsifiers

Peptide molecules possess the inherent characteristics of proteins, such as the biodegradability, biocompatibility and safety with additional advantages of low production costs with high purity and low batch-to-batch variations (Moreira et al. 2017; Li et al. 2016; Wibowo et al. 2017). In addition, exploitation of the amino acid chemical toolbox provides infinite possibilities of various chemical designs, enabling physicochemical tunability of peptide emulsifiers, for instance, tailoring hydrophilic and lipophilic properties and controlling the non-covalent electrostatic or hydrophobic interactions between molecules, which are essential for controlling interfacial self-assembly (Wibowo et al. 2017). Peptides can rapidly adsorb to interfaces at lower concentrations compared to proteins, facilitating droplet or bubble formation which is the most critical step for emulsion or foam formation, respectively. This is normally followed by interfacial adsorption into rigid cohesive structures surrounding the droplets or bubbles maintaining their attachment and stabilization of these system (Scott et al. 2016).

Interfacial adsorption of peptide molecules can lead to the formation of micellar structures, akin to conventional surfactant, where the hydrophilic heads are solubilized in aqueous phase and the hydrophobic tail projecting to the organic or gas phase (Fig. 13.1). Interfacial self-assembly into higher secondary structure (α-helices or β-sheets) could also happen, forming amphiphilic nanofibrous structure at the interface (Fig. 13.1). Additionally, self-assembly into nanofibrous network in the bulk aqueous phase could occur, leading to hydrogel formation, which increases the system’s viscosity and hinders emulsion coalescence or foam collapse, a system known as ‘emulgel’ (Fig. 13.1).

Fig. 13.1

The possible mechanisms of interfacial arrangement of peptide emulsifiers at the droplet interface through either micelle formation or self-assembly into nanofibrous structures, forming emulsions or emulgels

Based on their chemical structures, peptide emulsifiers can be classified into short aromatic, α-helix, β-sheet and surfactant-like peptides. In this section we will present these different classes and will discuss their interfacial adsorption mechanisms in more details.

13.2.1 Short Aromatic Peptide Emulsifiers

Short synthetic peptides were developed to self-assemble into a variety of higher nanostructures, as discussed in Chap. 4: short peptide and peptidomimetic nanomaterials. Interestingly, short aromatic peptides have been designed to self-assemble into amphiphilic nanofibrous structures which exhibit interfacial activity. For instance, the Ulijn group reported the ability of amphiphilic dipeptides, which are N-capped with aromatic moieties, to self-assemble into nanofibrous networks at organic/aqueous interfaces (Fig. 13.2a) (Bai et al. 2014). Unlike traditional emulsifiers, which work via adsorption of the molecular amphiphile at the biphasic interface forming micellar structures, these short aromatic peptides form interfacial nanofibrous stabilization matrices (Figs. 13.1 and 13.2a). The 9-fluorenylmethoxycarbonyl tyrosine-leucine (Fmoc-YL) is one example of these short aromatic peptide emulsifiers, which was successfully used to stabilize chloroform-in-water emulsion by hand shaking for few seconds at 80 °C (Table 13.2). Interfacial self-assembly of Fmoc-YL into a less ordered β-sheet-like assemblies occurred via interchain hydrogen bonding (H-bonding) between peptide backbone amides as well as aromatic π stacking of Fmoc capping groups, forming amphiphilic nanofibrous structures, which entangle into interfacial network film stabilizing the suspended chloroform droplets within the aqueous continuous phase (Figs. 13.1, 13.2a, g and i). The Fmoc-YL-based emulsion exhibited greater stability (months) at room temperature than the commercial emulsifier sodium dodecyl sulphate (SDS), with the latter showing phase separation within 2 weeks. In addition, better thermal stability as well as resistance to salting out were achieved with Fmoc-YL in comparison to SDS emulsifier.

Fig. 13.2

Fluorescent microscope images of chloroform-in-water emulsion droplets stabilized by (a) Fmoc-YL, (b) Fmoc-YA, (c) Fmoc-YS, (d) Fmoc-FF networks containing fluorescence isothiocyanate (FITC) in water phase, (e) water-in-chloroform emulsion droplets stabilized by Fmoc-FFF containing FITC in water phase and (f) Fluorescence microscopy of KYW chloroform-in-water emulsion droplets with overlay of Sudan II and Thioflavin T (ThT) staining, (g) Fmoc-YL and (h) Pyrene-YL emulsions. Scale bar is 50 μm except for KYW (10 μm). (i) SEM micrograph of air-dried Fmoc-YL microcapsules at chloroform/water interface. Scale bar is 2 μm. Adapted from (Bai et al. 2014), with copyrights permission from the American Chemical Society and (Scott et al. 2016), with copyrights permission from John Wiley and Sons. (j) left. Chemical structure of Fmoc-YpL and Fmoc-YL, Right, Schematic representation of the Fmoc-YpL before and after alkaline phosphatase dephosphorylation in a biphasic system, showing Fmoc-YL stabilized emulsions, while Fmoc-YpL acted like a conventional surfactant and system separated into two phases after 1 h. Adapted from (Moreira et al. 2016), with copyrights permission from Royal Society of Chemistry

Table 13.2 Classes of peptide emulsifiers and their application

On-demand demulsification of Fmoc-YL systems was made possible using thermolysin protease enzyme, which digested the interfacial peptidic film under mild physiological conditions. Fmoc-YL also showed to be a successful emulsifier for other organic phases, such as hexadecane and mineral oil. Replacement of L residue in Fmoc-YL with A or S also produced peptide emulsifiers that formed chloroform-in-water emulsions. However, these changes significantly affected the emulsion characteristics, such as globular size and critical emulsion concentration. In case of Fmoc-YA, reduction of sequence hydrophobicity arising from replacing L by A did not influence β-sheet nanofibres formation in water. However, introduction of S impaired the peptide tendency to self-assemble into nanofibrous structures; consequently, higher critical emulsion concentration and emulsion globular size were observed for Fmoc-YS compared to both Fmoc-YL and Fmoc-YA (Table 13.2 and Fig. 13.2a–c). In essence, increasing both hydrophobicity and H-bonding interactions enhanced interfacial adsorption and lowered the emulsion droplet sizes as well as critical emulsion concentrations. To increase sequence lipophilicity, Fmoc was replaced by pyrene or YL by FF, where the Fmoc-FF and Pyr-YL successfully created chloroform-in-water emulsions (Fig. 13.2d, h). Further increase in lipophilicity resulted in inverted emulsion (water-in-chloroform), when using the highly hydrophobic tripeptide Fmoc-FFF as emulsifier (Fig. 13.2e and Table 13.2).

Ulijn and co-workers have also introduced biocatalytic-controlled interfacial self-assembly of phosphorylated Fmoc-YL (Fmoc-YpL) at chloroform/water interface using a phosphatase enzyme (Fig. 13.2j and Table 13.2) (Moreira et al. 2016), inspired by Xu’s group strategy who first demonstrated that enzymatic dephosphorylation of Fmoc-tyrosine phosphate (Fmoc-Yp), using alkaline phosphatase, triggered self-assembly of the product Fmoc-Y into supramolecular hydrogel (Yang et al. 2004). Ulijn and co-workers demonstrated that the parallel arrangement of fluorenyl groups of Fmoc-YpL led to the formation of micelles, which when dephosphorylated by alkaline phosphatase into Fmoc-YL rearranged into fibrous structures at the organic/aqueous interface (Fig. 13.2j). The phosphorylated precursor Fmoc-YpL behaved as an amphiphile with surfactant-like adsorption, where the hydrophobic Fmoc tail projected to organic phase while the hydrophilic YpL chain head to water with low emulsion stabilization, with phase separation occurring within 1 h. Upon enzymatic dephosphorylation into Fmoc-YL, interfacial self-assembly into nanofibrous network was achieved via non-covalent interactions, as explained above, forming a very stable emulsion with no phase separation observed for months. The same group showed the possible conversion of a phase separated biphasic oil/water mixture containing Fmoc-YpL stored for up to 1 month into emulsion catalysed by enzymatic dephosphorylation. However, slower emulsification rate was observed compared to the immediate activation approach, with the formed emulsion exhibiting less stability and larger droplet size. This is attributed to the slower dephosphorylation of Fmoc-YpL leading to less ordered H-bonding, hence impairing the interfacial self-assembled fibrous network. Molecular dynamic simulation suggested the absence of permanent H-bonding between Fmoc-YpL monomers, while the dephosphorylated form showed Fmoc/L and Y/L H-bonding that agreed with FTIR results. Results of this study confirmed the importance of interfacial self-assembly of nanofibres for emulsion stabilization rather than individual peptide monomers (Fig. 13.1) (Moreira et al. 2016).

Although the Fmoc-capped dipeptides exhibit good surface activity, which was shown to be better than commercial emulsifiers in many cases, its potential translation to emulsifiers in food, pharmaceutical and cosmetic industries was withheld as the molecular design is based on the non-biological Fmoc group, which could potentially impair the biocompatibility of these materials. To overcome this limitation, Ulijn and co-workers developed unprotected tripeptide emulsifiers, with a peptide sequence design of two successive aromatic residues (Ar) flanked with a charged residue (X^+/−) either from the N- or C-terminus (i.e. X^+/−-Ar-Ar or Ar-Ar-X^+/−) (Scott et al. 2016). Two types of tripeptide emulsifiers were developed, the anionic (such as FFD and DFF), which formed bilayer-like assemblies at the oil/water interface, and the cationic (such as KFF, KYW and KYF), which formed interfacial nanofibrous networks that highly stabilized the rapeseed oil-based emulsions compared to the anionic type (Table 13.2). Interestingly, the cationic tripeptides self-assembled in both aqueous and biphasic media through H-bonding into β-stranded fibres, while the anionic tripeptides only showed interfacial self-assembly upon introduction of oil to the aqueous phase, where peptide chains stack into parallel orientation at the oil/water interface with lateral H-bonding interactions. The nanofibre forming peptide emulsifier KYF possessed superior thermal stability with no phase separation observed for the formed O/W emulsion at 80 °C, while KYW and KFF showed thermal stability only up to 50 °C and 40 °C, respectively, due to the breakdown of fibrous structure. Based on this, KYF was exploited as an emulsifier to formulate O/W nanoemulsion that can entrap oleic acid-platinum II conjugate for treatment of ovarian cancer (Dragulska et al. 2018). A nanoprecipitation method was employed by a dropwise addition of organic phase (oleic acid-platinum II conjugate in isopropanol) to aqueous phase (KYF dissolved in water at 37 °C and pH 7) with stirring overnight to prepare the nanoemulsion. The miniaturized droplet size (240 nm) enabled its ability for extravasation and passive targeting of tumours with long-term stability upon storage for months. This strategy can be employed to encapsulate and deliver other active moieties.

Besides nanoemulsion formation, Moreira et al. demonstrated the ability of the phosphorylated tripeptide analogue KYpF to act as on-demand emulsifier that stabilizes O/W emulsion and emulgel upon treatment with alkaline phosphatase (Table 13.2) (Moreira et al. 2017). Complete dephosphorylation occurred 24 hours after incubation with phosphatase enabling the transformation of a biphasic system containing few micelles of KYpF at the interface (40 mM) into an O/W emulsion at low enzyme concentration and O/W emulgel at high enzyme concentration where both were stabilized by the dephosphorylated KYF. Emulsion stabilization was achieved via creation of β-sheet nanostructures at the oil droplet surfaces lowering interfacial tension, while emulgel additionally has crosslinked β-sheet nanofibres in the aqueous phase that enhanced the continuous phase viscosity. Enzyme concentration was found to be a crucial factor in controlling the emulsifying capacity of the peptide, as it affects the rate of self-assembly and densities of fibrous networks at the interface hindering coalescence. Increasing enzyme concentration led to faster dephosphorylation, higher levels of π–π stacking, more organized H-bonding between backbone amide of peptide chains, and overall better self-assembly and control of emulgel properties. The lowest enzyme concentration tested (0.07 μM) did not form emulgel but stabilized the emulsion through short un-entangled thick fibres (14 nm width), while increased entanglement and dense thin fibres (3 nm) appeared with the highest concentration (6.6 μM), stabilizing the system for more than 1 week. On the other hand, there was no evidence of secondary structure formation of KYpF in the absence of enzyme. The failure of KYpF assembly was ascribed to the hydration of phosphate ions in water rather than the repulsion between the anionic groups, so modifying the peptide hydrophilicity can affect self-assembly. The best catalyst concentration of the enzyme was 3.3 μM that produced sharper peaks at 1620 and 1560 cm⁻¹ in FTIR. Interestingly, the biocatalytic emulsion formation was thermo-reversible displaying switchability between demulsification and re-emulsification upon heating to 50–60 °C then cooling overnight, respectively. This study also showed that controlled biocatalytic activation of KYpF gave more ordered structures than the use of KYF peptide per se, with less entangled slightly thicker fibres (~4.6 nm) formed.

Another triggering approach for self-assembly of the diphenylalanine-based tripeptide emulsifiers FFX (where X is a hydrophilic residue, such as S, F, D or H) that failed to form nanofibres at the oil/water interface (Scott et al. 2016), involved the conjugation of the organometallic moiety ferrocene (Fc) to the tripeptide sequence (Table 13.2) (Yang et al. 2017). The biocompatible moiety increased peptides hydrophobicity as well as steric bulk and imparted redox power to the peptide system. Taking FFH as example, the peptide failed to stabilize emulsion formation; however, the amphiphilic Fc-FFH conjugate monomers gathered at the ethyl acetate/phosphate buffer (pH 7.2) interface into thin shells (Yang et al. 2017). This orientation stabilized the fabricated nanoemulsion, which is composed of hollow nanovesicles (100–200 nm), for more than 4 months at ambient conditions and exhibited high thermostability at 70 °C. This perfect stabilization efficiency was due to an enthalpy-controlled equilibrium among the association energy within the self-assembled peptide structure and the interfacial free energy between the peptide monomers and solvent. Other Fc-FFX sequences also successfully stabilized the nanoemulsions (Yang et al. 2017). Interestingly, tailoring the nanoemulsion droplet size, its physical appearance and its redox, as well as catalytic activity, was performed through adjusting the ratio of solvents, temperature, oxidation state of Fc group and the sequence of tripeptide. Moreover, oxidation of the ferrocene peptide-based nanoemulsion led to emergence of small micellar structures that can be employed as drug delivery vehicle. Increasing the water to organic solvent ratio led to conversion of emulsion size from microscale to nanoscale with transparent appearance. Rising the temperature of the systems to 70 °C did not cause phase separation, but only increased droplet size was observed. However, upon cooling below 25 °C, Fc-FFH emulsion showed transition into hydrogel due to reduction of monomers’ aqueous solubility after cooling and increasing the association energy, which led to change of peptide conformation. Fc-FFH monomers reorganized themselves or dissociated from the interface followed by their assembly into twisted β-sheet fibres and then nanostructured hydrogels, where the FTIR of nanoemulsion showed an unstructured system (Peak at 1654 cm⁻¹), while hydrogel had a characteristic β-sheet peak at 1629 cm⁻¹. It has also been reported that His-based peptides can perform like an artificial hydrolase enzyme (Garcia et al. 2017; Wang et al. 2016; Huang et al. 2013; Guler and Stupp 2007). Based on this, Yang et al. found that Fc-FFH nanoemulsion was able to hydrolyse p-nitrophenyl butyrate (PNPB) into PNP, so Fc-FFH nanoemulsion also acted as an ideal artificial hydrolase enzyme with better storage stability than enzymes (Yang et al. 2017).

Other N-capped short aromatic peptides showed to stabilize foam systems, where Li et al. reported that naphthyl derivatives of the short aromatic dipeptide FF (NapFF and BrNapFF) self-assemble into thin interfacial films of β-sheet fibrous structure at the air/water interface using drop-casting technique (Table 13.2 and Fig. 13.3a–c) (Li et al. 2014). The peptide was dispersed in alkaline solvent (pH >10) where wormlike micellar structure formed followed by its dropwise addition onto a low pH solvent, leading to formation of woven strands of β-sheets configuration. The creation of interfacial self-assembled film required high peptide concentration (>0.35 wt%) besides quick protonation of peptide achieved by adjusting the subphase pH to be less than the pKa of the dipeptides, which was 3.5 and 2 for NapFF and BrNapFF, respectively. The β-sheet strands were in nanoscale width (40 nm) and when carboxylate group become protonated, bundled together to generate elastic fibres of 800 nm in thickness. They displayed an episodic wrinkle when compressed in a Langmuir trough conferring stabilization of large air bubbles for numerous days, with no coalescence or Ostwald ripening was observed. On the other hand, BrNapAV failed to build interfacial films at any concentration and different pH values although it exhibited β-sheet arrangement at high pH. This might be ascribed to the type of micelles (spherical micelles) formed at high pH differed from that formed with NapFF and BrNapFF (wormlike micelles) or the lower hydrophobicity of BrNapAV.

Fig. 13.3

Stabilization of foam using different NapFF concentrations, (a) 8 mg/mL, left, after adding CaCl₂, Right, without CaCl₂. (b) 5 and 2 mg/mL. (c) Confocal images of the air bubbles in foam showing non-spherical shapes upon bubbles distortion. Adapted from (Li et al. 2016), with copyrights permission from John Wiley and Sons. (d) Characteristics of four emulsions stabilized by 2 mM 2NapFF at 50 °C and the dipeptide were dyed with Nile blue in the confocal images. Adapted from (Aviño et al. 2017), with copyrights permission from Royal Society of Chemistry

The same group used the hydrophobic dipeptide NapFF (0.8 wt%) to formulate self-assembled viscoelastic hydrogel in presence of Ca²⁺ ions that can generate wet foams and stabilize them for more than 2 weeks (Table 13.2 and Fig. 13.3a) (Li et al. 2016). Due to interchain H-bonding, hydrophobicity of NapFF as well as external stimulation by cationic metals, the aromatic dipeptide self-assembled at the air/water interface into β-sheet oriented fibrous network films and stabilize the foam bubbles. The formed fibre (200 nm width) at the air/water interface (Li et al. 2016) was found to be thicker than fibres obtained from the drop casting method (40 nm width) (Li et al. 2014). In addition, the extra dipeptide moieties self-assembled in the bulk continuous phase forming a 3D structured hydrogel with high storage modulus (Li et al. 2016). Such viscoelastic hydrogel could arrest the migration of bubbles, and hence delay bubble size growth, as well as lower the liquid drainage out of the system pointing out the great stabilization of the formed foam. However, upon foam cessation, part of the aqueous phase separated from the hydrogel (Fig. 13.3a). On the other hand, the NapFF solution that was not treated with Ca²⁺ ions did not form a gel and so produced a very weak foam that ceased in less than 2 h (Fig. 13.3a) (Li et al. 2016).

Interestingly, system stabilization was shown to be concentration dependent, which might be enhanced by increasing the dipeptide concentration. For instance, 0.2 wt% NapFF formed thin fibres that collapsed in 7 days, while 0.5 wt% led to better stabilization where water leakage out of the system, reduction in foam volume and phase separation happened within 2 weeks (Fig. 13.3b, c) (Li et al. 2016). However, increasing peptide concentration to 0.8 wt% formed stable wet foam hydrogel system for more than 2 weeks (Fig. 13.3a) (Li et al. 2016). Moreover, the type of metal cation affected foam stabilization, where divalent cations can develop stiffer hydrogels than monovalent ones (Li et al. 2016; Chen et al. 2011a). In the view of this, it was found that foam induced by CaCl₂ addition was more stable (no coalescence) than MgSO₄/MgCl₂ followed by KCl that showed higher bubbles coarsening rate due to the absence of crosslinked fibres and presence of thin fibres 10 nm (Li et al. 2016).

Aviño et al. used extremely low concentration of 2NapFF (0.1 wt%) with two different MgSO₄ concentrations (18 and 142 mM) to induce self-assembled hydrogel for stabilizing foam and emulsions of different oils (Table 13.2) (Aviño et al. 2017). Micellar structure was formed after dissolving the peptide at high pH (11 ± 0.5) that converted to hydrogel after adding salt. The fibrous structure formed at the air/water interface upon adding salt to low dipeptide concentration was very loose that failed to hinder bubbles ripening leading to a weakly stabilized foam (Aviño et al. 2017). At low salt concentration, better foam quality was formed compared to higher salt concentrations, but both foams collapsed only after 2 h due to the preferential interfacial adsorption of the nanofibres leading to the weakness of the hydrogel in the bulk phase. A different behaviour was observed at the oil/water interface, where significantly more stable long-standing emulsions were produced with both MgSO₄ salt concentrations upon testing three different oils, isopropyl myristate, dodecane and silicone oils, while only octanol failed to be emulsified due to peptide accumulation in the oil and poor wetting of the interface (Fig. 13.3d). The isopropyl myristate-based emulsion was the optimum formulation with no obvious creaming due to the formed interfacial film that greatly stabilized the system. However, in case of silicone and dodecane oil-based emulsions, stabilization was only attributed to hydrogel formation in the continuous phase, as well as wetting properties of peptide at the interface. To that end, it could be concluded that gelation of naphthyl derivatives of short aromatic peptides is possible by dissolving the peptide in an alkaline medium then adding salts or lowering the pH.

13.2.2 α-Helix Peptide Emulsifiers

Bioinspired short amphiphilic peptide sequences that have tendency to acquire α-helical structures at organic/aqueous interfaces have been widely studied for their surface activity and emulsification capacity. For instance, the surface activity, conformational rigidity and helical propensity of the fungal structural hydrophobin proteins inspired the development of a library of short biotriggered helical emulsifiers, by Gazit and co-workers (Mondal et al. 2017). The helical heptad amphipathic peptides showed to stabilize silicon oil-based O/W emulsions. The design was based on the incorporation of the helical-forming non-proteinogenic amino acid, α-aminoisobutyric acid (Aib), at positions (i, i + 3), which enabled the formation of supramolecular helical structure both in aqueous phase and at oil/water interface (Fig. 13.4a). The parent peptide of this library was designed by modifying the previously reported helix-forming peptide, SHR-FF peptide (Ser-Aib-Phe-Ser-Aib-Phe-Aib) (Mondal et al. 2015), into SHR-FLLF (Phe-Aib-Leu-Ala-Aib-Leu-Phe) (Table 13.2). The terminal Phe groups of one strand allowed the π–π stacking with another adjacent one, while Leu inclusion offered zipper conformation via hydrophobic bonding in addition to increasing hydrophobicity by adding alanine instead of serine (Fig. 13.4a) (Mondal et al. 2017). Such structural modulation favoured the self-assembly into highly ordered supramolecular aggregates of helices, forming cylindrical micelles-like nanofibres (Fig. 13.4a) (Mondal et al. 2017). It was postulated the presence of 2 asymmetric structural units: (A) right-handed α-helix and (B) 3₁₀-helix, due to intrahelical H-bonding, with perpendicular orientation of unit A to B triggering hydrophobic and π stacking interactions forming knob-into-hole arrays (Fig. 13.4b) (Mondal et al. 2017). This arrangement created a helical hydrophobic column with dual hydrophilic terminals similar to bolaamphiphiles. SHR-FLLF showed to act as a biosurfactant at concentrations of 5, 10 and 15 mg/mL (critical assembly concentration, CAC = 2.8 mg/mL), emulsifying silicone oil in water (pH <2) forming O/W emulsions. The lowest peptide concentration (5 mg/mL) demonstrated stability for >1 week which was comparable to the commercial anionic surfactant sodium dodecyl sulphate (SDS), whereas higher concentrations (10 and 15 mg/mL) showed at least 2-month stability and some formulations did not show any signs of phase separation for up to 8 months. SHR-FLLF-based emulsions displayed highly charged droplets with similar emulsion properties to SDS due to the helical arrangement at the interface, in addition to interfacial self-assembly ensuring its strong surface activity. Additionally, emulsions formed at 10 and 15 mg/mL peptide concentrations demonstrated viscoelastic characteristics preventing phase separation, due to high viscosity of the formulations. The Tween 20-emulsified system showed larger droplet size than the SHR-FLLF peptide-based emulsions, leading to creaming and phase separation in 48 h. The replacement of alanine with a charged amino acid either glutamic acid or lysine (SHR-FLELF and SHR-FLKLF) was examined, which aimed to break the bolaamphiphilic behaviour while retaining helical propensity. Both charged analogues acquired the α- and 3₁₀ helical structures in water, where SHR-FLELF self-organized forming a micellar arrangement, while SHR-FLKLF stayed in unaggregated form. Consequently, SHR-FLKLF (5 and 7.5 mg/mL) stabilized the emulsion for less than a week, while the SHR-FLELF-based emulsion showed longer stability (>2 months at concentrations of 7.5 and 10 mg/mL) with viscoelastic shear-thinning behaviour.

Fig. 13.4

(a) Helical wheel representation and (b) perpendicular packing pattern of SHR-FLLF

A bioinspired switchable amphiphilic helical peptide emulsifier, Lac21, was also designed by the Middelberg’s group, which mimic the bacterial Lac repressor DNA-binding protein (Dexter et al. 2006). The Lac21 amphiphilic peptide (Ac-MKQLADSLMQLARQVSRLESA-CONH₂) is a hydrophilic heptad repeating motif abcdefg containing hydrophobic amino acids at the a and d positions spaced by alternating 3 and 4 amino acid residues of the bacterial Lac repressor protein sequence, with acetylated and amidated C- and N-terminals, respectively, to prevent interpeptide charge–charge interactions (Fig. 13.5) (Dexter et al. 2006). Lac21 did not form organized structure in water; however, it could adsorb and adopt a supercoiled α-helix structure at organic/aqueous and air/aqueous interfaces in a monomeric free “detergent” state reducing interfacial tension, in a similar behaviour to low molecular weight surfactants (Fig. 13.5) (Dexter et al. 2006). Despite the interfacial adsorption and helical structure formation, Lac21 failed to stabilize foams and emulsions as it lacks the ability to form interfacial cohesive film, although it is derived from protein that could form interfacial films. Interestingly though, Lac21 peptide was used as a parent sequence for the design of a library of peptides exhibiting robust intermolecular crosslinking at the interface generating stable cohesive interfacial films.

Fig. 13.5

Self-assembly of α-helix-forming peptides (Lac21 and its derivatives) in bulk phase and at interface

One example of these peptides is the 21-residue peptide AM1, which is a Lac21 sequence mutated with histidine (H) residues at positions 9 and 20 (Ac-MKQLADSLHQLARQVSRLEHA-CONH₂). AM1 showed to form and stabilize foams as well as 20%v/v toluene/water emulsions upon changing the composition of the continuous aqueous media using metal cations and by tuning pH to 7.4 (Table 13.2) (Dexter et al. 2006). This peptide adsorbed at fluid–fluid interface acquiring α-helical structure in a detergent mobile state where the hydrophobic moieties were directed towards the interface exhibiting interfacial tension lowering property like Lac21 (Fig. 13.5) (Dexter et al. 2006). However, the interfacial tension reduction due to AM1 occurred more rapidly than Lac21, which is attributed to the less organized structure of AM1 (random coil monomers) in bulk phase due to replacing the high helix-forming methionine (M) at position 9 in Lac21 with low helix-forming H residue. Such modification decreased the energy barrier against the adsorption of AM1 and led to its arrangement in α-helical structure at the interface. Interestingly, the H residues at positions 9 and 20 were oriented towards the bulk hydrophilic phase (Fig. 13.5) (Dexter et al. 2006). These residues bind to metal ions (100 μM Zn²⁺) at neutral pH allowing intermolecular crosslinking of the peptides located at the interface without metal binding in the bulk (Fig. 13.5) (Dexter et al. 2006). The interfacial assembly of AM1 developed a monolayer of a cohesive mechanically rigid film with elasticity modulus 121 mN/m comparable to the protein’s interfacial film, and no major effect on the lowered interfacial tension was recorded after metals. Additionally, the great spacing between the H residues in same sequence prohibited the intramolecular crosslinking using metal ions.

Treating of the AM1-based system with chelating agent (EDTA) or protonation of H residues by acidification to pH 3.6 did not change the interfacial tension but reversed the film state into the starting loose form (elasticity modulus <30 mN/m) in seconds due to abortion of crosslinking leading to emulsion break down and foam collapse. Adding excess metal ions or neutralizing the medium pH could return the interfacial peptides into their coherent film again and stabilize the system. AM1 stabilized the foam for about 10 min which collapsed within 1 min after acidification or metal sequestration by EDTA (Dexter et al. 2006). Such switchable behaviour of AM1 allowed a novel and easy demulsification technique for clean oil and water recovery without residues in any phases, which is better than other energy consuming and ineffective approaches. In another study, Dexter and Middelberg compared the effect of various metals on the strength of the formed interfacial film, where Zn²⁺ was the best crosslinker followed by Ni²⁺ and then CO²⁺, while Ca²⁺ and La³⁺ failed to create interfacial bridges, implying the different binding behaviour of metals that can influence film formation and the characteristics of formed films (Table 13.2) (Dexter and Middelberg 2007).

Middelberg et al. also compared between Lac21E (Ac-MEELADSLEELARQVEELESA-NH₂) and AM1 for foam stabilization (Table 13.2) (Middelberg et al. 2008). Lac21E formed a strong elastic interfacial layer of organized connected α-helices at pH 3 (close to isoelectric point 3.95) with lowered interfacial tension (ON state), followed by interfacial disassembly and dissociation from the interface into the bulk medium upon neutralizing the pH to 7.4 due to acquiring a high negative charge and loss of interfacial activity and elasticity (OFF state) that was reversed again by reacidification. So, Lac21E stabilized foams for 5 min at pH 3 but collapsed within 1 min at pH 7. The film formed by Lac21E was stronger than AM1 due to greater accumulation of Lac21E at the interface.

Another Lac21 peptide analogue reported by Dexter and Middelberg is AFD4 (Ac-MKQLADS LHQLAHKVSHLEHA-CONH₂), in which 4 H residues are present at positions (i, i + 3) and (i, i + 4) to endorse intramolecular crosslinking with ‘metal clip‘via sequential α-helical peptide turns to further stabilize the helix, in addition to the interfacial intermolecular bridging (Fig. 13.5 and Table 13.2) (Dexter and Middelberg 2007). Unlike Lac21 and AM1, AFD4 caused a more rapid decline in interfacial tension in the absence of metal ions, since substitution of R with H residue at position 13 widely destabilized the helical form in the bulk phase besides the electrostatic repulsion between cationic amino acids. Addition of metal ions to AFD4-based systems was found to reduce surface tension at a slower rate due to the slower adsorption kinetics after crosslinking in the bulk phase into higher helical structure (Fig. 13.5) (Dexter and Middelberg 2007). AFD4 yielded tougher cohesive films at the interface after metal incorporation than AM1 at same environmental conditions, which demonstrated to be as strong as the lysozyme and β-lactoglobulin films. The effect of metal ion type on AFD4 film strength was similar to AM1, where Zn²⁺ constructed the strongest intermolecular bridge, whereas Ni²⁺ allowed the most stable helix. Like AM1, AFD4 film could be simply switched to detergent state using EDTA or acidification. Foam stabilization property was studied for the three peptides (Dexter and Middelberg 2007), where Lac21 showed to reduce interfacial tension without bubble stabilization both in the presence and absence of metal ions. While AM1 failed in the absence of metal ions to stabilize foam with low foam height after 25 min (3 mm), which slightly increased after metal addition. Same behaviour was observed for AFD4, but the metal addition significantly increased the foam height to 50–53 mm after 25 min and the AFD4-Zn²⁺ systems stabilized the foam for more than 1 h without collapse.

Wibowo et al. introduced the helix-forming peptide AM-S, which is a modified analogue of the peptide bioemulsifier AM1, by conjugating a silken tail peptide inspired from Bombyx mori silkworm fibroin, which is composed of eight hydrophobic amino acid sequence of alternating G and A using PS residues as linkers (Ac-MKQLADS LHQLARQ VSRLEHA-PS-GAGAGAGY-CONH₂) (Fig. 13.5 and Table 13.2) (Wibowo et al. 2017). AM-S combined the AM1 features of facial amphiphilicity, interfacial activity as well as crosslinking ability in the presence of metal ions, in addition to improved air/water interfacial adsorption capability of silken tail, thus maintaining optimum rate of interfacial adsorption and enhancing interfacial stability. The P and S residues of the linker sequence provided free rotation of strands. As discussed earlier for AM1, AM-S peptide also exhibited a random coil structure in aqueous phase with quick diffusion to the interface lowering the interfacial tension to 52 mN/m using 10 μM peptide, which was not changed by introducing ZnCl₂. AM-S created interfacial multilayers of self-associated α-helical conformation rather than the monolayers produced by AM1 due to the hydrophobic bonding between molecules after the silk tail addition (Fig. 13.5). Interestingly, the thickness of AM-S/Zn²⁺ films was higher than AM1 one because of the peptide multiple layering, so greatly resisted desorption from the interface upon compression and stabilized the dense foams containing miniaturized bubbles (Wibowo et al. 2017). However, AM-S interfacial adsorption kinetics was slower than AM1 owing to its increased molecular weight by silk tail. Both AM1 and AM-S could not stabilize the foam, which collapsed within 2 min while they expressed film stability in the presence of ZnCl₂ due to film elasticity without increase in bubble size.

Wang and co-workers constructed another derivative of AM1, C8-AM, which is AM1 peptide sequence capped from N-terminus by a C8 hydrocarbon chain, which is a hydrophobic functionality for enhancing its interfacial anchoring propensity (Table 13.2) (Wang et al. 2017). In comparison to AM-S, C8-AM acquired an α-helical conformation in bulk phase rather than the random coiling of AM-S, where the latter only assembled at the interface and did not form micelles in bulk, but only soluble dimers to tetramers. C8-AM (20 μM) adsorbed at oil/water interface at a slower rate than AM-S, due to its ordered helical structure in bulk in addition to the hydrophobic interaction between C8 chains. In the absence of Zn²⁺, C8-AM reduced interfacial tension to 8.4 mN/m within 400 s compared to 17 mN/m in case of AM-S. In the presence of Zn²⁺, C8-AM crosslinking happened via H residues forming a cohesive interfacial film (Fig. 13.5). After treatment with EDTA, C8-AM-based emulsion was disrupted after 6 hours while AM-S emulsion broke down within 20 min due to the robust binding of C8 groups to the interface even after interrupting the metal ions-mediated crosslinking. Interestingly, C8-AM created 2% Miglyol 812-based O/W nanoemulsion of smaller globular size (147 nm) than AM-S (180 nm) after sonication, where a concentration of 100 μM of C8-AM was enough to stabilize the system for 3 weeks, whereas higher concentration of AM-S (400 μM) was required to achieve equivalent stabilization. Unlike AM-S, C8-AM did not show flexible reversibility at the interface due to rigid binding of C8 hydrophobic group to the interface.

Middelberg and co-workers deigned a novel bifunctional peptide SurSi (Ac-MKQLAHSVSRLEHA-RKKRKKRKKRKKGGGY-CONH₂), by conjugating the R- and K-rich peptide segment, Si module (RKKRKKRKKRKKGGGY), to AM1 (Wibowo et al. 2014). Hui et al. made further study on SurSi, which revealed the ability of this peptide to stabilize nanoemulsion at 400 μM concentration for at least 2 weeks (Table 13.2) (Hui et al. 2016). Furthermore, the ability of coating oil droplets by silica shell was investigated, in order to formulate stable oil-filled nanocapsules near neutral pH at ambient conditions with no harmful ingredients showing sustained release property for the entrapped drugs (Wibowo et al. 2014). AM1 peptide could not perform this function where precipitation of silica occurred in continuous phase not at the interface, forming irregular aggregates. On the other hand, SurSi successfully created the nanocapsule, as the dual functionalized peptide combined the surface active AM1-derived moiety responsible for nanoemulsion formation with the R- and K-rich Si module, which developed biosilicification at oil/water interface (Fig. 13.5). The modulated AM1 moiety contained solely the first and last heptads of the parent AM1, while D at position 7 in the first heptad was replaced with H to retain the metal binding capacity for crosslinking at the interface to enable the formation of cohesive films. The use of 2 heptads only instead of 3 was to enhance interfacial adsorption, as there is an inverse relation between the peptide molecular weight and the square of interfacial adsorption rate (Middelberg et al. 2000).

SurSi interfacial activity and surface coverage were very low at neutral pH as compared to AM1, due to the delayed diffusion to the interface resulting from the strong repulsive forces between peptide chains (theoretical charge = 13.44), leading to large spacing between the strands and hence lack of interfacial film formation even upon adding Zn²⁺. However, neutralizing the charge through increasing the pH could enhance the adsorption kinetics and interfacial activity of SurSi. Indeed, SurSi peptide successfully emulsified Miglyol 812 oil containing oil-soluble moiety (fipronil) in water at neutral pH 7.5, forming nanoemulsion of droplet size 66.5 nm, which was larger in size, lower in zeta potential and less stable than AM1-based emulsion. Increasing the peptide concentration had a positive effect on reducing the droplet size of emulsion (Hui et al. 2016). The addition of tetraethoxysilane (TEOS) to the system produced silica nanocapsules after 20 h via different mechanisms (Fig. 13.5). It was suggested that E, Y, S and H residues hydrolysed TEOS via nucleophilic attack on silicon atom resulting in silanolate anion and silanol (Wibowo et al. 2014). Furthermore, silanolate anions developed electrostatic attraction with the cationic residues K and R, in addition to the H-bonding between silanol, as well as with hydroxyl groups of Y and S. Silanolate anion and silanol acted as nuclei inducing further silica precipitation. The thickness of nanocapsules shell could be amplified by elevating pH to increase the levels of silanolate, reaction time and levels of TEOS (silica species precursor) to slow the release rate of entrapped compounds.

Middelberg and Dwyer crosslinked four AM1 peptides via DPS short peptide linker to enhance folding propensity, developing a soluble and stable protein emulsifier DAMP4 [MD(PSMKQLADSLHQLARQ-VSRLEHAD)₄] (Table 13.2) (Middelberg and Dimitrijev-Dwyer 2011). The D anionic residue induced repulsive forces between strands and P would terminate helix formation, whereas S residue supported the free rotation. The cationic K residue was deprotonated at pH 8.5, so DAMP4 acquired a single negative charge and stabilized the foams created, while decreasing pH by 1 unit (pH 7.4) expedited the foam collapse due to neutralization of the charge, though there was no change of its effect on interfacial tension at both pH values. Destabilization effect of divalent cation Ca²⁺ on the foam formed at pH 8.5 was higher than that of the monovalent Na⁺, due to greater charge screening effect of Ca²⁺. Interestingly, the use of kosmotropic electrolyte such as Na₂SO₄ salt was able to share the intrinsic surface hydration layer of DAMP4 and support the foam stability at pH 7.4 where DAMP4 net charge was zero. Na₂SO₄ salt allowed the electrostatic interaction of SO₄²⁻ ion to the basic residues of DAMP4 (R and K), resulting in amplified repulsive force ‘hard-sphere repulsion’, as well as highly oriented robust hydration film and better foam stability (Middelberg and Dimitrijev-Dwyer 2011; Parsons and Ninham 2010; Tavares et al. 2004; Besseling 1997). On the other hand, the chaotropic salt (NaSCN) does not have a tightly held hydration layer and debilitate the hydration layer of DAMP4 (Middelberg and Dimitrijev-Dwyer 2011; Gao et al. 2009).

In a further study, Dwyer and co-workers investigated the influence of molecular size of peptides and proteins and their bulk structure on their respective interfacial activities, by comparing the long well-organized protein DAMP4 to its simple unorganized peptidic precursor DAMP1 as well as studying their mixture (Table 13.2) (Dwyer et al. 2013). DAMP1 was a slightly modified form of AM1 via adding P and S residues to the N-terminus and a D residue to the C-terminal (PSMKQLADS-LHQLARQVSRLEHAD), while DAMP4 was the quadriad form of DAMP1 with the inclusion of a M and D to N-terminus [MD(PSMKQLADS-LHQLARQ-VSRLEHAD)₄]. The unstructured DAMP1 adsorbed quickly to the air/water interface as monomers of α-helices whereas DAMP4, as mentioned before, formed a robust package of 4 helices, so it was slowly attracted to the interface after unfolding into a chain of engaged DAMP1 forming interfacial monolayers and recording greater interfacial elasticity (Fig. 13.5). Both the peptide and protein quantities were identical at bulk and interface indicating that adsorption energy was affected by the chemical structure only rather than the size which only influenced the adsorption rate. Strikingly, a faster interfacial tension drops upon mixing DAMP1 and DAMP4 compared to individual components, implying augmentation effect of the mixture surface activity.

Inspired by DAMP4 molecular design, Zhao et al. developed the anionic protein emulsifier SP16 [MD(P-S-ANSVAESLANLAESVSELVSNA-D)₄] to create and tailor foams, which can be controlled by pH and ionic strength (Table 13.2) (Zhao et al. 2017). SP16 is composed of four repeats of one peptide sequence linked by an acid cleavable sequence DP followed by S residue as a spacer. The hydrophobic part of the structure maintained the 4-helix bundle conformation, implying heat resistance property besides the hydrophilic and hydrophobic faces created by the orientation of amino acids which guaranteed the surface activity of SP16. Being anionic in nature, SP16 had a very low isoelectric point (pI) of 2.98, leading to the high pH responsiveness of the folded structure and hence the foaming properties. Minor foam formation was achieved at the pI, but altering the pH above or below pI by 1 pH point led to folding of SP16 into helix bundle followed by quick interfacial adsorption, as well as reduction of interfacial tension with enough charge to stabilize foam formation. Increasing the pH to 6 and above, negatively charged the protein hindering its folding, consequently decreased interfacial adsorption propensity and disrupted the orientation of hydrophilic and hydrophobic residues at the interface as well as surface activity, though charge screening by NaCl recovered the foaming activity. Interestingly, at pH 5, SP16 was able to fold into helix bundle in the bulk; however, due to its high negative charge there was a delay in interfacial adsorption behaviour leading to weak foam formation, unless NaCl was added.

Besides the hydrophobin and Lac analogous peptides, other designs of α-helical peptide emulsifiers were reported. For instance, anionic poly(γ-benzyl l-glutamate)_n polypeptides (PBLG) were developed by Morikawa and co-workers, with different polymerization numbers (1a, n = 12 and 1b, n = 32) (Table 13.2) (Morikawa et al. 2005). PBLGs could self-associate into α-helical robust rods and form O/W microemulsion (CH₂Cl₂/water = 1:2 or CH₂Cl₂/methanol/water = 1:1:1, v/v), which were used as templates for formulation of stable hollow microcapsules upon evaporation of the volatile organic phase, a system that could be used to deliver both hydrophilic and lipophilic drugs. Only 68% of the shorter oligopeptide 1a formed α-helices in the emulsion (with the appearance of some β-sheet structure after air drying), while the longer one 1b had 97% of α-helix conformation confirming that longer sequences provided more helical propensity. The amphiphilic polypeptide was adsorbed to the organic/aqueous interface and upon evaporation more monomers were attracted to the interface forming small multilayered microcapsules of assembled peptides (1–5 μm) containing aqueous phase in the core. These microcapsules were able to entrap a hydrophobic molecule (pyrene) via binding to the shell. The dichloromethane emulsion showed aqueous layers separation after 5 min and the emulsion layer contained droplets of 5–60 μm in size. Adding methanol to the system decreased the droplet size within 3 min to <10 μm and enabled longer stabilization for 24 h before separation of aqueous layers.

Amphiphilic diblock copolypeptide emulsifiers were also developed by Hanson et al., with N-terminus hydrophilic block formed of cationic poly-l-lysine sequence (20–100 residues) conjugated to a hydrophobic racemic leucine (rac-L) sequence (5–30 residues) from the C-terminus (Fig. 13.6) (Hanson et al. 2008). The surfactant-like design imbued the polypeptide with surface activity and interfacial stabilization ability owing to the amphiphilicity and intermolecular H-bonding. These emulsifiers K₂₀(rac-L)₁₀, K₄₀(rac-L)₁₀, K₄₀(rac-L)₂₀, K₄₀(rac-L)₃₀, K₆₀(rac-L)₂₀ and K₁₀₀(rac-L)₁₀ could develop W/O/W double emulsions (Table 13.2 and Fig. 13.6) via easy homogenization and microfluidic techniques without the aid of other emulsifiers, as well as providing long-term stabilization for the formed emulsions (>9 months), while the less hydrophobic emulsifiers (5 residues of rac-L) stabilized the system for 12 months with bigger droplet size. The use of hand operated homogenization or ultrasonic mixing yielded microemulsion which was converted to nanoemulsion after microfluidic homogenization (Fig. 13.6). The flexibility of orientation of rac-L chain in diblock copolypeptide enhanced its solubility in oil with minimal interchain H-bonding and packing and hence weakly stabilized the inner water globules in the external oil phase. However, on the outer oil/water interface, intense packing of the rac-L in oily phase resulted in strong crosslinking by H-bonding that inhibited the conversion into a simple emulsion and stabilization of the double emulsion without inclusion of other additives. Interestingly, the R₄₀(rac-L)₁₀, E₄₀(rac-L)₁₀, K₆₀(rac-V)₂₀ and K₆₀(rac-A)₂₀ also stabilized double emulsions due to structure similarity to K_x(rac-L)_y, with the flexibility of design versatility (Table 13.2 and Fig. 13.6). On the other hand, the K₆₀L₂₀ emulsifier only produced simple O/W emulsions owing to the low solvation of the homochiral l-leucine part in oil phase upon acquisition of α-helical structure, while the hydrophilicity of the homopolypeptide block K₆₀ favoured rapid phase separation (Table 13.2 and Fig. 13.6). Thus, for this type of emulsifiers, the interfacial helical propensity impaired surface activity rather than enhancing it.

Fig. 13.6

Emulsification technique to generate both simple and double emulsions by block copolypeptide surfactants. Adapted from (Hanson et al. 2008), with copyrights permission from Springer Nature

13.2.3 β-Sheets Peptide Emulsifiers

β-sheet forming peptides have been widely studied for the development of amyloid-like nanofibrous structures for the formation of hydrogels in aqueous media (Wychowaniec et al. 2020; Gazit 2007; Aggeli et al. 2001; Zhang et al. 1994). In addition to hydrogelation properties, amphiphilic β-sheet forming peptides have been shown to possess interfacial activity (Fig. 13.7a). Dexter was the first to introduce amphiphilic β-strand peptides as emulsifying agent that adsorbed from bulk to the interface (Dexter 2010). Inspired by the β-sheet forming amphiphilic sequence design of the Tirrell group that form monolayer at the air/water interface (Rapaport et al. 2000), Dexter developed the following four short nonapeptides B-14 (Ac-PDFDFDFDP-CONH₂), B-15 (Ac-PHFHFHFHP-CONH₂), B-16 (Ac-PEFEFEFEP-CONH₂), and B-17 (Ac-PKFKFKFKP-CONH₂) of alternating hydrophobic residues (P or F) with same charged residues, which are either anionic (E, D) or cationic (K or H), to hinder any possible electrostatic crosslinking of the β-sheets in bulk phase and aid in the interfacial adsorption (Table 13.2 and Fig. 13.7b) (Dexter 2010). These peptides showed surface activity at different pH values in a similar manner to traditional ionic emulsifiers, but with good biocompatibility and biodegradability. When the peptide was uncharged by adjusting the pH of B-14 and B-16 to acidic or B-15 and B-17 to basic, it acquired β-hairpin conformation rather than β-sheet in bulk phase, owing to the aromatic interaction of F residues at positions 3 and 7. Formation of strong rigid film of peptide aggregates at the interface is mediated by H-bonding and hydrophobic interactions, forming emulsion with large droplet size ~1500 ± 100 nm except for B-17 (~400 nm) due to the reduced zeta potential which allowed coalescence to happen (Fig. 13.7b). Increasing zeta potential through pH change led to formation of polyproline II-form helices in bulk and decreased the interfacial spreading with better emulsification due to electrostatic repulsion, so intermediate charging of peptide was the optimum. Based on this, B-14 to B-17 could stabilize emulsions at different pH ranges; however, they were more successful as ionic emulsifiers than film forming emulsifiers. The kinetics of B-14 and B-16 peptides adsorption were similar at all pH ranges, while decreasing charges of B-15 and B-17 slowed their adsorption rate due to aggregation in aqueous solution.

Fig. 13.7

(a) Schematic presentation of the general self-assembly mechanism of β-sheet forming peptides into amphiphilic β-sheet nanofibres, forming hydrogels in aqueous media and emulsions/emulgels in biphasic media. Adapted from (Wychowaniec et al. 2020), with copyrights permission from the American Chemical Society published under Creative Commons Attribution 4.0 International license. (b–e) Examples of β-sheet forming peptide emulsifiers, (b) Amphiphilic β-strand forming peptides; B-14, B-15, B-16 and B-17, which self-assemble at oil/water interface into rigid peptide films, stabilizing emulsion formation. Adapted from (Dexter 2010), with copyrights permission from the American Chemical Society. (c) Self-assembly of Q11 peptide at oil/water interface, forming W/O emulsion through formation of microgel spheres surrounding the aqueous media, Left: Fluorescent nitrobenzoxadiazole (NBD)-containing microgels produced with a paddle-type stirrer, Middle: TEM of the microgel fibrous structure (scale bar 100 nm), Right: Laser scanning confocal microscopy of Congo red-stained microgels. Adapted from (Tian et al. 2011), with copyrights permission from Royal Society of Chemistry. (d) Self-assembly of A₉R peptide into hydrogel in water and O/W emulsion. Left: Laser scanning confocal microscopy, Middle and Right: cryo-SEM of 0.05 wt% A₉R emulsion. Adapted from (Castelletto et al. 2019), with copyrights permission from the American Chemical Society. (e) Self-assembly of Phg4 peptide into β-sheet nanofibres, forming hydrogel and O/W emulgel, Left: Fluorescence microscopy of Phg4 O/W emulsion (stained with FITC), Middle: SEM of Phg4 vacuum-dried O/W emulgel, Right: SEM of Phg4 air-dried O/W emulgel. Adapted from (Wychowaniec et al. 2020), with copyrights permission from the American Chemical Society published under Creative Commons Attribution 4.0 International license

Another β-sheet forming peptide is the fibrillized short peptide Q11 (Ac-QQKFQFQFEQQ-Am), which was reported by Tian et al. to self-assemble at oil/water interface (Table 13.2) (Tian et al. 2011). Interfacial self-assembly of Q11 is triggered by the addition of Dulbecco’s phosphate buffer saline to stabilize W/O emulsion via formation of microgel spheres surrounding the internal aqueous phase (Fig. 13.7c). It was possible to control the microgel size and size distribution using characteristics of speed and type of blade mixer, as well as further shear steps beyond the gelation. Interestingly, NIH 3T3 fibroblasts and C3H10T½ mouse pluripotent stem cells were encapsulated within the microgel spheres with acceptable biocompatibility during the emulsion preparation process, by dispersing the cells within the peptide aqueous solution before addition of buffer for emulsification.

Surfactant-like β-sheet forming peptides were also designed to act as emulsifiers, where Castelletto and co-workers introduced the surfactant-like A₉R peptide which formed hydrogel in water and stabilized O/W Pickering emulsions, in a similar way to proteins (Table 13.2) (Castelletto et al. 2019). Unlike traditional surfactants, A₉R did not from interfacial monolayers, but rather the emulsion droplets were wrapped by a network of self-assembled β-sheet fibrils (25 nm diameter) of A₉R when exceeding the critical aggregation concentration (0.05–0.07%) without the need for pH tuning due to the strong packing of A residues in the core and surface exposure of the R residue to aqueous phase (Fig. 13.7d). The emulsion was stable by A₉R emulsifier for 3 days instead of 1 day only in absence of the peptide. On-demand demulsification was achieved by the addition of the proteolytic enzyme elastase, which could be exploited for site-specific release of entrapped drugs. In addition, A₉R showed a remarkable antimicrobial action towards G-ve bacteria including P. aeruginosa and E. coli.

Ionic self-complementary peptides are characterized by alternation of a hydrophobic residue with counter charge residues and are the most used sequences for the formation of β-sheet assembled structures. Elsawy and co-workers were the first to develop the shortest ionic self-complementary tetrapeptide Phg4 (PhgEPhgK, where Phg is phenylglycine), which can self-assemble into thermodynamically stable β-sheet nanofibres (diameter ~ 7–11 nm) that are not only capable of hydrogel formation in monophasic aqueous medium but can also stabilize O/W emulsions in biphasic media (Table 13.2) (Wychowaniec et al. 2020). Emulsification was pH and concentration dependent, where surface activity was only achieved at the assembly pH range (4.5–8) and concentration (>2%W/V). Electron microscopy (both SEM and TEM) showed the adsorption of peptide nanofibres at the oil/water interface forming nanofibrillar microspheres (∼50–200 μm) encapsulating oil droplets within the nanofibrous network of the continuous phase (Fig. 13.7e). These structural features stabilized the formation of viscoelastic emulgels, which showed to be shear thinning and thixotropic. Emulsification was pH switchable, as emulgel formation only happened at self-assembly pH range (4.5–8), where the overall charge of the peptide chain is neutral favouring both assembly and interfacial adsorption. While outside this pH range, phase separation occurred owing to the repulsion of charged peptide chains and failure to self-assemble into the emulsifier β-sheet nanofibre form. Phg4 successfully formed O/W emulgels with chloroform and melissa oil as the organic phase at a range of oil:water ratios and peptide concentrations. The peptide emulsifier stabilized melissa emulgels for longer than 5 weeks and showed to withstand harsh environmental conditions (heating at 60 °C for 3 hours and salting out for 1 week) with no signs of phase separation, which was superior to commercial emulsifiers, such as cetrimide, SDS and tween 80.

13.2.4 Miscellaneous Surfactant-Like Peptides

In addition to peptide emulsifiers that assemble into well-defined secondary structures, there has been endeavours for molecular engineering of peptides that mimic surfactant structures. One example of these emulsifiers is the synthetic pseudopeptide gemini amphiphile (GAP), which was reported by Lotfallah et al. (Lotfallah et al. 2015). GAP was designed to have a hydrophilic head of valine-derived pseudopeptide exposed to aqueous phase with 2 hydrophobic tails that solubilize in oil phase by van der Waals force with variable critical aggregation concentration (4–25 mM) depending on the ionic strength and pH of the system (Fig. 13.8). Amphiphilic GAP thus self-associates at oil/water interface forming hollow microspheres (1–5 mm in diameter), therefore stabilizing O/W emulsion containing <1% of isopropyl myristate oil, after strong mixing by magnetic stirrer for 30–90 min (Table 13.2 and Fig. 13.8). The formed emulsion displayed a superior stability against exposure to strong acids, heating at 45 °C for 4 months, mechanical stress (centrifugation for 30 min at 3000 rpm with good system recovery if centrifuged for longer time) and long-term storage. However, proteolysis with thermolysin enzyme or alkalinisation to pH 12 led to disassembly and phase separation. It was suggested that surface activity of GAP required total or partial protonation of its amino groups to stabilize the emulsion droplets, so it could be exploited as a carrier for delivery of anionic nucleic acids. The system can also be used for the encapsulation of a variety of hydrophobic compounds, which was demonstrated by the successful encapsulation of the fluorescent hydrophobic model compounds N,N-diethyldansylamide and 9,10-dimethylanthracene.

Fig. 13.8

Diagram of GAP self-assembling to form an O/W emulsion with macroscopic and microscopic images of the formed emulsion. Adapted from (Lotfallah et al. 2015), with copyrights permission from Royal Society of Chemistry

Lipopeptides also showed to exhibit good surface activity thanks to the amphiphilicity of the molecular design. The lipodipeptide emulsifier, C₁₃-KR, was purposely designed by Lv et al. to stabilize emulsions in acidic environment, where its surface activity was compared to whey protein isolate and Tween 80 (Table 13.2) (Lv et al. 2019). The peptide-based emulsion showed the smallest droplet size (<1 μm) and the largest surface charge (100 mV), electrostatic repulsion and creaming stability, compared to the emulsions formed by the other two emulsifiers. The low molecular weight and high surface charge of C₁₃-KR offered a fast interfacial adsorption rate (400 Hz). Besides, the enhanced surface activity due to the good interfacial anchoring of the hydrocarbon chain decreased droplet size of the emulsion, with further size reduction achieved by either increasing the peptide concentration or prolongation of sonication time. Formation of stable highly elastic films at the interface in addition to the good surface activity enhanced storage stability. In addition, the high zeta potential of the C₁₃-KR stabilized droplets, conferring resistance to salting out and thermal aggregation.

Another lipopeptide emulsifier and gelator was formed via Schiff’s base formation upon mixing the precursors A, hydrophilic peptide NH₂-lauroylGGGH, and B, lipophilic 1,4-phthalaldehyde, at ambient temperature with molar ratio of A:B 2:1 (Table 13.2) (Nishida et al. 2017). The product was a mixture of ABA and AB that self-assembled into nanofibres (30–100 nm) entangled into thicker bundles at pH 7.4 without heating. Entangled nanofibres stabilized emulsions in biphasic media and formed hydrogel in aqueous medium within short period, 1 and 15 min, respectively. The hydrogel storage modulus was higher than its loss modulus after 3 min of precursors blending and progressively increased by time confirming the self-assembly and gelation process. Gel–sol transition occurred by heating to 75 °C or changing the pH to low or high values due to protonation or hydrolysis of the Schiff’s base. Fabrication of emulsion was triggered by homogenizing 4 parts by volume of aqueous phase containing A with 1 part of organic phase containing B with the amount of precursor A was double B quantity. Homogenization for 1 min yielded thinner nanofibril bundles of ABA and AB (30–40 nm thickness) at the interface only, to produce globules of 22.4 μm diameter, without gelation in the bulk phase. This system was not formed upon the use of single precursor only and this was confirmed by confocal laser scanning microscopy. Acidification of the system with HCl or heating to 80 °C led to disassembly of the interfacial nanofibres and droplets fusion into larger size. The formed emulsion droplets were employed as stimuli-sensitive microreactors to regulate a click reaction of an alkyne with an azide by entrapping each reactant in separate oil droplets of the emulsion stabilized by the nanofibres and then initiating the reaction on demand when disassembly induced by external stimulus (heating or acidification).

13.3 Protein Emulsifiers

Proteins have been widely used as emulsifying and foaming agent in food industry (Wilde et al. 2004; McClements 2004). They are readily available, natural and non-toxic; moreover, they can produce and stabilize emulsions with desirable physicochemical properties. Proteins adsorb to the interface and allow droplet dispersion through reduction of interfacial tension while hindering droplet coalescence via either steric stabilization or electrostatic repulsion of the formed protective layer (Fig. 13.9) (McClements 2004; Walstra 2003). In this section, we will expand more on the main protein emulsifiers, which have been commonly used in industry, such as milk proteins, hydrophobins, gelatin and pea proteins.

Fig. 13.9

The possible emulsion stabilization mechanisms using proteins by electrostatic repulsion or steric stabilization

13.3.1 Milk Protein Emulsifiers

Milk proteins are mainly divided into two families, caseins and whey proteins.

13.3.1.1 Caseins

Caseins, belonging to phosphoproteins, are known to be the principal protein constituents in milk (80%) (Tomadoni et al. 2020). Caseins include four major flexible proteins, namely, α_s1-, α_s2-, β- and κ-casein. They are unstructured negatively charged proteins that do not form α-helices or β-sheets due to the numerous P residues in their sequence as well as absence of disulphide bonds (Bouyer et al. 2012; Tomadoni et al. 2020; Sarkar and Singh 2016). Caseins acquire self-assembled micellar forms upon binding to calcium phosphate in milk that interconnect via non-covalent bindings (Dickinson 2006). α_s1 and β caseins are highly efficient emulsifiers that have good tendency to adsorb at oil/water interface and reduce interfacial tension as a result of their amphiphilicity (Table 13.3) (Bouyer et al. 2012). β-casein exhibited tail-like anchoring behaviour to the interface via its terminal part, while α_s1 casein attached to the interface in a loop-like orientation through the mid-part of its sequence (Fig. 13.10).

Table 13.3 Classification of the protein emulsifiers and their uses

Fig. 13.10

Milk proteins interfacial adsorption where α_s1 and β caseins acquire random coil structure in solution that adsorb to the emulsion interface and then β caseins exhibit tail-like anchoring behaviour to the interface via its terminal part, while α_s1 casein attached to the interface in a loop-like orientation through the mid-part of its sequence. Moreover, the folded structure of β-lactoglobulin in water shows interconnected antiparallel β-sheets and flanked α-helices that rearranges upon adsorption to the interface. Adapted from (Dickinson 1998), with copyrights permission from Royal Society of Chemistry and (Zhai et al. 2011), reference with copyrights permission from the American Society of Chemistry

β-casein showed better emulsification properties, more reduction of interfacial tension and higher preferential interfacial adsorption with thicker film formation than α_s1 (Sarkar and Singh 2016). β-casein can also prevent the coalescence of globules for long time via two mechanisms: firstly steric as well as electrostatic repulsive forces through extended saturated monolayer film of 10 nm width as reported by Dickinson and Davies (1999) as well as Fang and Dalgleish (1998) and secondly the viscoelastic behaviour of the adsorbed protein film at the interface as illustrated by Bantchev and Schwartz (2003) and Gau et al. (1994) (Fig. 13.9). Despite its prominent surface activity, casein-based emulsions can easily flocculate when treated with calcium ions, due to screening of the protein’s negative charge and thinning of its interfacial film (Dalgleish 1997; Agboola and Dalgleish 1995).

Sodium caseinate is the commonly used form as emulsifier which is the sodium salt of mixture of all casein proteins, as it is not possible to use individual types of casein because of the high purification cost of individual components (Bouyer et al. 2012; Sarkar and Singh 2016). Sodium caseinate is produced by reducing the pH of skimmed milk to 4.6 to precipitate the casein followed by solubilization using sodium salts at neutral pH and finally drying. Stabilization of the sodium caseinate-based O/W emulsions is affected by protein concentration, molecular orientation in the bulk phase and the components of protein interfacial film (Sarkar and Singh 2016).

The enzymatic hydrolysis of β-casein yielded 2 peptide sequences: a hydrophilic peptide (β-CN, f1–25) of 25 amino acids and a hydrophobic one (β-CN, f193–209) of 17 amino acids (Lee et al. 1987). These peptides showed poor emulsification properties at neutral pH, which was ascribed to the short length of these hydrolysates. However, surface activity improved at extreme pH (below 4 and above 8) in case of the β-CN, f193–209 sequence and at acidic pH for β-CN, f1–25 peptide. These casein-derived peptides showed formation of thick films around the oil droplets suggesting the possible association of peptides at the interface.

13.3.1.2 Whey Proteins

Milk whey proteins are primarily composed of 80% β-lactoglobulin and 15% α-lactalbumin with minor proteins of lactoferrin, serum albumin and immunoglobulins (Bouyer et al. 2012; Tomadoni et al. 2020). Unlike caseins, they have secondary, tertiary and quaternary structures connected by disulphide bridges, therefore not as flexible as casein (Bouyer et al. 2012). β-lactoglobulin and α-lactalbumin hold ideal emulsification characteristics and were utilized successfully to stabilize food emulsions; however, both possessed a slight lower stability than casein-based emulsion in similar environment (Tcholakova et al. 2006; Dickinson 1997; Hunt and Dalgleish 1994; Dalgleish 1995). β-lactoglobulin, a dense folded protein, is constituted of 162 amino acid sequence containing 2 disulphide bridges as well as 1 free thiol moiety. Its folded structure shows interconnected antiparallel β-sheets and flanked α-helices (Fig. 13.10) (Sarkar and Singh 2016). After interfacial adsorption, partial unfolding is initiated with intermolecular binding of β-sheets through their free sulfhydryl groups to form disulphide bridges between molecules and develops a compact interfacial layer of 2 nm width irreversibly fortified upon storage with the increment of polymerization degree (Dickinson 1998; Dalgleish 2004). β-lactoglobulin is able to stabilize emulsions at neutral medium due to the protein negative charge that promoted electrostatic repulsion between emulsion globules besides reduction of interfacial tension (Kim et al. 2002; Wüstneck et al. 1999; Paulsson and Dejmek 1992). However, like most proteins, β-lactoglobulin is thermolabile, where it unfolds at 70 °C, resulting in protein aggregation and droplet coalescence (Kim et al. 2002; Ye 2010).

The second abundant whey protein, α-lactalbumin, is a calcium metalloprotein held together via 4 intrachain disulphide bridges without free thiol groups, unlike β-lactoglobulin (Sarkar and Singh 2016). α-lactalbumin possesses more C residues and less P moieties than β-lactoglobulin (Ng-Kwai-Hang 2003). In addition, α-lactalbumin thermally degrades at lower temperature (66 °C) than β-lactoglobulin without aggregation due to the absence of free thiols (Sarkar and Singh 2016; Considine et al. 2007). Dickinson et al. studied the competitive interfacial adsorption between both proteins and suggested the irreversible adsorption of the firstly added protein, which predominated the interface and could not be displaced by the secondly added one (Dickinson et al. 1989).

Lactoferrin is another whey protein component present in a minor level, composed of 700 amino acids capable of binding to iron and exhibits a high positive zeta potential (+50 mV) at neutral pH. It can therefore form stable emulsions of cationic droplets at a wider pH range (3–7), unlike caseinates and β-lactoglobulin that can stabilize emulsion at neutral pH only where their pI ranged between 4.5 and 5.5 (Baker and Baker 2005; Ye and Singh 2006). Wahlgren et al. reported the possibility of electrostatic attraction between the anionic β-lactoglobulin and the cationic lactoferrin (Wahlgren et al. 1993). Based on this, Ye and Singh developed stable O/W emulsions with thick films of multilayered binary protein mixture at neutral pH (Ye and Singh 2007). Although the net charge of the protein mixture is zero, the dense multilayers offered robust steric hindrance that prevented coalescence of the emulsion droplets.

The marketed forms of whey protein emulsifiers are whey protein concentrate (WPCs) (25–80% protein) and whey protein isolate (WPIs) (>90% protein) extensively utilized in food industries (Sarkar and Singh 2016). WPI has been also utilized to develop O/W emulsions using simple stirring methods (Table 13.3) (Sedaghat Doost et al. 2019; Teo et al. 2016; Ng et al. 2017; Zhu et al. 2018; Fioramonti et al. 2015).

To summarize, milk proteins are considered the most widely used bioemulsifier in food industry being a natural amphiphile, generally regarded as safe (GRAS), i.e. biocompatible, non-toxic, biodegradable, offer great nutritional value and have antioxidant properties and easily sourced from milk, hence sustainable (Ha and Lee 2020; He et al. 2011). However, they suffer from batch-to-batch variabilities and immunogenicity where cow’s milk is one of the eight main food categories causing 90% of food allergies (Nutten et al. 2020; Broersen 2020).

13.3.2 Hydrophobins

Hydrophobins are a group of globular fungal proteins that are believed to behave like small molecule emulsifiers. As a result of their high amphiphilicity, as well as rapid interfacial adsorption due to their relatively small sizes, hydrophobins showed the greatest surface activity between all types of protein emulsifiers (Paukkonen et al. 2017; Berger and Sallada 2019). They can be classified according to their self-assembled conformation, hydropathy patterns and solubility into two types: class I as SC3 that yields insoluble rodlets with amyloid orientation solely soluble in strong acids, and class II, like HFBI and HFBII, which acquire highly organized 2D interfacial monolayers. Class II also are widely soluble in organic solvents and detergents and easily solubilized in water even at high concentration (100 mg/mL) (Paukkonen et al. 2017; Berger and Sallada 2019). Hydrophobins possess 4 disulphide bridges in their tertiary structure and show intermolecular interactions which allow self-association as amphipathic elastic monolayer films at oil/water and air/water interfaces. Hence, they decrease interfacial tension and can change the behaviour of surfaces from lipophilic nature to hydrophilic and vice versa (Paukkonen et al. 2017; Paananen et al. 2013). Therefore, hydrophobins are suitable for formulation of foams and O/W emulsions with long-term stability up to 4 months at ambient storage conditions (Table 13.3) (Paukkonen et al. 2017; Tchuenbou-Magaia et al. 2009; Green et al. 2013; Cox et al. 2007, 2009) and are considered as non-toxic and non-immunogenic pharmaceutical excipients (Aimanianda et al. 2009; Ebbole 1997).

13.3.3 Gelatin

Gelatin is a high-molecular weight protein obtained by hydrolysis of animal collagen via boiling at acidic pH to yield Type A gelatin or basic pH to produce Type B gelatin (Bouyer et al. 2012; Taherian et al. 2011). The difference in their preparation method led to differences in their physicochemical properties where pI of Type A gelation is ranged from 7 to 9, whereas Type B gelatin is 4.8 to 5.1 (Djagny et al. 2001). Such variation in characteristic was attributed to that basic hydrolysis of collagen had converted N and Q into D and E that yielded Type B gelatin (Veis 1964).

Gelatin proteins showed to have some surface activity. Both low- and high-molecular weight fish gelatin demonstrated to stabilize 20% corn oil in water as O/W emulsion at ≥4% w/w total gelatin concentration as reported by Surh and co-workers (Table 13.3) (Surh et al. 2006). However, large oil droplets (>10 μm) were observed in the emulsion due to low surface activity of gelatin, which could lead to coalescence. Low-molecular weight fish gelatin produced greater fractions of large droplets than high-molecular weight one; however, emulsions formed by the earlier were more stable against creaming. Emulsions stabilized by fish gelatin showed resistance to phase separation at pH range between 3 and 8, heating for 30 min at 90 °C and NaCl salt up to 250 mM.

Land snails-derived gelatin also possessed interfacial activity. Zarai et al. studied the emulsifying properties of this gelatin and suggested the possibility to act a bioemulsifier, since it demonstrated to stabilize olive O/W emulsions at total gelatin concentrations of 0.5–4% w/v (Table 13.3) (Zarai et al. 2016). At high gelatin levels, a crosslinked 3D matrix was formed in the bulk phase, which, in addition to its interfacial adsorption, help trapping the oil globules inside the hydrogel network forming a strong emulgel. Land snails-derived gelatin produced emulsions with higher stability than other gelatins from different origins (smooth hound, zebra blenny, barbel and grey triggerfish), could be attributed to the differences in composition, protein conformation and properties. Snail gelatin also produced dense stable foams at high concentrations 1–4%w/v due to the formation of interfacial thick gelatin films, but markedly collapsed after 1 hour of mixing (Zarai et al. 2016).

13.3.4 Pea Proteins

Pea proteins include two main constituents: vicilin, a 7S storage globulin, and legumin, a 11S storage globulin (Gueguen et al. 1988). Legumin protein consists of linked subunits via non-covalent binding in a hexameric structure, while vicilin protein shows a trimeric conformation (Bouyer et al. 2012; Burger and Zhang 2019). The ratio between legumin and vicilin, as well as their structures, differs depending on the production technique and species, consequently variation of their functions. In general, vicilin possessed higher surface activity and emulsion stabilization efficiency than legumin (Burger and Zhang 2019). Pea proteins successfully produced emulsions through interfacial adsorption that occurred in two stages. Firstly, protein migration from aqueous phase and anchoring to the oil/water interface via the exposed hydrophobic part of the molecule. Secondly, protein reorientation occurred to position the hydrophilic parts towards the aqueous phase and solubilize the lipophilic areas in the organic phase forming a viscoelastic rigid interfacial layer that stabilized the system by steric effect and electrostatic repulsions (Burger and Zhang 2019; Sridharan et al. 2020). In essence, pea proteins are capable of reducing interfacial tension (Ducel et al. 2004) and stabilize emulsion formation at low pH both by Pickering mechanism and hydrogelation (Table 13.3) (Liang and Tang 2014). However, pea proteins emulsions could not withstand changing salts concentration, where increasing NaCl concentration of 50 to 200 mM enlarged the emulsion droplet from 6 to 14.05 μm and the medium pH should be away from its pI (4.5) where the extreme pH values (pH 3 and 8) were preferred to allow the charging of protein molecules, so it was necessary to add another emulsifier, such as pectin, to stabilize emulsions and prevent phase separation in harsh conditions (Gharsallaoui et al. 2010a, b; Ettoumi et al. 2016; Ladjal-Ettoumi et al. 2016).

13.4 Protein–Polysaccharide Mixed Emulsifiers

On one hand, proteins are characterized by the quick interfacial adsorption and film formation that prohibits emulsion coalescence, though stability can always be impaired either by changing pH near to pI or by salting out (Ozturk and McClements 2016; Yang et al. 2019). On the other hand, polysaccharides can develop highly stable emulsions but using extremely high polymer levels (Chen et al. 2011b; Kharat et al. 2018). Therefore, combining both proteins and polysaccharides via covalent conjugation, electrostatic attractions or by physical mixing combines the attributes of both components, fast interfacial adsorption of protein in addition to the viscosity increment and steric hindrance of polysaccharides, thus creating highly stable emulsions (de Oliveira et al. 2016; Setiowati et al. 2020). In this section we will discuss protein–polysaccharide mixed emulsifiers formed by covalent conjugation and non-covalent interaction.

13.4.1 Protein–Polysaccharide Covalent Conjugates

Controlled Maillard reaction was used to conjugate amino groups of protein with the carbonyl functionalities of polysaccharides to yield composites having higher emulsifying power than either polysaccharides or proteins alone, such as sodium caseinate/maltodextrin (O’Regan and Mulvihill 2010), β-lactoglobulin/dextran (Wooster and Augustin 2006) and whey protein isolate/maltodextrin (Akhtar and Dickinson 2007). These complexes also offered resistance to undesirable environmental conditions, such as freezing, heating or great ionic strength (Bouyer et al. 2012; Setiowati et al. 2020). For instance, O’Regan and Mulvihill previously showed the augmented stability of emulsion formulated using sodium caseinate maltodextrin for 7 days at 45 °C in comparison to sodium caseinate (O’Regan and Mulvihill 2010). Setiowati and co-workers also demonstrated that chemical conjugation of whey protein isolate to low methoxy pectin resulted in a conjugate of superior interfacial activity and emulsion stabilization than the electrostatic complex (Diah Setiowati et al. 2019; Setiowati et al. 2017).

Zhang et al. showed the effect of glycosylation of soy protein isolates with maltodextrin using Maillard reaction on the emulsification of fish oil compared with soy protein isolate/maltodextrin mixture where smaller droplet size, lower polydispersity index and improved physical stability (3-week storage) were observed with both hydrolysate of soy protein isolate/maltodextrin conjugate and hydrolysed soy protein isolate/maltodextrin conjugate-based O/W emulsions (Fig. 13.11) (Zhang et al. 2014, 2015).

Fig. 13.11

O/W emulsion stabilization through conjugation between the hydrolysed soy protein and maltodextrin or hydrolysis of the conjugated soy protein and maltodextrin using Maillard reaction. Adapted from (Zhang et al. 2015), and (Zhang et al. 2014), with copyrights permission from Elsevier

13.4.2 Protein–Polysaccharide Physical Mixtures

Electrostatic interactions between counter charge polysaccharides and proteins at the interface were claimed to boost the emulsion stability (Li et al. 2019). There are two main reported methods for mixing emulsifiers, where one method involved one pot mixing of the two components to prepare the complex necessary for emulsion production (Yang et al. 2019; Lutz et al. 2009; Laplante et al. 2006; Yildiz et al. 2018). However, the other method is layer by layer electrostatic deposition, where initially a primary protein-based emulsion was developed followed by incorporating the polysaccharide to the system that interacted at droplet interface with the protein to form a second coat around droplets (Khalloufi et al. 2009). These deposition steps could be repeated several times to produce multicoats stabilizing the emulsion.

Initial emulsification with protein is essential, as protein emulsifiers exhibit fast stabilization kinetics and reduction of interfacial tension. The protein–polysaccharide mixed emulsifiers showed greater resistance to environmental conditions including dehydration, high salt concentrations and temperature changes. For example, coalescence of casein-based emulsion near the pI of casein was substantially hindered by adding octenyl succinic anhydride modified starch (OSAS) and this blend demonstrated higher oxidative stability than OSAS-based emulsion (Yang et al. 2019).

Another physical mixing approach, not involving electrostatic interaction, depends on addition of a protein emulsifier to the biphasic system for interfacial adsorption and formation of small globules of the internal phase. This is followed by the introduction of the polysaccharide component as a viscosity modifying agent, which enhances the viscosity of the continuous aqueous phase via thickened networks and hampers droplet movement, thus preventing creaming/sedimentation of the internal phase droplets. Several mixtures were prepared by the viscosity enhancement method, such as whey protein isolate/xanthan gum (Sun and Gunasekaran 2009), sodium caseinate/locust bean gum (Perrechil and Cunha 2010), sodium caseinate/xanthan gum (Moschakis et al. 2005) and whey protein isolate/flaxseed gum (Khalloufi et al. 2008).

13.5 Summary

Short peptides are emerging new class of bioemulsifiers that have been shown to stabilize different foams and emulsions and have proven to be superior to traditional commercial emulsifiers in long-term stabilization of these colloidal systems. Besides, peptides are similar to proteins in being safe, biodegradable, biocompatible and of low toxicity to human health and environment, thus overcoming the common limitations of the currently used traditional emulsifiers. On the top of these, the physicochemical properties of peptide emulsifiers can be finely tuned by the rational chemical design of the amino acid sequence, providing a plethora of design options for controlling their behaviour at the organic/aqueous and gas/aqueous interfaces. In general, peptide-based emulsification was reported to occur either via micelle-like structure formation, akin to conventional emulsifier, or through self-assembly into amphiphilic nanofibrous structure at the interface and in bulk continuous phase. Proteins have also been widely used in food industry to stabilize emulsions, through steric stabilization and/or electrostatic repulsion. However, proteins have some limitations, mainly due to the possible batch-to-batch variation, chemical and physical instability as well as potential immunogenicity. Protein–polysaccharide mixed emulsifiers have also been introduced but have shown to be less efficient in stabilizing emulsions and foams when compared to peptides and proteins. In brief, peptides and proteins are important emerging subclasses of bioemulsifier armamentarium, though there is still a lot of work to be done for their wide utilization in industry, when it comes to large-scale production of these molecules, as well as tackling stability issues during storage and manufacturing.