Structural Basis and Molecular Mechanism of Yeast Biosorption: for Food Safety and Nutrition

ABSTRACT

Yeast is widely used in food processing, yet its unique cellular structure and the resulting potential for biosorption remains underutilized. A literature search was performed in databases including PubMed and ScienceDirect using a combination of relevant keywords, including yeast biosorption, food contaminants, bioactive compounds, and biosorption enhancement. The search was specifically focused on studies pertaining to food or food‐related systems. This narrative review elucidates that rapid, passive surface biosorption mediated by the yeast cell membrane (primarily the mannan layer) conforms to pseudo‐second‐order kinetics and the Langmuir/Freundlich isotherm, while active, energy‐driven intracellular accumulation dominates at trace levels. For food safety, optimized yeast‐mediated processes achieve 70%–90% removal of heavy metals (Pb²⁺, Hg²⁺, Cd²⁺, Cu²⁺) and mycotoxins (patulin, ochratoxin A) in complex matrices, significantly reducing their in vivo bioavailability. In addition, yeast can serve as a protective carrier for sensitive bioactive substances, improving their gastrointestinal stability and bioavailability (e.g., the bioavailability of tea polyphenols increased from 12.2% to 73.2%). Biosorption is further upgraded via surface‐display genetic engineering, alkaline pretreatments (improving polyphenol loading by 93.2%), and physical intensification (ultrasonic treatment or vacuum perfusion, etc.). Ultimately, transitioning to industrial food applications requires characterization of complex yeast‐matrix interactions and the strategic utilization of active metabolic networks in viable yeast platforms.

1. Introduction

Despite sustained growth in global food production, challenges related to contamination and nutritional quality remain significant. Heavy metals, originating from industrial emissions and agricultural inputs, can accumulate in fruits via plant uptake, thereby contaminating derived products such as fruit juices and wines. Concurrently, these metals can also enter the dairy supply chain when livestock consume contaminated water and feed, leading to their presence in milk and other animal products (Roky et al. 2026; Abdollahimajd et al. 2026). In addition, improper storage and handling further elevate the risk of mycotoxin contamination—for instance, aflatoxins (AFs) may arise in milk when cattle ingest moldy feed, whereas Ochratoxins can occur in wines produced from contaminated fruits (Jiang et al. 2021; Ortiz‐Villeda et al. 2021). Furthermore, modern processing of staple foods often results in reduced levels of naturally occurring bioactive compounds, creating a growing demand for functional ingredients and effective delivery systems (F. Zhang et al. 2025).

In this context, yeast biosorption has gained increasing attention in food systems. Yeast is the cornerstone of the modern food, beverage, and baking industries. Advances in the understanding of yeast cellular structure and proteome have expanded its functional applications beyond fermentation (Botstein et al. 1997). In particular, yeast cells can sequester exogenous compounds through surface biosorption and intracellular uptake, providing a basis for applications in both food safety and nutritional enhancement (Aeini et al. 2025). To date, yeast biosorption has demonstrated effective removal of pollutants from fluid food matrices; for example, Saccharomyces cerevisiae biomass has shown effective biosorption of toxic heavy metals, including Pb^{2 +} and Cd^{2 +}, as well as mycotoxins such as aflatoxins and patulin (PAT), in contaminated fluid food matrices (Fateminasab et al. 2024; Alimi et al. 2025). Beyond contaminant removal, yeast cells have been used as natural carriers for bioactive compounds such as polyphenols, vitamins, and lipophilic nutrients, enabling their adsorption, protection against degradation, and controlled release in food formulations (Dadkhodazade et al. 2021; Gil‐Martín et al. 2022).

The effectiveness of yeast in binding contaminants is largely attributed to the physicochemical properties of its cell wall (CW) components, including polysaccharides, proteins, and lipids. These components enable diverse interactions with a wide range of chemical species, even under complex food matrix conditions (R. Massoud, Khosravi‐Darani, et al. 2019; X. Liu et al. 2020). Key advantages of yeast‐based biosorbents include low cost, food‐grade safety, and compatibility with heterogeneous food environments (Fomina and Gadd 2014; J. Wang and Chen 2006). Furthermore, the majority of yeast strains employed in relevant studies have a well‐established history of use in the food industry and are recognized as food ingredients or additives in major jurisdictions, thereby offering enhanced safety assurance and facilitating regulatory approval and market translation.

Despite growing interest in yeast‐based biosorption, research in food systems remains fragmented. For heavy metals, studies have largely emphasized environmental remediation over food safety (J. Wang and Chen 2006; Sagar Jena et al. 2022). Regarding mycotoxins, biosorption has typically been treated as a secondary pathway alongside enzymatic degradation, with limited targeted synthesis (Alimi et al. 2025; Jiang et al. 2024). Few reviews have comprehensively integrated yeast biosorption in food systems from mechanistic principles to practical applications. To address these gaps, this narrative review presents a structured synthesis of yeast biosorption in food‐related applications. It first outlines the structural and physicochemical mechanisms underlying biosorption, followed by a discussion of its roles in the removal of heavy metals and mycotoxins, as well as in the extraction, stabilization, and delivery of bioactive compounds. Finally, strategies to enhance biosorption performance—including cell surface engineering, strain selection, and pretreatment approaches—are critically evaluated.

2. Materials and Methods

This narrative review provides a comprehensive overview of yeast biosorption in food systems. A literature search was performed in ScienceDirect, PubMed, and Wiley Online Library using following the keywords and their combinations: “yeast,” “S. cerevisiae,” “biosorption,” “heavy metal,” “mycotoxin,” “bioactive compounds,” “food matrix,” “wine,” “beverage,” coupled with additional terms for enhancement strategies such as “cell surface display” and “genetic engineering.” Additional records were identified by screening reference lists of included studies and relevant reviews. After duplicate removal, titles and abstracts were screened, yielding 238 articles for full‐text assessment. Studies were selected based on relevance to yeast biosorption in food‐related matrices, prioritizing those reporting biosorption capacity, kinetics, isotherms, or enhancement strategies. Studies on environmental wastewater treatment or those addressing non‐biosorption mechanisms were deprioritized. A narrative synthesis was subsequently conducted, organizing findings by mechanistic insights, application domains, and enhancement strategies.

3. The Material Basis and Potential Mechanisms of Yeast Biosorption

3.1. The Structure of Yeast Cell

The yeast cell wall (CW) is the primary mediator of biosorption, accounting for 15%–30% of the cellular dry weight. It is predominantly composed of polysaccharides (80%–90%), with minor, strain‐specific fractions of proteins and lipids (Lipke and Ovalle 1998; Sentandreu et al. 1984). Structurally, the CW exhibits a bilayered architecture: an inner scaffolding and an outer electron‐dense layer. The inner layer consists of a three‐dimensional lattice of β‐1,3‐glucan cross‐linked with β‐1,6‐glucan and chitin, providing mechanical rigidity. Glycosylphosphatidylinositol (GPI)‐dependent proteins and proteins with internal repeats (Pir) are anchored to this glucan matrix via covalent or alkali‐sensitive linkages. The outer layer is dominated by highly glycosylated mannoproteins (> 90% glycan content). Phosphodiester bonds within the O‐ and N‐linked mannan side chains impart a net negative electrostatic charge to the cell surface, facilitating the sequestration of cationic heavy metals and mycotoxins (Figure 1B) (Klis et al. 2002).

FIGURE 1.

The yeast cell structure and mechanisms of yeast biosorption. (A) Yeast biosorption of metal ions, mycotoxins, and active compounds. (B) Yeast cell surface structure contributing to biosorption. (C) The mechanism of yeast surface biosorption (ZEA, zearalenone, AFB1, Aflatoxin B1). (D) The mechanism of yeast intracellular bioaccumulation.

The structure–function relationship of the CW in biosorption was demonstrated in Pichia pastoris X33 (Chen et al. 2021). While intact cells achieved a Cu²⁺ removal efficiency of 41.1% (6.2 mg/g), mechanical disruption increased the capacity to 16.13 mg/g, likely due to the unmasking of cryptic binding sites within the inner ultrastructure. Fractional analysis revealed that isolated CWs possessed a higher biosorption capacity (11.53 mg/g) and faster kinetics (equilibrium within 15 min) compared to cytoplasmic fractions (8.87 mg/g; 30 min). Furthermore, the mannan fraction exhibited significantly higher removal efficiency (34%) than the glucan fraction (12%). These findings underscore the CW, particularly its mannan‐rich outer layer, as the principal functional site for the physicochemical sequestration of exogenous molecules.

3.2. The Potential Mechanisms of Yeast Biosorption

The sequestration of exogenous molecules by yeast is a complex biphasic process comprising metabolism‐independent surface biosorption and metabolism‐dependent intracellular bioaccumulation. These pathways differ significantly in their kinetics, energy requirements, and the specific molecular interactions involved.

3.2.1. Metabolism‐Independent Surface Biosorption

Surface biosorption is a rapid, passive process that occurs at the cell envelope, predominantly driving sequestration when extracellular target concentrations are high. This phenomenon involves non‐covalent interactions between exogenous ligands and functional groups within the CW matrix, such as amine, carboxyl, and sulfhydryl groups, alongside physical entrapment within the porous ultrastructure (Jiang et al. 2024). A critical component of this phase is the ion exchange mechanism, where the binding of metal cations to negatively charged polysaccharides is coupled with a concomitant stoichiometric efflux of protons (H⁺) or light metal ions from the CW. Because this process is purely physicochemical and requires no metabolic energy, the sequestration capacity remains functionally conserved across living, heat‐inactivated, and nonviable yeast biomass. Furthermore, the reversible nature of these surface interactions allows for the chemical desorption of sequestered heavy metals using chelating agents like EDTA, which facilitates both the regeneration of the yeast biosorbent and the recovery of captured ions for downstream applications (Shao et al. 2025).

3.2.2. Factors Affecting the Efficiency of Passive Biosorption

The efficacy of passive biosorption is profoundly dictated by the macro‐morphology and biochemical composition of the CW. Morphologically, yeast strains exhibiting a higher CW thickness‐to‐cell diameter ratio demonstrate robust sequestration capacities, particularly for mycotoxins such as ochratoxin A (OTA) and zearalenone (ZEA), whereas lower ratios correspond to reduced binding efficiency (Armando et al. 2012). Architecturally, the conformation of CW polysaccharides is a decisive factor; for instance, alkali‐insoluble fractions of β‐d‐glucan that adopt single‐ or triple‐helix conformations—rather than highly cross‐linked networks with chitin—enhance the steric accessibility of reactive functional groups, thereby augmenting ZEA‐binding capacity (Yiannikouris et al. 2004). In the context of bioactive compounds, the yeast CW acts as a versatile functional scaffold for the encapsulation of sensitive molecules such as polyphenols, anthocyanins, and carotenoids (Dadkhodazade et al. 2021; Stafussa et al. 2016). This sequestration is primarily driven by hydrophobic interactions within the protein–lipid microdomains and extensive hydrogen bonding between the hydroxyl groups of phenolic compounds and the oxygen atoms of the glucan‐mannan backbone. Biochemically, the distribution of polysaccharides, proteins, and lipids confers distinct electrostatic and hydrophobic properties to the yeast cell surface. In general, a higher net negative surface charge is associated with enhanced sequestration of divalent cations (e.g., Pb²⁺, Cd²⁺, and Cu²⁺), primarily through electrostatic interactions and ion exchange. By contrast, increased surface hydrophobicity may reduce binding affinity for these metal ions but can promote interactions with lipophilic bioactive compounds (Kordialik‐Bogacka 2011). Mechanistically, heavy metal binding is dominated by charge‐driven interactions, whereas the sequestration of mycotoxins and bioactive molecules relies more on hydrophobic interactions, hydrogen bonding, and physical entrapment within the CW matrix (Figure 1C).

3.2.3. Metabolism‐Dependent Intracellular Bioaccumulation

In contrast to surface adsorption, intracellular bioaccumulation emerges as the predominant pathway when extracellular ions are present at trace levels. This metabolism‐dependent, energy‐driven process relies on the coordinated action of plasma membrane transporters, intracellular metallochaperones, and sink enzymes to internalize and compartmentalize ions (J. Wang and Chen 2006). Using copper as a paradigm, uptake in S. cerevisiae is mediated by a highly regulated sequential process (Figure 1D). Extracellular Cu²⁺ is initially reduced to Cu⁺ by cell surface metalloreductases, Fre1 and Fre2, followed by internalization via high‐affinity transmembrane proteins Ctr1 and Ctr3. The extracellular N‐terminus of Ctr1 is enriched in methionine and serine residues, providing specific coordination sites for Cu⁺ (Puig et al. 2002). While early physiological studies hypothesized that uptake was coupled to K⁺ efflux, contemporary structural analyses clarify that Ctr1 functions as a selective Cu⁺ channel, with K⁺ efflux representing a secondary charge compensation response to maintain electrochemical gradients (De Rome 1987; Ren et al. 2019). Once internalized, Cu⁺ is rapidly sequestered by intracellular metallochaperones, such as Cox17 for mitochondrial delivery and Lys7 for incorporation into Cu/Zn superoxide dismutase (Sod1). This efficient intracellular trafficking maintains an extremely low level of free cytosolic Cu⁺, thereby preserving a strong thermodynamic gradient that favors continued uptake (Shi et al. 2021).

To avoid toxicity, copper homeostasis is tightly regulated. Under copper‐replete conditions, transcription of CTR1/2/3 is downregulated, and surface Ctr1 proteins are rapidly degraded, whereas copper limitation induces the mobilization of vacuolar copper stores via Ctr2 (L. Liu et al. 2012; Yonkovich et al. 2002). Notably, the relatively broad substrate specificity of these metal transport systems also permits the uptake of nonessential or toxic metal ions. For instance, the low‐affinity iron transporter Fet4 can mediate the accumulation of cations such as Ni²⁺, Cd²⁺, and Co²⁺, which may compete with Fe²⁺ for binding (Hassett et al. 2000; Eide 1998). This functional flexibility provides a mechanistic basis for the capacity of yeast to accumulate toxic elements from the environment, even at low concentrations.

3.3. Kinetics and Equilibrium Isotherms of Yeast Biosorption

The quantitative assessment of yeast biosorption relies on kinetic and isotherm frameworks to differentiate between transport‐limited and reaction‐limited processes (Aeini et al. 2025). Kinetic modeling, primarily through pseudo‐first‐order (PFO) and pseudo‐second‐order (PSO) equations, provides insights into the rate‐limiting steps of sequestration. In this approach, residual adsorbate concentrations measured at different time points are fitted to the kinetic models, and the slopes and intercepts derived from these equations yield the kinetic parameters. PFO models exhibit multi‐linear characteristics when ln(q_e  − q_t ) is plotted against time (t), revealing a segmented adsorption pathway. The initial rapid phase typically corresponds to the external mass transfer and surface accumulation of target molecules, leading to monolayer formation. This is followed by a secondary, attenuated phase governed by intraparticle diffusion or the structural rearrangement of sorbates within the porous yeast biomass. Despite the utility of multiphase PFO analysis, the PSO model is more commonly used in yeast biosorption kinetics, which often yields high coefficients of determination (R ²). This widespread empirical adoption suggests that the rate‐limiting step is often dominated by chemisorption involving valence forces through the sharing or exchange of electrons between yeast CW functional groups and the sorbate (Ho and McKay 1999). Thermodynamic equilibrium and biosorption capacities are evaluated using isotherm models, frequently the Langmuir and Freundlich equations, from which the mass of adsorbed solute required to completely saturate a unit mass of adsorbent (mg/g) can be derived using the corresponding equilibrium uptake values. Conformity to these models provides insights into the structural interaction between the sorbate and the yeast surface. Farhan and Khadom (2015) evaluated the transport of Pb²⁺, Cd²⁺, Cr³⁺, Co²⁺, and Zn²⁺ in S. cerevisiae employing kinetic and isotherm models. The kinetic study was conducted over a range of initial metal ion concentrations, with residual concentrations measured at five discrete time points during the reaction. Comparative regression analyses revealed that the PSO model provided a superior fit (R ² > 0.95) relative to the PFO equation. They identified a three‐stage adsorption kinetic profile, consisting of a rapid initial phase within the first few minutes, followed by two subsequent stages that are clearly separated by a plateau, the characteristics of which depend on the metal ion concentration and availability in the solution. To further characterize the thermodynamic equilibrium, these data were fitted to Langmuir and Freundlich isotherms, both of which achieved high predictive accuracy (R ² > 0.95). The conformity to the Langmuir model, in particular, indicates a relatively homogeneous distribution of active sites on the yeast surface, facilitating the formation of an adsorptive monolayer. Similar high‐fidelity results were reported by Y. Zhang et al. (2010), where Langmuir and Freundlich isotherms described the temperature‐dependent Cu²⁺ biosorption capacity with R ² values exceeding 0.95, confirming the stability of the adsorptive layer across varying thermal conditions.

The application of these mathematical frameworks extends to the recovery of bioactive compounds, where the choice of model often reflects the complexity of the yeast–sorbate interaction. Rubio et al. (2018) integrated PSO kinetics with Freundlich, Temkin, and Dubinin–Radushkevich isotherms to analyze the biosorption of polyphenols from grape pomace using pretreated yeast biomass. The results indicated that the optimal isotherm fit was highly sensitive to the pretreatment method (e.g., autoclaving, acid, or alkaline disruption), which modifies the surface heterogeneity and binding energy distribution of the biomass. For instance, while PSO kinetics remained robust across all sets (R ² > 0.85), the shift in isotherm fidelity highlighted how structural changes in the CW, such as the unmasking of β‐glucan or the degradation of mannoproteins, alter the thermodynamic affinity for bioactive molecules. Collectively, these kinetic and isotherm analyses provide the quantitative foundation necessary for scaling yeast‐based biosorption processes in food and environmental applications (Table 1).

TABLE 1.

Kinetic and isotherm models in yeast biosorption.

Model	Equation	Parameters	Reference
Kinetic
Lagergren first‐order reaction rate model (pseudo‐first order adsorption kinetics)	$\log (q_e-q_t)=\log q_e-\frac{K_1}{2.303}$	K 1: the equilibrium rate constant of pseudo‐first adsorption kinetics K h: the equilibrium rate constant of pseudo‐second adsorption kinetics q t: the amount of adsorbate on adsorbent at time t (mg/g) q e: the equilibrium uptake (mg/g)	(Farhan and Khadom 2015)
Ho's second‐order reaction rate model (pseudo‐second order adsorption kinetics)	$\frac{t}{q_{\mathrm{t}}}=\frac{1}{K_{\mathrm{h}}{q}_{\mathrm{e}}^2}+\frac{1}{q_{\mathrm{e}}}t$		(Farhan and Khadom 2015; Rubio et al. 2018)
Isotherm
Langmuir model	$q_{\mathrm{e}}=q_{\mathrm{m}}\frac{K_{\mathrm{L}}{C}_{\mathrm{e}}}{1+K_{\mathrm{L}}{C}_{\mathrm{e}}}$	q e: the amount adsorbed at time t (mg/g) q m: the mass of adsorbed solute completely required to saturate a unit mass of adsorbent (mg/g) C e: the equilibrium concentration (mg/L) K L: a constant related to the energy or net enthalpy of adsorption (L/mg)	(Farhan and Khadom 2015)
Freundlich model	$q_{\mathrm{e}}=K_{\mathrm{F}}{C}_{\mathrm{e}}^{\frac{1}{n}}$	q e: the amount adsorbed at time t (mg/g) C e: the equilibrium concentration (mg/L) K F: Freundlich equilibrium constants indicative of adsorption capacity n: Freundlich equilibrium constants indicative of adsorption intensity	(Farhan and Khadom 2015; Rubio et al. 2018)
Temkin model	$q_{\mathrm{e}}=\frac{RT}{b_{\mathrm{t}}}\ln A+\frac{RT}{b_{\mathrm{t}}}\ln C_{\mathrm{e}}$	q e: the amount adsorbed at time t (mg/g) C e: the equilibrium concentration (mg/L) A: equilibrium binding constant (L/g) RT/b t: Temkin constant (J/mol) related to the sorption energy	(Rubio et al. 2018)
Dubinin–Radushkevich model	$\ln q_{\mathrm{e}}=\ln q_{\mathrm{s}}-\beta {\epsilon }^2$	q e: the amount adsorbed at time t (mg/g) q s: Dubinin–Radushkevich monolayer capacity (mg/g) $\beta$: Dubinin–Radushkevich constant [mol2 /(KJ2)] $\epsilon$: Polanyi potential	(Rubio et al. 2018)

4. Biosorption of Heavy Metal by Yeast

4.1. Common Heavy Metal Hazards in Food

Heavy metal contamination in the food chain threatens food safety and quality. Plants accumulate toxic metals (Cd, Cr, Pb, Hg) via nonselective uptake pathways for essential micronutrients, a process exacerbated by contaminated agrochemicals and irrigation water (Ccopi et al. 2026). The toxicological mechanisms are metal‐specific: Cd accumulates in renal and hepatic tissues; Pb disrupts calcium signaling and displaces essential cations, posing neurodevelopmental risks; and Hg forms stable bonds with sulfhydryl groups, inducing neuro‐ and nephrotoxicity (Hu et al. 2016; Kanwal et al. 2017; Dadkhahfar et al. 2026). Even essential Cu becomes hazardous at elevated concentrations, particularly in fermentation, where residues inhibit yeast sugar transporters and glycolytic enzymes, leading to stuck fermentations and reduced ethanol yields (J. Wang et al. 2023). Collectively, these contaminants underscore the urgent need for effective mitigation strategies to protect both consumer health and food processing viability.

4.2. Yeast Biosorption of Heavy Metal Ions in Food Systems

Yeast‐mediated biosorption of heavy metals in food systems spans a continuum from in situ fermentation processes to engineered treatment systems and increasingly complex food matrices, all governed by a shared mechanistic basis involving coupled surface binding and intracellular sequestration.

4.2.1. In Situ Sequestration During Fermentation Processes

In fermented beverages, S. cerevisiae serves a dual role as both the fermentative agent and an in situ biosorbent. Research in viticultural matrices has highlighted a complex interplay between yeast vitality and metal removal. Early studies on enological strains such as SN9 and SN41 demonstrated significant Cu²⁺ attenuation in grape must, though the relative contributions of surface adsorption versus intracellular accumulation remained a point of contention (Brandolini et al. 2002). Initial investigations using TEM‐EDS on brewing strains under high Cu²⁺ loads suggested that sequestration was localized exclusively at the cell envelope via passive ion exchange (X.‐Y. Sun et al. 2015). However, subsequent evaluations under realistic physiological conditions (≤ 19.2 mg/L Cu²⁺) revealed a dynamic temporal shift: Cu²⁺ initially localized at the cell surface but was internalized into the cytosol as fermentation progressed (X. Sun et al. 2016). The significantly higher removal efficiency of metabolically active cells (72%–81%) compared to nonviable biomass (17%–23%) underscores that while passive adsorption initiates the process, active metabolic transport ultimately dominates sequestration at trace metal levels. This mechanism is critical for the beverage industry, as it not only remediates contaminants but also prevents the “stuck” fermentations typically associated with copper‐induced metabolic repression.

4.2.2. Ex Situ Engineered Systems for Liquid Food Remediation

Building upon the mechanisms observed in fermentation systems, yeast has also been developed as a dedicated biosorbent in engineered, ex situ treatment processes for liquid food streams. Continuous‐flow biosorption utilizing active yeast biomass has demonstrated robust capacity for Cu²⁺ and Pb²⁺ across concentrations ranging from 10 to 180 mg/L (Amirnia et al. 2015). These systems typically adhere to the Langmuir isotherm model, indicating a structured, monolayer‐based sequestration process. The superior performance of living cells in these continuous setups corroborates findings from wine matrices, confirming a bipartite mechanism where initial physicochemical surface binding is synergistically enhanced by active intracellular bioaccumulation. The development of such autonomous biotechnological strategies provides a pragmatic, cost‐effective alternative to traditional chemical precipitation, particularly for treating large volumes of aqueous food ingredients where trace metal purity is paramount.

4.2.3. Biosorption Optimization in Complex Food Matrices

Beyond relatively simple liquid systems, the application of yeast biosorption is increasingly being extended to complex food matrices, such as dairy products, where high levels of proteins and lipids introduce additional challenges. Navigating the competitive binding and matrix interference inherent in these systems requires advanced statistical modeling. Response surface methodology (RSM) with central composite design (CCD) has proven essential for optimizing parameters like pH, biomass density, and contact time to achieve high‐efficiency sequestration at parts‐per‐billion (ppb) levels. For instance, optimized yeast‐based processes have achieved nearly 90% removal efficiency for Hg²⁺ in refined systems (Hadiani et al. 2018). Similarly, systematic CCD‐based studies in milk matrices have successfully mitigated Hg²⁺, Pb²⁺, and Cd²⁺ with efficiencies exceeding 70% (Table 2) (Massoud et al. 2019, 2020, 2021). These advancements demonstrate that with precise parametric optimization, yeast‐based biosorption can be effectively tailored to ensure the safety of diverse and chemically complex agricultural commodities.

TABLE 2.

Yeast absorbents for toxic metal ions removal in food systems.

Yeast (status)	Strain number	Matrices	Initial yeast concentration	Initial metal ion concentration	Experimental condition	Removal rate (or other reported adsorption metrics)	Reference
Saccharomyces cerevisiae (active)	SN41, SN9	Sterilized grape must	104 cells/mL	Cu2+: 32 ppm (mg/L) and 320 ppm	100 mL of red grape must, 25°C, fermentable sugar 19%, pH 3.15, 7 days	SN41: 89.07% (32 ppm initial Cu2+ concentration), 39.31% (320 ppm initial Cu2+ concentration)	(Brandolini et al. 2002)
	BH8, AWRI R2, Freddo	Model synthetic medium simulating components of standard grape juice	106 cells/mL	Cu2+: 0.50 mM (32 mg/L), 1.00 mM (64 mg/L) and 1.50 mM (96 mg/L)	400 mL MSM,28°C, 120 r/m in thermostatic shaker, 14 days	AWRI R2: highest removal rate (67.37%) in 0.50 mM Cu and highest adsorption efficiency (15.82 mg/g) in 1.50 mM Cu	(X.‐Y. Sun et al. 2015)
	BH8, AWRI R2, Freddo	Model synthetic medium simulating components of standard grape juice	5 × 105 cells/mL	Cu2+: 0.15 mM (9.6 mg/L), 0.20 mM (12.8 mg/L), 0.25 mM (16.0 mg/L), and 0.30 mM (19.2 mg/L)	400 mL MSM,28°C, 120 r/m in thermostatic shaker, 14 days	AWRI R2: 81% (0.20 mM initial Cu2+ concentration), 76% (0.25 mM initial Cu2+ concentration), 72% (0.3 mM initial Cu2+ concentration)	(X. Sun et al. 2016)
	47 S. cerevisiae strains	Sulfited grape must	107 cells/mL	Cu2+: 300 µmol/L	100 mL sulfited (50 mg/L) grape must, 240 g/L sugar, pH 3.5, 26°C	MPR2‐24: 2.050 mg/kg copper residual content detected in cells	(Capece et al. 2018)
	PTCC 5020	Milk	Pb: 22 × 108 CFU, Cd: 30 × 108 CFU, Hg: 22 × 108 CFU	Pb: 70 µg/L, Cd: 80 µg/L, Hg: 80 µg/L	4°C–40°C, 0–50 rpm shaking rate (both temperature and shaking rate show insignificant effects on removal rate), Pb: 4 days, Cd: 4 days, Hg: 3 days	70% Pb, 70% Cd, 70% Hg (highest removal rate)	(Massoud et al. 2019; Massoud et al. 2020, Massoud et al. 2021)

Collectively, these findings establish that heavy metal biosorption proceeds via a bipartite mechanism, involving an initial phase of passive physicochemical adsorption onto CW constituents, followed by a subsequent phase of energy‐dependent intracellular transport. Active yeast cells are particularly effective for sequestering trace heavy metal concentrations in food matrices, with the removal process occurring concurrently with fermentation and potentially extending over several days until equilibrium is attained. Importantly, elevated heavy metal ion concentrations can compromise yeast metabolic activity, which in turn may adversely affect both fermentation kinetics and biosorption efficiency.

5. Biosorption of Mycotoxins by Yeast

5.1. Hazards of Mycotoxins in Food

Mycotoxins—low‐molecular‐weight metabolites from Aspergillus, Penicillium, and Fusarium—pose a pervasive threat to global food safety, entering the food chain through infected staple crops and carrying over into animal‐derived products (Alimi et al. 2025). Detection rates for ZEA and deoxynivalenol (DON) frequently reach 60%–80% in certain regions, exacerbated by climate‐driven shifts in fungal distribution (Eskola et al. 2019; Zhu et al. 2026). Among diverse mycotoxins, AFs, OTA, ZEA, and PAT exemplify distinct toxicological mechanisms: AFB₁, a potent hepatocarcinogen (IARC Group 1), causes hepatic damage and genotoxicity (Martínez et al. 2023; Jiang et al. 2021); OTA induces nephrotoxicity via mitochondrial disruption and protein synthesis inhibition (Khoury and Atoui 2010); ZEA acts as an endocrine disruptor through estrogen receptor binding (El‐Sayed et al. 2022); and PAT inactivates enzymes via sulfhydryl group modification, causing gastrointestinal (GI) and neurological damage (Saleh and Goktepe 2019). Despite structural diversity, these mycotoxins share the capacity to compromise food safety, necessitating mitigation strategies throughout the food supply chain.

5.2. Yeast Biosorption of Mycotoxins in Food Systems

Mycotoxin mitigation strategies are categorized into physical, chemical, and biological interventions. While conventional physical and chemical methods often suffer from low efficiency and high costs, biological strategies, specifically microbial biosorption, offer a sustainable alternative due to their high molecular specificity and environmental compatibility (Guan et al. 2021). Yeast, particularly S. cerevisiae, presents distinct advantages as a Generally Recognized as Safe (GRAS) agent. Unlike bacteria that may produce metabolites of concern, yeast is nonallergenic and genetically stable (Table 3) (Jiang et al. 2024).

TABLE 3.

Yeast absorbents for mycotoxin removal in food systems.

Yeast (status)	Strain number	Matrices	Initial yeast concentration	Initial mycotoxin concentration	Experimental condition	Removal rate (or other reported adsorption metrics)	Reference
Saccharomyces cerevisiae (cell wall components)	WT292, FKS1, MNN9, SC1026	Water	100 µg/mL	ZEA: 2 µg/mL, 4 µg/mL, 6 µg/mL, 8 µg/mL, 10 µg/mL, 20 µg/mL	37°C, 200 rpm in rotary shaker, 1.5 h	Extracted alkali‐insoluble ‐d‐glucan fraction: maximal amount of toxin bound:14.53 µg/mL, Affinity rate: 50.4%	(Yiannikouris et al. 2004)
Saccharomyces cerevisiae, Torulaspora delbrueckii, Schizosaccharomyces pombe, and so forth (active)	S10c (S. cerevisiae), IMIAT‐70 (Torulaspora delbrueckii), Schp.3 (Schizosaccharomyces pombe), and so forth	White wine must, red wine must	106 cells/mL	OTA: 2 mg/L	20°C, 36 days	IMIAT‐70: 46.83% to 52.16 in white wine, Schp.3: 53.21%–70.23% in red wine	(Cecchini et al. 2006)
Saccharomyces cerevisiae (in beer fermentation residue)	N/A	Water	108 cells/mL	2.0 µg/mL of each mycotoxin (AFB1, ZEA, OTA, DON)	10 mL of buffer solution (pH 3.0 or pH 6.5), 25°C, 60 min	75.1% at pH 3 and 77.5% at pH 6.5 for ZEA, lower binding capacity for AFB1, OTA, DON (less than 60% and 40% at pH 3.0 and 6.5, respectively)	(Campagnollo et al. 2015)
Kazachstania servazzii (extracted from Kefir grains, active)	KFGY7	Milk	1%, w/v	11.0 µg/mL of each mycotoxin (AFB1, ZEA, OT)	25°C, 24 h	74% AFB1, 62% OTA, 95% ZEA	(Taheur et al. 2017)
Saccharomyces cerevisiae (inactive, nanoparticle nano‐Fe3O4 functionalized)	YS‐3	Apple juice	0.1 g/L	PAT: 0.5 mg/L	30 mL patulin working solution, rotary shaker 150 rpm, 30°C, about 75 h	Adsorption capacity 8.63 µg/g	(Qiu et al. 2018)

The yeast cell envelope is the primary functional site for mycotoxin sequestration, as evidenced by comparative studies of viable cells, heat‐inactivated biomass, and protoplasts. Investigations into PAT removal by S. cerevisiae CCTCC 93161 revealed that intact cells exhibit maximal efficiency, while protoplasts show significantly lower capacity (p < 0.05), confirming the necessity of the CW matrix (Z. Zhang et al. 2019). Sequestration kinetics typically display a biphasic profile: an immediate passive adsorption phase followed by a slower, time‐dependent phase that may involve enzymatic biotransformation. Fourier‐transform infrared (FTIR) spectroscopy has identified specific participation of hydroxyl (‐OH), amine (‐NH), and C–O groups within CW polysaccharides in the binding process, as indicated by characteristic bathochromic shifts in their respective absorption bands during toxin interaction. Consequently, the biosorption capacity of yeast can be significantly augmented by modulating CW biosynthesis. For example, exogenous supplementation with calcium ions (Ca²⁺) at an optimal concentration of 10 mM has been shown to maximize CW yields and increase the content of β‐1,3‐glucan and β‐1,6‐glucan (Luo et al. 2019). This structural enrichment is mediated by the activation of the calcineurin‐dependent transcription factor Crz1, which upregulates key CW biosynthesis genes, including MID1 and CRZ1. This transcriptional cascade enhances the activity of β‐1,3‐glucanase and β‐1,3‐glycosyltransferase, leading to a more robustly cross‐linked glucan network. Such structural modifications have successfully improved PAT removal rates in both aqueous solutions and complex matrices like simulated apple juice.

Besides effective biosorption ability for mycotoxins, for effective detoxification, the yeast–mycotoxin complex must remain stable under physiological GI conditions. In studies involving kefir‐derived yeast strains like Kazachstania servazzii KFGY7, the yeast–mycotoxin complex released an average total rate of merely 33% of its bound toxins, compared with 65% for whole kefir grains and 46% and 55% for two high‑adsorption‑efficiency kefir‑derived lactic acid bacteria strains. This high retention confirms that the yeast CW effectively sequesters mycotoxins throughout intestinal transit, thereby facilitating safe fecal excretion (Taheur et al. 2017). While initial binding may be pH‐dependent—driven by the protonation state of mycotoxin carboxyl groups and their subsequent hydrogen bonding with the β‐glucan matrix—yeast cell wall extracts (YCWE) maintain significant biosorptive capacity even after sequential proteolytic treatment (Vartiainen et al. 2020). In vivo trials in poultry models have demonstrated that YCWE supplementation (4 kg/ton) can reduce the hepatic burden of OTA by approximately 30%. This protective effect is attributed to the ability of the yeast CW to limit the intestinal absorption and systemic bioavailability of mycotoxins, facilitating their excretion.

Collectively, in contrast to heavy metal biosorption, yeast‐mediated mycotoxin sequestration relies primarily on passive physicochemical adsorption onto CW constituents rather than intracellular uptake. These findings establish yeast CW preparations as robust, versatile, and biocompatible agents for mycotoxin decontamination, with demonstrated efficacy across both aqueous model systems and complex food matrices, alongside promising performance in animal trials. Concurrently, enzymatic biotransformation by metabolically active cells may further contribute to overall mycotoxin reduction (P. Wu et al. 2026).

6. Biosorption of Food Bioactive Compounds by Yeast

6.1. Sequestration of Bioactive Compounds

The yeast CW and cytoplasm contain abundant functional groups—including proteins, polysaccharides, and lipids—that effectively coordinate with polyphenolic ligands, thereby rendering yeast biosorption a promising approach for extracting bioactive compounds from food matrices (Table 4). The yeast CW and cytoplasm contain abundant functional groups—including proteins, polysaccharides, and lipids—that effectively coordinate polyphenolic ligands. Tao et al. (2019) utilized ultrasound‐assisted biosorption (394.2 W/L) with spent brewer's yeast to recover blueberry anthocyanins, demonstrating that sequestration efficiency is profoundly dictated by the target molecule's stereochemical conformation. Chromatographic analyses revealed that malvidin‐3‐O‐glucoside exhibited a higher affinity (34.4%) than its galactoside counterpart (25.0%). Within the flavonol subclass, hyperoside achieved a maximal efficiency of 39.8%, significantly outperforming rutin (25.0%) and myricetin (24.2%). Spectroscopic evidence (FTIR and XRD) confirmed the physical intercalation of these phenolic ligands into the amorphous yeast matrix, characterized by wavenumber shifts in the C═C aromatic stretching region (1650 cm⁻¹).

TABLE 4.

Yeast absorbents for bioactive compounds extraction.

Yeasts (status)	Matrices	Initial yeast concentration	Target bioactive compounds	Initial concentration	Experimental condition	Adsorption rate (or other reported adsorption metrics)	Reference
Baker's yeasts (dehydrated)	Unclarified sherry type white wines	2.8 g/L	Phenolic compounds	16.3 mg/L	Browning wine, 15.5% ethanol content, 20°C, 24 h	Significant adsorption for six phenolic compounds (catechin, epicatechin, procyanidin, etc.), highest adsorption rate for procyanidin B3 (about 32%)	(Razmkhab et al. 2002)
Saccharomyces cerevisiae (4CV, 3VA, 7EV, etc.), Saccharomyces uvarum (S6U) (active)	Crushed grapes	3 × 106 CFU/mL	Anthocyanins and derivatives	430.47 mg/L	5 L crushed grapes, target product with an alcohol level of 13.5% v/v and a pH of 3.6	3.07% for total anthocyanin and derivatives, 16.24% for p‐coumaryl	(Morata et al. 2003)
Saccharomyces cerevisiae (yeast protein extracts)	Model wine	154 mg/L yeast proteins	Red wine polyphenols	2 g/L	2 mL samples, 12% v/v ethanol, 2 g/L tartaric acid, pH 3.5, 50 mM ionic strength, 25 mg/L SO2, room temperature, 24 h	17% for grape skin tannins (devoid of monomers and oligomers, presented a high mean degree of polymerization), 2.5% for red wine polyphenols (free anthocyanins and their derivatives, cinnamic acids, flavonols, flavanol monomers), 8% for red wine tannins (devoid of monomers and oligomers, mainly composed of oxidation products and tannin‐bound anthocyanins) at an initial polyphenol concentration of 2 g/L	(Mekoue Nguela et al. 2016)
Saccharomyces cerevisiae (waste yeast with alkaline pretreatment)	Grape pomace	4 g/L	Phenolic compounds	50 g/L grape pomace	12.5 mL sample, 40% (v/v) ethanol, thermostatic shaking bath at 150 rpm and 25°C, 360 min	Highest adsorption capacity: 190.28 mg/g	(Rubio et al. 2018)
Saccharomyces cerevisiae (waste yeast)	Phenolic mixture from blueberry	6 g/L	Phenolic compounds	1600 mg/ L	50% ethanol, pH 2.0 in solvent (without blueberry pomace extract), 20‐kHz sonication acoustic energy density 394.2 W/L, 40°C, 120 min	Maximum adsorption capacity: 186.1 mg/g	(Tao et al. 2019)

The protective encapsulation capacity of yeast was further elucidated by Ribeiro et al. (2021) using S. cerevisiae to sequester tea polyphenols. Under optimized conditions (alkaline pretreatment, pH 10.0), equilibrium capacities reached 7.07 mg/g for white tea and 6.22 mg/g for green tea. Attenuated total reflectance mid‐infrared (ATR‐MIR) spectroscopy corroborated this capture through a significant absorbance differential at the 3291 cm⁻¹ band, corresponding to the O–H stretching vibrations characteristic of flavonoid phenolic hydroxyl groups. These findings substantiate that the structural configuration of disparate phenolic compounds fundamentally governs their differential binding kinetics within the yeast matrix (Table 4). Recently, Pedro et al. (2026) employed brewer's spent yeast (BSY) (S. cerevisiae) to recover bioactive compounds from ethanolic extracts of fruit juice residues. The saline‐treated yeast exhibited a biosorption capacity of 51.97 mg/g for total phenolic compounds in ethanolic grape extracts (Pedro et al. 2026).

6.2. Enhancement of Stability and Bioaccessibility

In addition to effectively coordinating bioactive compounds via physicochemical adsorption, the ultrastructure and biomacromolecules of yeast cells facilitate their physical retention and stabilization, thereby creating a protective microenvironment that shields them not only from oxidative and thermal degradation within food matrices but also from digestive enzymes and pH fluctuations in the oral cavity and GI tract (Dikit et al. 2010; J. Wu et al. 2015; Kofuji et al. 2012; Kavetsou et al. 2019).

The resulting yeast‐bioactive complexes offer significant functional advantages regarding GI stability and targeted delivery. In the aforementioned studies on yeast biosorption for bioactive compound extraction, Ribeiro et al. (2021) reported that yeast‐encapsulated tea polyphenols achieved a bioaccessibility of 73.2%, compared to only 12.2% for the free extracts. Similarly, Pedro et al. (2026) found that phenolic compounds adsorbed onto untreated and saline‐treated yeast exhibited bioaccessibility values of 24.73–68.79% and 33.72–59.76%, respectively—substantially higher than the bioaccessibility observed for ethanolic extracts, which remained below 8% during simulated GI digestion. Young et al. (2020) corroborated this protective capacity using multimodal imaging, showing that high bile salt concentrations (HBS, 25 mg/mL) triggered a rapid burst release of curcumin (> 80% within 1 h) without matrix disintegration, whereas gastric pepsin exerted no discernible effect on CW integrity. Mechanistically, curcumin is likely associated with the hydrophobic regions of biopolymers located in the cell interior. Low concentrations of bile salts may be insufficient to perturb the conformational folding of these structures, thereby failing to solubilize the bioactive compound.

Yeast biosorption also provides a robust strategy for the stabilization of sensitive vitamins, which lack de novo synthesis pathways in humans and are highly susceptible to degradation during thermal processing and irradiation. The yeast matrix functions as a biological carrier that shields these micronutrients within its lipid‐protein microdomains. Soares et al. (2025) successfully encapsulated Vitamin D3 within BSY biomass through vacuum biosorption. The fit of the isotherm models suggests that the interaction between nonpolar vitamin D₃ and yeast cells was governed by van der Waals forces. In vitro simulated digestion assays demonstrated that the BSY microparticles completely shielded Vitamin D3 throughout the oral and gastric phases, followed by a targeted burst release of 69.14% strictly within the intestinal phase—the primary physiological site of absorption. This encapsulation strategy ensures that the bioactives remain stable and bioavailable, reinforcing the nutritional integrity of agricultural commodities without compromising their structural profiles.

6.3. Capture of Volatile Flavor Compounds

For food flavor enhancement, yeast encapsulated volatile compounds must rapidly release their payload upon introduction to the oral cavity—a microenvironment characterized by distinct shifts in hydration, salivary enzyme activity, and inorganic salt composition. This release requirement is fundamentally different from that of oral drug delivery systems, which typically prioritize gastric stability and targeted intestinal release. Consequently, the sequestration and subsequent release of volatile flavor compounds are governed by mechanisms primarily determined by the selective permeability of the yeast CW and the thermodynamic partition coefficient (log P) of the flavor ligands. The outer mannoprotein layer typically limits diffusion to molecules with a molecular weight below 700 Da and a hydrodynamic radius less than 0.85 nm (Scherrer et al. 1974). Dardelle et al. (2007) established that lipophilic compounds (logP > 2.0) consistently achieve encapsulation efficiencies (EE) exceeding 50%. However, Pham‐hoang et al. (2016) demonstrated that molecular polarity often outweighs lipophilicity; for instance, hexanoic acid achieved an EE of 45% compared to only 4% for hexanal, despite similar logP values. Structural nuances further fine‐tune this affinity: C10 ethyl esters achieve a maximal EE of only 15%–20%, while the expanded hydrophobic surface area of branched ethyl esters and trans‐configured isomers facilitates superior matrix binding. Trans‐CW diffusion is primarily thermodynamically limited by the aqueous solubility of the volatiles within the hydrated pores, allowing for stable entrapment until hydration‐triggered release (Normand et al. 2005).

Yeast biosorbents serve dual functions in the extraction and encapsulation of bioactive compounds. Analogous to mycotoxin biosorption, this process is governed primarily by physicochemical adsorption onto the CW and membrane. Efficacy in extracting target compounds from complex food matrices, as well as encapsulating target compounds for protective delivery and controlled release, has been demonstrated in relevant studies. Beyond acting as a delivery vehicle, the yeast matrix (e.g., β‐1,3‐d‐glucan) provides intrinsic therapeutic synergy (Han et al. 2022; Yan et al. 2025). Given that endogenous enzymes in metabolically active cells may biotransform or degrade bioactive compounds, some studies have tended to employ inactivated whole cells or cytoplasm‐depleted CW particles to maximize stability and biosorptive capacity. Nevertheless, certain investigations have reported superior biosorption performance by intact, viable cells, a phenomenon that may be attributable to the additional binding sites provided by intracellular organelles.

7. Strategies for Enhancing Yeast Biosorption

7.1. Yeast Cell‐Surface Display–Based Metal Binding Engineering

Yeast cell‐surface display (YSD) relies on glycosylphosphatidylinositol‐anchored proteins (GPI‐APs) to immobilize exogenous proteins within the cell envelope (Chatterjee and Mayor 2001). Glucanase‐extractable CW proteins in S. cerevisiae are typically synthesized with a C‐terminal GPI anchoring signal that directs their covalent linkage to β‐1,6‐glucan biopolymers during maturation (Pittet and Conzelmann 2007). The agglutinin system is the most extensively utilized scaffold for YSD due to its stability and high display density. In this system, α‐agglutinin and a‐agglutinin mediate cell–cell adhesion during haploid mating. Engineering strategies utilize these scaffolds in two ways: (1) the α‐agglutinin system involves a direct translational fusion of the target protein to the C‐terminal GPI anchor, and (2) the a‐agglutinin system employs a heterodimer consisting of the GPI‐anchored Aga1p subunit and the target‐fused Aga2p subunit, which are covalently linked via interchain disulfide bridges (Figure 2A) (Lipke and Kurjan 1992; Kondo and Ueda 2004).

FIGURE 2.

Strategies for yeast biosorption enhancement. (A) Yeast surface display system using α‐agglutinin and a‐agglutinin. (B) Hybridization and trait segregation of yeast strains. (C) Pretreatment for yeast biosorption enhancement.

The surface display of high‐affinity motifs, such as hexa‐histidine (hexa‐His) tags and metallothioneins (MTs), significantly enhances the sequestration of divalent cations. Early proof‐of‐concept work by Kuroda et al. (2001) utilized the hexa‐His motif fused to α‐agglutinin (strain GPMH6). Under alkaline conditions (pH 7.8), GPMH6 maintained cellular proliferation in 4.0 mM Cu²⁺, whereas wild‐type (WT) and control strains (GPM) were inhibited at concentrations exceeding 1.0 mM. Quantitative analysis revealed an eightfold increase in surface Cu² biosorption capacity in GPMH6, with 50% of the metal localized to the extracellular surface. Competitive desorption using EDTA confirmed the reversibility of the binding, allowing for cyclic biosorbent regeneration.

To further amplify capacity, Kuroda and Ueda (2006) engineered tandem repeats of yeast metallothionein (YMT) (Kuroda and Ueda 2006). By integrating tetrameric (4×) and octameric (8×) YMT sequences into the pGPM vector, they proportionally increased Cd²⁺ sequestration. Stoichiometric binding assays, quantified via ICP‐AES, demonstrated that Cd²⁺ recovery for the 8× YMT strain was 8.7‐fold higher than that of the monomeric counterpart. This underscores that increasing the density of biologically evolved detoxification scaffolds is a viable strategy for maximizing surface‐mediated biosorption.

Combinatorial peptide libraries and in silico modeling provide powerful platforms for identifying metal‐binding motifs with high stereochemical specificity. Kotrba and Ruml (2010) demonstrated the display of short NP peptides derived from the CXXEE motif of Cupriavidus metallidurans for selective Pb²⁺ binding. At concentrations of 100–150 µM, engineered strains achieved 3.5‐ to 5.2‐fold increases in Pb²⁺ sequestration, removing up to 95% of the metal. Unlike the YMT system, multimerization of NP peptides did not result in a proportional increase in binding capacity, suggesting that sequestration is governed by site‐specific nucleation processes rather than simple increases in ligand density. Transmission electron microscopy further revealed localized microprecipitate formation on the CW, indicating that peptide‐mediated nucleation plays a central role in metal immobilization. While intact engineered cells lost their activity at concentrations above 300 µM due to toxicity, isolated engineered CWs retained high adsorption capacity even at 500 µM, highlighting the robustness of cell‐wall‐based biosorbents.

Beyond molecular‐level design, system‐level constraints also become critical under industrially relevant conditions. Recent advances have further expanded this concept through phage display‐derived peptide screening. S. Wang et al. (2024) identified a Cd^{2 +}‐binding peptide via biopanning against a Cd^{2 +}‐chelating resin. When displayed on the Aga2p system in S. cerevisiae EBY100, the engineered yeast showed a 35% increase in Cd^{2 +} sequestration (S. Wang et al. 2024). Moreover, immobilization within a calcium alginate hydrogel matrix further enhanced overall removal efficiency to 55.7%, enabling a viability‐independent and recyclable biosorption platform.

7.2. Intracellular Sequestration and Genetic Optimization Strategies

7.2.1. Rational Genetic Engineering via Metallothioneins

Beyond surface display, the intracellular expression of heterologous transgenes—specifically MTs from hyperaccumulator plants—provides a metabolic‐driven mechanism for enhancing heavy metal tolerance and sequestration. De Oliveira et al. (2020) demonstrated that the heterologous expression of the Populus trichocarpa metallothionein gene, PtMT2b, substantially mitigates Cd²⁺ toxicity in S. cerevisiae. Transgenic strains harboring a mutated variant, PtMT2b(Y) (featuring a single non‐synonymous point mutation), exhibited a 37% increase in cellular proliferation under 20 µM Cd²⁺ compared to WT controls. In phenotypic assays with a toxic dose of 50 µM Cd²⁺, the WT strain suffered complete growth arrest, whereas PtMT2b(Y) and WT PtMT2b(C) variants maintained active proliferation. Quantitatively, the PtMT2b(Y) strain achieved an intracellular Cd²⁺ sequestration capacity 30‐fold greater than that of the WT counterpart at 30 µM Cd²⁺. This intracellular sequestration strategy has been extended to other toxic elements, with recombinant S. cerevisiae expressing tailored MTs showing robust efficacy in concentrating both Cr⁶⁺ and Pb²⁺ (Table 5) (R. Zhang and Yi 2017, 2024).

TABLE 5.

Surface display for yeast biosorption enhancement.

yeast	strain number	displayed protein	target of biosorption	display system	enhancement result	reference
Saccharomyces cerevisiae	MT8‐1	Histidine oligopeptide (hexa‐His)	Copper ions	α‐Agglutinin	Surface‐engineered yeast adsorbed three to eight times more ions than parent strains and had higher copper resistance (4 mM) than parent (below 1 mM at pH 7.8); half of absorbent can be recovered by EDTA treatment	(Kuroda et al. 2001)
	MT8‐1	Yeast metallothionein (YMT) with 1, 2, 4, 8 encoding sequences repeat, respectively	Cadmium ion	α‐Agglutinin	Providing more binding sites for copper ions, engineered yeast with 4 repeats showed 5.9 times higher adsorption, and yeast with 8 repeats showed 8.7 times higher adsorption for copper ions.	(Kuroda and Ueda 2006)
	N/A	Mutant protein NikRm from E. coli	Uranyl ions	α‐Agglutinin	Adsorption abilities of strains displaying tandemly fused metal‐binding domains of NikRm were significantly higher than those in control group; adsorption rate was about 57%, and 77% uranyl could be recovered by citrate buffer (pH 4.3) treatment	(Kuroda et al. 2014)
	EBY100	Metal‐binding peptides gE1, gE3, gE6, and gE6 selected from bacteriophage display peptide library	Cadmium ion	α‐Agglutinin	EBY100‐gE1 had the highest adsorption rate improvement of 35%. And when this recombinant strain was immobilized with sodium alginate, its adsorption rate is 55.7% higher than control yeast(immobilized).	(S. Wang et al. 2024)

7.2.2. Evolutionary and Metabolic Reprogramming

To enhance the copper storage capacity of yeast, He et al. (2022) employed atmospheric and room‐temperature plasma (ARTP) mutagenesis to construct a S. cerevisiae strain with high copper accumulation. The resulting mutant strain, designated H247, exhibited the highest intracellular copper content—55.8% higher than that of the parental strain Cu‐5 (He et al. 2022). Subsequently, transcriptomic analyses revealed that the differentially expressed genes between H247 and Cu‐5 were significantly enriched in the positive regulation of lipid biosynthetic processes. These findings suggest that lipid droplets serve a dual role: they not only function as a storage pool for copper but also potentially enhance copper trafficking to subcellular compartments, including mitochondria, vacuoles, and the Golgi apparatus. This work uncovers a novel pathway for enhancing intracellular copper accumulation.

7.2.3. Natural Genetic Variation as a Basis for Diverse Sequestration Traits

While targeted engineering focuses on specific pathways, the innate genetic background of yeast dictates complex structural and biochemical variations—such as β‐glucan/mannoprotein ratios and endogenous MT expression—that profoundly influence total biosorption capacity. This genetic diversity is particularly evident in the remediation of organic toxins, which often relies on a combination of cell‐wall interaction and intracellular processing. Cecchini et al. (2006) identified superior innate OTA biosorption in S. cerevisiae, Saccharomyces bayanus var. uvarum, and S. bayanus during wine fermentation, noting that interstrain variability is significantly more pronounced in red wine matrices than in white wine.

To dissect the specific genetic mechanisms driving this variability, Caridi et al. (2012) assessed the OTA biosorption capacity of parental wine strains (TP5 and TT173) and their 46 meiotic progeny. Substantial phenotypic divergence was observed, which the authors attributed to structural differences in mannosyl phosphate content, alongside variations in cell volume and flocculation capacity. Following a 30‐day microfermentation in OTA‐spiked grape must, transgressive segregation was observed: while parental strains left residual OTA concentrations of 1.42–1.60 ng/mL, select progeny achieved residual levels as low as 0.60 ng/mL. Flow cytometric and phenotypic distribution analyses confirm that OTA biosorption is a polygenic, quantitative trait governed by multiple loci rather than simple Mendelian segregation. This complex genetic architecture provides a theoretical foundation for targeted selective breeding programs aimed at optimizing commercial yeast strains for both heavy metal and mycotoxin remediation (Figure 2B).

7.3. Pretreatment Strategies for Enhanced Biosorption and Loading

Prior to their utilization as microencapsulation matrices, yeast cells are subjected to pretreatment protocols designed to maximize intracellular payload capacity. Autolysis and plasmolysis are the two most prevalent techniques. Autolysis leverages endogenous hydrolases to degrade cytosolic constituents, typically under optimal conditions of pH 5.5 and 50°C. This enzymatic degradation increases plasma membrane permeability and expands the available void space for bioactive sequestration. In contrast, plasmolysis utilizes hyperosmotic agents to induce cellular dehydration, causing the plasma membrane to retract from the CW. While this cytosolic contraction increases loading volume, it may impair the encapsulation efficiency (EE) for ligands that require specific non‐covalent interactions with intracellular macromolecules, particularly if the target organelles are structurally compromised during osmotic shrinkage. Additionally, physical interventions such as ultrasonication, electroporation, and high‐pressure homogenization are employed to selectively disrupt the cellular envelope and enhance encapsulation yields (Tan et al. 2021).

For surface‐mediated biosorption, thermal and chemical pretreatments are employed to unmask reactive binding sites on the CW (Figure 2C). Göksungur et al. (2005) demonstrated that ethanol, alkaline (NaOH), and thermal treatments universally enhance heavy metal adsorption. Specifically, ethanol permeabilization provides the greatest improvement for Cd²⁺ and Pb²⁺ sequestration, while alkaline treatment maximizes Cu²⁺ biosorption. Mechanistically, these interventions strip away masking outer layers and denature CW biopolymers, thereby exposing previously inaccessible functional groups.

These principles have been effectively applied to the upcycling of industrial waste yeast for polyphenol recovery. Rubio et al. (2018) investigated the comparative efficacy of thermal (autoclaving at 121°C for 20 min), acid (0.1 M H₂SO₄), and alkaline (0.1 M NaOH) pretreatments on the biosorption of grape pomace polyphenols. Quantitative analysis revealed that alkaline‐pretreated yeast achieved a superior biosorption capacity of 198.28 mg/g, representing a 93.2% increase over native biomass. Mechanistically, alkaline treatment not only increases the density of available binding sites but also enhances the thermodynamic stability of the resulting yeast–polyphenol complexes. This was evidenced by a 196% increase in residual phenolic content and a 147% boost in bioaccessibility during in vitro simulated GI digestion, suggesting that surface modification is critical for both the initial uptake and the subsequent functional protection of the target ligands.

A recent study has shown that saline treatment represents a promising, low‐cost, and less toxic enhancement method for yeast biosorption in the food industry. Pedro et al. (2026) found that salt‐treated yeast exhibited a biosorption capacity 517% higher than that of unmodified yeast, while the equilibrium time for the biosorption reaction decreased from 360 to 60 min. This enhancement may be attributed to osmotic stress, which removes water and soluble compounds from the cell, thereby increasing intracellular space and facilitating the osmoporation process that promotes transport of fluid mass into the cell. Additionally, saline treatment increases surface heterogeneity and may expose new binding sites, thereby enhancing the interaction between yeast and phenolic compounds (Pedro et al. 2026).

7.4. Physical Intensification of Biosequestration and Encapsulation

Mass‐transfer resistance frequently constrains the efficiency of yeast‐based sequestration. To overcome these limitations, physical intensification strategies—primarily ultrasonication and vacuum‐assisted impregnation—are employed to accelerate ligand transport through distinct biophysical mechanisms. While ultrasonication utilizes acoustic cavitation to disrupt boundary layers, vacuum‐assisted biosorption leverages transmembrane pressure differentials to drive convective flow into the cellular architecture.

7.4.1. Ultrasonication‐Mediated Kinetic Enhancement

The mechanistic basis of ultrasonic intensification lies in acoustic cavitation, where the implosion of microbubbles generates localized hydrothermal extremes, microturbulence, and microstreaming (Vo et al. 2025). These effects attenuate the boundary liquid film resistance and induce microperforations in the cell envelope, thereby unmasking internal binding sites and thermodynamically favoring ligand‐matrix interactions (Breitbach and Bathen 2001).

Tao et al. (2019) demonstrated that low‐frequency ultrasound (20 kHz; 394.2 W/L) enhanced the biosorption capacity for blueberry phenolics onto spent brewer's yeast by 62.7% (reaching 67.2 mg/g) compared to mechanical agitation. Kinetic modeling revealed that ultrasonication increased the external mass transfer coefficient (k _s) by 128.5% and the surface diffusion coefficient (D _s) by 74.3%. This acceleration was partially attributed to structural comminution, where acoustic shear fragmented the yeast cells, thereby shortening the intracellular diffusion paths.

Beyond adsorption, ultrasonication facilitates the intracellular loading of hydrophobic nutraceuticals. In the encapsulation of fisetin, an acoustic energy density (AED) of 333.3 W/L yielded an EE of 67.7% while preserving 90% of the antioxidant activity (de Andrade et al. 2022). Scanning electron microscopy (SEM) revealed that cavitation‐induced pressure waves ablated the natural topological roughness and bud scars of the yeast surface, altering the supramolecular conformation of the glucan‐mannoprotein network. This suggests a multi‐mechanistic enhancement: increased membrane permeability, partial efflux of cytosolic components, and the disruption of the native lipid–protein phase architecture, which collectively facilitate transmembrane diffusion and stabilize non‐covalent interactions with intracellular sites.

7.4.2. Vacuum‐Assisted Impregnation and Structural Integrity

In contrast to cavitation‐driven methods, vacuum‐assisted biosorption is a nonthermal technique utilizing biphasic pressure transitions. Initial depressurization evacuates interstitial and intracellular gases, followed by a rapid restoration of atmospheric pressure that mechanically drives the ligand‐rich solvent into the porous cellular cavity. This convective mechanism is particularly efficacious for loading poorly water‐soluble bioactives into intact matrices.

The superiority of vacuum‐driven loading was evidenced by Young et al. (2017), who achieved a 288‐fold acceleration in process kinetics and a threefold increase in curcumin mass loading compared to passive diffusion (1.0 kPa for 5 s). Notably, the EE in intact cells was significantly higher than in isolated yeast cell wall particles (YCWPs), with fisetin and curcumin loading being 4.7‐fold and 1.7‐fold greater, respectively. This disparity suggests that intact intracellular structures—including mitochondria, cytoskeletal networks, and nucleic acids—provide supplementary binding interfaces that are lost during the harsh chemical extraction of YCWPs.

The importance of cellular integrity is further underscored in storage stability. Costa et al. (2025) reported that while vacuum‐assisted impregnation (−101.3 kPa) achieved high loading (55.09%–96.95%) of Vitamin D3, intact yeast cells provided superior structural protection during long‐term storage. Specifically, intact cells retained 31.2% of the payload over 60 days, compared to only 20.6% in plasmolyzed biomass. These findings indicate that while physical intensification maximizes initial loading, the maintenance of the native cellular ultrastructure is paramount for the targeted release and enhanced bioavailability of the sequestered cargo within the intestinal compartment.

8. Conclusion and Future Perspectives

This review systematically elucidated the structural basis, kinetic mechanisms, and enhancement strategies of yeast‐mediated biosorption, emphasizing its dual role in decontaminating heavy metals and mycotoxins while facilitating the targeted encapsulation of bioactive nutrients. By integrating advanced genetic engineering, targeted pretreatments, and physical intensification methods such as ultrasonication and vacuum perfusion, yeast‐based biosorbents have demonstrated remarkable efficiency in diverse food‐related applications. However, the research on yeast biosorption in food matrices is considerably less in‐depth than biosorption in aqueous models, while a significant gap remains between laboratory‐scale aqueous models and the physicochemical complexities of real food matrices. The key difference lies in the presence of biomacromolecules—including polysaccharides, proteins, and lipids—which can compete with yeast for sorbates, form physical barriers, and affect yeast metabolic activity, thereby profoundly influencing ligand bioavailability and biosorbent integrity. Additionally, certain food matrix molecules may compete for binding sites, further complicating the biosorption process. To facilitate methodological translation, direct comparative evaluations between real food matrices and laboratory‐scale aqueous models should be conducted. Future studies performed in authentic food matrices, employing advanced analytical approaches with rigorous variable control and quantitative target monitoring, will be instrumental in advancing the field. Furthermore, although most yeast strains used in food‐related biosorption studies are GRAS, the safety of genetically engineered or chemically treated biosorbents requires rigorous certification. In the United States, genetically engineered yeast may be reviewed via the FDA's GRAS pathway; in the EU, such products require premarket authorization as novel foods (Regulation (EU) 2015/2283); and in China, they must undergo safety evaluation through the “Three New Foods” pathway. For chemically treated biomass, it is equally important to ensure that reagents and their residual levels comply with local food safety standards. Future research should focus on bridging the gap between simplified proof‐of‐concept models to complex food systems through interdisciplinary validation and comprehensive techno‐economic assessments.

Author Contributions

Zeting Chen: writing – original draft, writing – review and editing, visualization. Jiemei Shen: writing – review and editing. Xingyi Yang: writing – review and editing. Zhibin Liu: writing – review and editing. Jinzhi Han: writing – review and editing. Li Ni: writing – review and editing. Wangxin Liu: writing – review and editing, writing – original draft, visualization, supervision, resources.

Funding

This work was supported by funding from Fujian Provincial Higher Education Testing Center Valuable Instruments and Equipment Open Testing Fund (No. 2026T046) and Fuzhou University Talent Start‐up Fund (No. 511597).

Conflicts of Interest

The authors declare no conflicts of interest.