Microencapsulation of Monascus Red Pigment Using Saccharomyces cerevisiae Ghosts: Process Optimization and Characterization.

Introduction
Microbial pigments are sustainable colorful secondary metabolites produced by microorganisms to adapt to challenging ecological conditions [1]. Their ease of harvesting, cost-effectiveness, and unique physicochemical and biological properties make them highly desirable as nontoxic biocolorants in the food, pharmaceutical, personal care, and textile industries [2, 3]. Beyond their commercial applications, microbial pigments are gaining increasing attention in biomedicine for their potential antioxidant, photoprotective, antimicrobial, and anticancer activities [4, 5]. These pigments belong to diverse chemical classes, including carotenoids, terpenoids, melanin, phenazines, and azaphilones [6, 7].
Among microbial pigments, Monascus pigments (MPs) are mixtures of azaphilones produced by Monascus species. These fungi synthesize three groups of differently colored pigments, each having two structurally related compounds: yellow (monascin and ankaflavin), orange (rubropunctatin and monascorubrin) and red (rubropunctamine and monascorubramine) pigments, which have similar structures but differ in the length of side chains. While yellow and orange MPs are directly biosynthesized by the fungus, red MPs are formed through a chemical reaction between orange MPs and compounds containing primary amino groups under neutral or mildly acidic conditions [8]. Red pigments exhibit poor photostability and low aqueous solubility at low pH [9–11].
Monascus pigments (MPs) have been traditionally used as safe and vibrant biocolorants in Asian countries for generations [12, 13]. Beyond their coloring properties, MPs exhibit various therapeutic benefits, including anti-obesity effects [14], improvement of lipid metabolic disorders and modulation of gut microbiota dysbiosis [15, 16], and antimicrobial activity [17] which can be further enhanced by nanoconjugation with metal oxides [18]. However, a major safety concern associated with MPs is the potential co-production of citrinin, a mycotoxin known to cause liver and kidney toxicity and potentially carcinogenic effects. This led to the development of various strategies to inhibit citrinin synthesis and improve the safety of MPs [19].
Nanoencapsulation was shown to improve MPs’ stability and bioactivity. For instance, adding MPs to a double emulsion system made of guar gum and sodium caseinate improved the pigments’ resistance to light and heat [20]. It has also been demonstrated that ova albumin-based core–shell nanoparticles (NPs) enhance the stability of MPs in addition to their antioxidant activity and pancreatic lipase inhibitory effect [21].
Focusing on Monascus red pigments (MRP), literature reports indicated potential use in antimicrobial therapy as active agents against several microorganisms [17, 22, 23] as well as photosensitizers in antimicrobial photodynamic therapy [24]. Other possible clinical uses include enhancing cardiac performance in animals with heart failure using mesenchymal stem cells by scavenging reactive oxygen species and controlling the phenotype of macrophages [25]. Moreover, MRP showed a kidney protective effect against renal toxicity caused by hydroxyapatite nanoparticles [26]. An eco-friendly MRP nano-suspension used as a natural colorant of cellulose fabric offered promise as a safer and more effective alternative to some synthetic dyes in the design of personal care items [27].
To broaden the scope of applications of MRP in foods and medicine, several attempts have been made to improve its physicochemical properties and biological activities. For instance, enhancement of the solubility and stability of MRP could be achieved by modulating its biosynthetic pathways or via direct protein microgelation [28, 29]. Its release properties were improved by incorporation into Ca2+-mediated chitosan/sodium alginate hydrogel beads [30] while its anticancer activity was boosted by lecithin/chitosan nanoparticles (NPs) [31] and liposomes [32].
Although biomaterial-based carrier systems can enhance the physicochemical and biological performance of MRP, nature-derived biocarriers such as S. cerevisiae cells offer unique and compelling advantages for biomedical applications. These include abundant and consistent availability, low production costs, and ease of large-scale manufacturing. From a formulation perspective, S. cerevisiae cells are well-characterized in terms of safety, physical dimensions (averaging 7.5 × 5.5 µm), surface charge, cell wall composition, and inherent biological activities, making them a promising and practical vehicle for MRP delivery [33, 34]. Consequently, S. cerevisiae has been explored as a bioactive carrier for a diverse range of therapeutic agents, including berberine [35], oils [36], nucleic acids [37], as well as polymer- and lipid-based nanoparticles [38, 39]. Furthermore, yeast-derived structures have also been leveraged in advanced drug delivery systems. For instance, daunorubicin has been encapsulated within S. cerevisiae-derived lysosomes and reassembled vacuoles [40, 41], doxorubicin in extracellular vesicles [42], curcumin in glucan particles [43], and mCherry fluorescent protein in L-BC capsids viral-like particles [44].
However, studies focusing on the encapsulation of natural pigments by S. cerevisiae [45] and the application of S. cerevisiae ghosts (ScGs) for drug delivery remain limited [46, 47]. To address this research gap, we selected ScGs as biocarriers for MRP. Probiotic ghosts are generated by removing cytoplasmic and genetic material while preserving the structural integrity of the cell envelope and surface functional groups [48–50]. Compared to conventional biomaterial-based delivery systems, ScGs offer several distinct advantages: their hollow internal structure allows for efficient loading and controlled release of therapeutic agents, while their intact cell wall ensures structural stability in biological environments, enhances safety, and supports intrinsic targeting capabilities.
The present study aimed to develop an innovative S. cerevisiae-microencapsulated MRP as a bio-microstructure that integrates the bioactivities of the pigment with the carrier functionalities of S. cerevisiae ghosts (ScGs) for diverse biomedical applications. A Box–Behnken design (BBD) was employed to optimize the microencapsulation process. Four experimental variables were assessed to maximize ScGs loading with the natural red pigment. The resulting MRP-ScGs constructs were thoroughly characterized for their morphological features, structural integrity, cell wall properties, MRP loading and release profiles, cytotoxicity, cellular uptake, and potential molecular targets. These evaluations provide valuable insights into their potential as multifunctional bioactive carriers for biomedical applications.

Materials and Methods

Preparation of Purified M. purpureus Red Pigment (MRP)
A scratch of the fungal spores of the M. purpureus strain from an agar slant was used to inoculate potato dextrose agar (PDA, Oxoid) plates under aseptic conditions. The plates were incubated in a static incubator (JSGI-100 T, Korea) at 28–30 °C for 7 days. Discs of the fungal growth on the plates were generated by using a sterilized cork borer. These discs were further used as a seed culture to inoculate the core production medium of Monascus pigments.
For pre-inoculum cultivation, three discs of the mother culture of M. purpureus were inoculated into 100 mL of pre-inoculum culture medium containing (60 g/L glucose, 20 g/L peptone, 10 g/L KH2PO4, 10 g/L NaNO3, and 5 g/L MgSO4) in 250-mL Erlenmeyer flasks. This cultivation was carried out at 30 °C and 180 rpm for 30 h. Batch fermentation and Monascus pigment extraction were carried out as reported with slight modification [51]. An aliquot of pre-inoculum culture (5 mL) was inoculated into 50 mL of production medium (50 g/L rice powder, 20 g/L NH4NO3, 3 g/L NaNO3, 1.5 g/L KH2PO4, 1 g/L MgSO4·7H2O, 0.2 g/L MnSO4) in 250-mL Erlenmeyer flasks and incubated at 30 °C and 180 rpm for 7 days [51]. The culture broth was centrifuged at 672 × g for 10 min at ambient temperature (⁓25 °C) to separate the supernatant from the biomass. For determining dry cell weight (DCW), the mycelium was washed twice with distilled water and dried at 60 °C in an oven for 24 h. The dried biomass was cooled in a desiccator and weighed. For the extraction of intracellular Monascus pigments, about 0.5 g of dry powder ground from biomass was transferred into a 10 mL centrifuge tube and extracted in triplicate with 3 mL of 75% ethanol for 30 min in an ultrasonic bath and centrifuged at 672 × g for 10 min. The total supernatant was pooled and passed through RC 0.22-µm bacterial filters.
The crudely extracted Monascus pigments (yellow, orange, and red) were purified through a two-step procedure [52]. Firstly, these were mixed with n-hexane in a ratio of 1:1 in a suitable separating funnel. The organic n-hexane layer containing the yellow Monascus pigments was separated and removed. The aqueous layer containing orange-red Monascus pigments was allowed to pass through a solid phase extraction C18 column cartridge (500 mg/6 mL, Finisterre by Technorama, Spain), pre-conditioned with methanol followed by distilled water to prevent loss of the aqueous phase). The orange Monascus pigment was eluted first. The adsorbed red Monascus pigment was then desorbed from the column using methanol.

Characterization of Monascus pigments

UV/Vis Analysis Spectrophotometry
The absorbance spectra of the M. purpureus intracellular extracts were recorded by UV/Vis spectrophotometry (UV-160A double-beam spectrophotometer, Agilent tech., USA) from 350 to 700 nm at 1 nm intervals. The absorbance of intracellular yellow, orange, and red pigments was determined at the specific wavelengths 410 nm, 470 nm, and 485 nm, respectively. A calibration graph was constructed for MPR at λmax = 485 nm.

Fourier Transform Infrared (FTIR)
The red pigment was mixed with IR grade KBr (1:100) (Sigma-Aldrich, St Louis, USA) and scanned over a range of 4000–450 cm−1 at ambient temperature [53].

Liquid Chromatography-Tandem Mass Spectrometry (LC–MS/MS)
Analysis of the red pigment content was performed by LC–MS/MS using the 3200Q-TRAP LC–MS/MS System (AB Scitex, Framingham, MA, USA) with an EMS mode and a Prominence UFLC (Shimadzu, Kyoto, Japan). The pigment was separated on a Mightysil RP18 column (150 mm × 2 mm) with a linear gradient from a mobile phase consisting of acetonitrile–water containing 0.1% formic acid (60:40, v/v) to acetonitrile–water containing 0.1% formic acid (100:0, v/v) at a flow rate of 0.2 mL/min, oven temperature 40 °C, and run time 25 min. MS/MS was used to detect pigment compounds using electro-spray ionization in the positive ion mode (IS, 2000; CUR, 40; CAD, set to “high”; TEM, 250; GS1, 50; GS2, 80; scan range, m/z 100 to 1000; scan speed, 4,000 Da/s). Data analysis was performed using the Analyst 1.5.1 software version. The separated pigment was also examined at 250–600 nm using a photo diode array detector (SPD-M20A) [54].

Preparation of S. cerevisiae ghosts
S. cerevisiae was purchased from WFCC-MIRCEN, the World Data Centre for Microorganisms (http://www.wdcm.org/), Faculty of Agriculture, Ain Shams University, Cairo, Egypt. The ScGs were prepared using the modified sponge-like reduced protocol (SLRP) [55] with modification involving removal of two chemicals, calcium carbonate (CaCO3) and hydrogen peroxide (H2O2). Briefly, different volumes of 10% sodium dodecyl sulfate (SDS) and 10% NaOH solutions (100, 200, 300, 400, 500, 600, 700, 800, 900, and 1000 µL) were added to S. cerevisiae pellets pre-washed with 0.85% NaCl solution and suspended in 8 mL of sterile water. The volume was adjusted to 10 mL to detect the least concentration of both chemicals necessary for the fastest production of ScGs without complete lysis of Sc cells as verified by light microscopy. The mixtures were incubated at 37 °C for 24 h with agitation at 120 rpm. The S. cerevisiae pellets were harvested by centrifugation for 5 min at 672 × g and washed twice with a 0.85% NaCl solution. The supernatants were analyzed spectrophotometrically at 260 nm and 280 nm for DNA and proteins, respectively, using Uv–Vis spectra (Thermo Scientific-VISION proSoftware V4.30). To check the viability of prepared ghosts, 100 µL of S. cerevisiae ghosts suspension was inoculated onto a Sabouraud Dextrose agar plate and incubated at 30 °C for 72 h.

Characterization of S. cerevisiae Ghosts

Integrity of Cell Walls by Light Microscopy
Light microscopy with multiple staining was used to examine the integrity of ScGs cell walls. Both S. cerevisiae cells and ScGs were examined under a light microscope after staining with trypan blue, toluidine blue, nigrosin, and crystal violet stains, separately. Briefly, S. cerevisiae and ScG suspensions were smeared on the surface of clean and dry microscopic glass slides and fixed by exposing the slides briefly to a Bunsen flame [56]. The fixed slides were separately stained with 200 µL of crystal violet (1% w/v for one min) or 200 µL of toluidine blue (0.1% w/v for 10 min) and then washed with water. The slides were air-dried and examined under the oil immersion lens of the light microscope. For staining with trypan blue, 50 µL of trypan blue (0.2% w/v stock solution) was mixed with 50 µL of live S. cerevisiae and ScGs suspensions separately. The mixture was set aside for 5 min and then checked using a hemocytometer under the 40 × lens of the light microscope as reported [57]. For nigrosin staining (negative staining), 50 µL of nigrosin dye (10% w/v) was added near the end of a clean and dry microscopic glass slide. A 50 µL aliquot of S. cerevisiae or ghost suspension in 0.85% NaCl was mixed with the nigrosin stain, and the mixture was spread using another slide, producing a broad, even, and thin smear [58].The smear was left to dry without heating, then examined under the 100 × lens of the light microscope.

Morphological Characteristics

Scanning Electron Microscopy (SEM)
Live S. cerevisiae cells and ScGs were fixed using 2.5% glutaraldehyde for 2 h at 4 °C to preserve native cell morphology. After fixation, the samples were washed with PBS and then dehydrated through a graded ethanol series. Finally, the samples were smeared and allowed to air-dry. The smear surface was coated with gold (JEC-1100 E Sputter Coater). The coated samples were scanned by SEM (SEM, JEOL JSM-5300).

Transmission Electron Microscopy (TEM)
Live S. cerevisiae cells and ScGs were quickly immersed in glutaraldehyde and fixed for at least 1 h at room temperature before being post-fixed in osmium tetroxide. Fixed cells were embedded in agar and processed. The specimens were dehydrated in graded concentrations of ethanol and propylene oxide and embedded in Spurr’s plastic. Semi-thin sections were cut from blocks with a glass knife, and the blocks were chosen for thinning. Diamond-cut thin sections were placed on copper grids, impregnated with uranyl acetate and lead citrate, and scoped [59]. Samples were examined by TEM (TEM 100 CX, Jeol-Japan).

Lyophilization of S. cerevisiae Ghosts
ScGs were washed twice with sterilized physiological saline. The ghosts were suspended in 1 mL of water and 10% reconstituted skim milk. Samples, 1 mL each, were frozen at − 20 °C and desiccated under vacuum (50 m Torr) for 48 h at − 20 °C (Freeze-dry system and stoppering tray dryer; Laconco, Kansas, USA). The lyophilized ghosts were stored at − 20 °C [60].

Microencapsulation of Monascus Red Pigment (MRP) by S. cerevisiae Ghosts

Optimization Using Box–Behnken Design (BBD)
The ScGs-microencapsulated MRP (MRP-ScGs) was prepared by incubating a known amount of lyophilized ScGs (10 mg) with a known volume of a pigment solution in DMSO in a shaking incubator under controlled conditions. The process parameters were optimized to maximize pigment loading using the BBD for response surface methodology (RSM) [61] created by Minitab software 17.3. Four independent variables that significantly affected the pigment loading and one response (dependent variable) were identified in a set of preliminary single-point experiments. Independent variables included: incubation temperature (X1), initial MRP concentration (X2), shaking rate (X3), and incubation time (X4). Each independent variable was tested in three coded levels, denoted as −1 (low level), 0 (center level), and + 1 (high level), through a design matrix of 27 experimental trials (Table S1). Loading efficiency % (LE %) was selected as the dependent variable (response).

Characterization of MRP-ScGs

Morphological Properties
The MRP-loaded ScGs were examined in comparison with blank ScGs by light microscopy without staining, confocal microscopy, and TEM. For light microscopy, a suspension of MRP-ScGs and blank ScGs was investigated without staining using a hemocytometer under the 40 × lens of the light microscope. The quality of cells was judged by examining the integrity of their surface structures. MRP-ScGs and ScGs were also visualized by confocal microscopy using a Leica TCS SPE equipped with a Leica LAS X interface [62] after fixation of samples for 10 min in 10% neutral buffered formalin [63]. TEM imaging was performed as described above.

Loading Efficiency
The amount of MRP loaded by ScGs was determined as reported by Xie et al. [64] with some modifications. Briefly, after the loading process, the mixture was centrifuged at 672 × g for 5 min. The pellet (10 mg) was washed three times with PBS to remove residual unbound pigment and then suspended in 70% ethanol for the extraction of loaded MRP by vigorous shaking for 10 min. The concentration of the liberated pigment under different loading conditions was determined spectrophotometrically at λmax 485 nm. The loading efficiency percent (LE %) was calculated using the following equation [65]:

Pigment Release
The release of MRP from MRP-ScGs was investigated at 37 °C using a dialysis method and PBS pH 7.4 as the release medium. Tween 80 was added to PBS to enhance MRP release and ensure sink conditions. The concentration of Tween 80 was selected based on a preliminary experiment for the solubility of MRP in PBS pH 7.4 containing Tween 80 in different concentrations (0.1, 0.2, 0.4, 0.8, and 1.6%). A known volume (1 mL) of an MRP solution in DMSO (300 mg/mL) was placed in a dialysis bag (VISKING® dialysis tubing MWCO 12,000–14,000, Sigma-Aldrich) and shaken in 20 mL of the release medium at 100 rpm for 48 h protected from light. At predetermined time intervals (1, 2, 4, 6, 24, and 48 h), 2-mL samples of the release medium were withdrawn, filtered through a 0.45-µm membrane filter, and assayed spectrophotometrically for MRP at λmax 485 nm. The withdrawn sample was immediately replaced with the same volume of fresh release medium adjusted to 37 °C. For the release of MRP from MRP-ScGs, 1 mL of a 0.5% suspension of MRP-ScGs in saline was placed in the dialysis bag, which was suspended in 20 mL of the selected release medium (PBS pH 7.4/1% Tween 80) and the procedure was completed as described. Experiments were run in triplicate.

Cytotoxicity
The cytotoxicity of MRP-ScGs in comparison with blank S. cerevisiae cells and MRP was assessed using the lung cancer A549 cell line obtained from the Center of Excellence for Regenerative Research in Medicine and its Applications (CERRMA), Faculty of Medicine, Alexandria University. Cytotoxicity studies were performed using the 3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyl-2H-tetrazolium bromide (MTT) assay. Prior to the cytotoxicity study, the viability of cancer cells was assessed using trypan blue dye exclusion staining and a dual chamber hemocytometer for manual cell counting [66]. Briefly, 20 µL of cell suspension was combined with 20 mL of a 0.4% trypan blue solution in 0.8% NaCl in a 1.5-mL Eppendorf tube. A 10 µL volume of the mixture was placed into each chamber of the hemocytometer to assess cell viability under the microscope. Cells were counted using the × 10 objective [67]. Equation 2 was used to compute the number of cells [68]. For most applications, 7 × 104 cells/mL were used for seeding [69].
The percentage of unstained viable cells was calculated using Eq. 3 [70].
The A549 cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin, then maintained in a 37 °C incubator with 5% CO₂ under humidified conditions. MRP was dissolved in DMSO, while ScGs were suspended in 1 mL PBS and all diluted in Dulbecco’s modified Eagle medium (DMEM) to reach the specified concentrations. Based on hemocytometer cell counting data, cancer cells were seeded at 7000 cells per well and maintained overnight at 37 °C. The medium was aspirated, and 100 µL of the treatment media containing MRP (0.25–50 µg/mL) [71] or ScGs (0.05 × 1012, 0.1 × 1012, 0.2 × 1012, 0.5 × 1012, 1 × 1012 CFU/ml) or a combined system with the same range of MRP and ScGs was added to the cells. MRP-ScGs at a fixed MRP concentration (0.25–50 µg/mL) were added to all wells, and cytotoxicity was measured after 24 h. Cells were incubated with 10 µL MTT (5 mg/mL) for 4 h, and the formed formazan crystals were dissolved in 100 µL of DMSO. Finally, the absorbance was measured at 590 nm using a Tecan microplate reader, USA. All experiments were carried out in triplicate. The median inhibitory concentration (IC50) values were calculated according to Eq. 4 [72]:

MRP Uptake by Cancer Cells
The effect of microencapsulation by ScGs on the uptake of MRP by cancer cells was assessed by confocal microscopy using the A549 cell line. The A549 cells were cultured at a density of 3 × 105 cells per well in 6-well plates, and after attachment, they were treated with 20 µg/mL of MRP either in the free or microencapsulated form for 24 h. Subsequently, cells were fixed with methanol and treated with Hoechst to stain nuclei for examination under the confocal microscope [73].

Molecular Target Prediction for MRP and Molecular Docking
Based on the structure of MRP, molecular targets were predicted in an attempt to further understand the mechanism by which MRP lowered the viability of cancer cells. The structure of MRP was retrieved from ZINC database (ZINC58563822) as the following Simplified Molecular Input Line Entry System (SMILES) layout:C/C = C/C1 = CC2 = C(CO1)C(= O)[C@]1(C)OC(= O)[C@H](C(= O)CCCCC)[C@H]1C2

To predict potential cellular targets in human cells, MRP was screened for molecular targets using the Swiss Target Prediction tool (http://swisstargetprediction.ch/). Gene set enrichment analysis was performed for the top 50 identified targets using enrichR tool [74–76]. Finally, a potential target was selected for docking using SWISSDOCK (http://www.swissdock.ch/docking). The 3D structures of the selected targets were downloaded from the Protein Data Bank (PDB) website (https://www.rcsb.org/) [77].

Statistical Analysis
Data were analyzed using Graph Pad Prism ® version 6 software (GraphPad Software Inc., CA, USA). Data were expressed as mean ± SD. A value of p < 0.05 indicated statistical significance. Minitab software 17.3 was used to generate Box–Behnken design and analyze the data obtained. Statistica 10 was used to construct the three-dimensional surface plots.

Results

Monascus Red Pigment Preparation and Characterization
By the end of the incubation time using the fermentation medium described above, the maximum dry M. purpureus cell weight was approximately 10.5 g/L, and the maximum yield of purified MRP was 14 mg/g of dry weight mycelia. The purified MRP was characterized by different spectral methods (Fig. 1a–c). The obtained MRP could be detected at 485 nm upon scanning of the UV/Vis absorbance in the range 350 nm to 700 nm (Fig. 1a). The FTIR spectrum of MRP (Fig. 1b) showed peaks at 3300–3500 cm−1 indicative of N–H stretching in secondary amide, 3000–3100 cm−1 due to C-H2 stretching in − CH, 2850–2990 cm−1 assigned to C-H3 stretching in the CH group, 1630–1730 cm−1 denoting C = O stretching of the 5-membered cyclic ketone-lactone system and 1440–1640 cm−1 related to C = C stretching of the conjugated double bond system. Analysis of the MRP by high-performance liquid chromatography coupled with a mass spectrometer LC–MS/MS (Fig. 1c) showed a large peak at m/z 353 that was consistent with the (M + H)+ ion of the compound and a smaller peak at 381, indicating a molecular weight of 353 kDa for rubropunctamine (C21H23NO4) and 381 kDa for monascorubramine (C23H27NO4).

S. cerevisiae Ghost Preparation and Characterization
In preparing ScGs via the modified sponge-like protocol, as the concentration of the chemicals used in the treatment increased (from 0.01 to 0.1 mg/mL), the amount of DNA and protein released into the supernatant increased, indicating more cells were being evacuated. However, this release reached a maximum point (at chemicals concentration of 0.1 mg/mL) (Fig. 2a and b). Further increases in chemical concentration (0.12–0.14 mg/mL) did not result in more DNA release while resulting in complete lysis of S. cerevisiae cells, as confirmed by the absence of cell pellets following centrifugation. Moreover, non-viability of the prepared ScGs was confirmed by the absence of colony formation on Sabouraud Dextrose Agar plates after 3 days of incubation. This indicates that the treatment successfully generated non-viable “ghost” cells. The lack of growth served as evidence that the treatment effectively compromised cellular integrity, preventing any possibility of recovery or replication.

Light Microscopy Imaging
The prepared ScGs were first differentiated from native S. cerevisiae cells by light microscopy using different stains. Figure 3a–h shows light microscopic images of ScGs in comparison with S. cerevisiae cells using trypan blue, toluidine blue, nigrosin, and crystal violet stains. Successful ghost formation and the morphological characteristics of the prepared ScGs were evaluated via the detection of evacuated dead S. cerevisiae cells as well as the preservation of the integrity and negativity of their cell envelopes.
Trypan blue, a viability indicating dye, was excluded from native S. cerevisiae cells and acquired by ScGs, which appeared blue with a clear cell outline under the 40 × lens of the light microscope (Fig. 3a and b, respectively). Trypan blue staining indicated successful formation of nonviable ghosts with an empty envelope. Toluidine blue, which binds to acidic polysaccharides and nucleic acids, resulted in a deep blue coloration of native S. cerevisiae cells and faint blue staining of ScGs (Fig. 3c and d, respectively) implying a lack of genetic materials in the ghosts. The morphology and intactness of the prepared ghosts relative to native cells was assessed using nigrosin, an acidic dye, which stains the background, leaving the bacterial cells unstained. As shown in Fig. 3e and f, the glass slide was stained with a black tint, but native S. cerevisiae and ScGs excluded the negatively charged nigrosin staining and appeared as clear specks on the dark backdrop. Crystal violet stained native S. cerevisiae cells with a deep purple color attributed to the binding of the cationic dye to both the cell walls and nucleic acids, while ScGs lacking nucleic acids displayed a much lighter purple color of the cell envelope (Fig. 3g and h, respectively).

Electron Microscopy Imaging
Scanning electron microscopy (SEM) imaging (Fig. 4a and b) of native S. cerevisiae cells (Fig. 4a) in comparison with that of ScGs (Fig. 4b) indicated that ScGs preserved the yeast 3D structure and an intact and non-denatured surface structure showing tiny pores at the tip of the cells (yellow arrows), through which the intracellular contents were discharged. TEM examination (Fig. 4c and d) revealed that native S. cerevisiae had intact cell walls with an obvious nucleus, cytoplasmic materials, and a vacuole (Fig. 4c) whereas ScGs showed an empty intracellular space, rigid cell walls in addition to an obvious pore (Fig. 4d).

Microencapsulation of Monascus Red Pigment by S. cerevisiae Ghosts
The microencapsulation of MRP by ScGs was optimized using response surface methodology (RSM) with a three-level, four factors BBD. Table 1 displays the experimental values of the response (loading efficiency % of MPR into ScGs). Data demonstrated that the experimental values of the response ranged from 9.4 to 61.7% mg MRP/mg ghost, which in turn reflected the model aptness to describe the variations in the response from run to run, which was further supported by R2 (0.92), adjusted R2 (0.82), F-model (9.66), and P-model (0.000176). Multiple non-linear regression analysis of BBD data revealed that there were two model terms (X1 and X2X2) imposing significant influence (P < 0.001) on the response. The interaction among the independent variables, imposing significant effect on the response, was assigned to the quadratic form (X2X2) only. The relationship between the four independent variables and the dependent variable (response) could be portrayed in a full polynomial equation from the second order (Eq. 5) including both significant and non-significant model terms. After exclusion of the non-significant model terms, Eq. 6 was established to describe the relationship between the independent variables and the dependent variable (response). Furthermore, the impact of each two independent variables on the response was illustrated using the three-dimensional surface plots shown in Fig. 5A–F. The highest loading efficiency 61.4 ± 2.3% was realized in run #27 with X1 (25 °C), X2 (125 mg/mL MRP), X3 (300 rpm), and X4 (90 min).

Characterization of MRP-Loaded S. cerevisiae Ghosts (MRP-ScGs)
The optimized MRP-ScGs in comparison with their blank counterparts were characterized morphologically by digital photography of their respective pellets as well as light microscopy under the 100 × lens of the light microscope without staining, TEM, and confocal differential interference contrast (DIC) microscopy (Fig. 6). Blank ScGs (Fig. 6, upper row) appeared visually as a creamy pellet in the Eppendorf tube (Fig. 6a) and colorless capsules under the oil immersion lens of the light microscope (Fig. 6c). In addition, they appeared as empty vesicles with internal dimensions of 2.88 µm and 5 µm, an inner volume of about 26 µm3, and showed a pore of approximately 1 µm in size in the cell wall by TEM at × 15,000 (Fig. 6e) and non-fluorescent vesicles by DIC microscopy (Fig. 6g). On the other hand, MRP-ScGs (Fig. 6, lower row) appeared as a red pellet by digital photography (Fig. 6b) and red capsules by light microscopy (Fig. 6d). TEM of MRP-ScGs at × 25,000 (Fig. 6f) showed an obvious increase in the density of the loaded ghost due to MRP entrapment with no damage to the ghost cell wall, while DIC confocal imaging (Fig. 6h) revealed a uniform distribution of the red fluorescent MRP throughout the internal space of the ScGs.

Release of MPR
The MRP-ScGs were also characterized for MRP release in PBS pH 7.4 containing 1.6% Tween 80 at 37 °C and 100 rpm using a dialysis method (Fig. 7). Release of free MPR from a DMSO solution was used for comparison and to verify MPR dializability. Fast and complete release of free MPR indicated dializability of the drug under the study conditions. In contrast, the release profile of microencapsulated MPR indicated sustained release of MRP according to a biphasic pattern, attaining approximately 8% release in 48 h.

Cytotoxicity and Cellular Uptake
The cytotoxicity of ScGs, MRP, and MRP-ScGs to A549 lung cancer cell line was assessed using cell viability assay, IC50 values, and DIC confocal imaging of the test cells post treatment. The cell viability curves for A549 cells upon treatment with the test preparations (Fig. 9a) indicated a concentration-dependent effect within the concentration ranges of samples used. The median inhibitory concentrations (IC20) determined using GraphPad were > 40, 10, and 3 µg/mL for ScGs, MRP, and MRP-ScGs, respectively, indicating enhancement of anticancer activity of MRP upon microencapsulation by ScGs.
Confocal DIC imaging of the A549 lung cancer cells upon treatment with 20 µg/mL of free or microencapsulated MRP to visualize cellular uptake and distribution of the pigment revealed that free MPR entered the cells and diffused into the cytoplasm with higher concentration at the cell periphery (Fig. 8b, upper panel). However, encapsulation of MRP by ScGs increased cellular uptake of the pigment, which was more uniformly distributed within the cells (Fig. 8b, lower panel). As regards to MRP-ScGs, they accumulated in the proximity of cancer cells, as marked by the arrows (Fig. 8b, lower panel, DIC).

Molecular Target Prediction
It was of interest to investigate possible mechanisms and targets for MRP in human cells that account for its cytotoxic effects. Molecular targets of MRP were predicted using the SwissTargetPrediction tool. The top 50 identified targets included kinases, proteases, and G-protein coupled receptors (Fig. 9a). Gene set enrichment analysis of these top 50 targets revealed enrichment in several pathways related to cancer, such as epithelial-mesenchymal transition, hypoxia, and DNA repair (Fig. 9b). Genes involved in these pathways included MMP1, MMP3, MAPK8, and MAPK10 (Fig. 9c). Upon docking MRP on MMP1, a protein known to play essential roles in many cancer types, the pose with the lowest energy showed interaction with amino acids of the active site of MMP (Fig. 9d). However, more studies are needed to validate these in silico results.

Discussion
The current study describes for the first time the use of ScGs as a biocarrier for MRP. Characterization data of the prepared MRP by UV–Vis spectroscopy, FTIR, and LC/MS/MS (Fig. 1) indicated that the determined λmax (485 nm), FTIR peaks, and the large peak at m/z 353 of the LC/MS/MS spectrum were generally consistent with those reported for MRP pigments [78].
In the preparation of ScGs according to the modified sponge-like reduced protocol (SLRP) reported for the production of Candida albicans ghosts [55], the concentrations of SDS and NaOH were adjusted for inducing poration of the yeast cell wall to discharge cellular proteins and DNA without extensive lysis of the wall. The ScGs pellets were harvested by centrifugation at 672 × g for a relatively short time (5 min) to avoid deformation of the empty cells [79]. Based on the concentrations of proteins and DNA released as a function of the chemicals concentration, an adjusted 0.1 mg/mL concentration of SDS and NaOH (Fig. 2) allowed for the fastest production of ScGs with preservation of cell wall integrity, which was checked microscopically. SDS is known to disturb the cell membrane of yeast, promote the release of cellular proteins, and prompt the signaling pathway of cell wall integrity [80, 81] while NaOH enhances the permeability of the cell wall of yeast cells and extraction of their protein content [82]. Successful production of ScGs by the modified SLRP used for Candida albicans can be attributed to the similarity of cell wall composition of the Saccharomyces and Candida yeasts [83].
Light microscopy examination of the produced ScGs in comparison with native S. cerevisiae cells as control using different stains allowed differentiation of ScGs from native S. cerevisiae cells and affirmed preservation of the integrity and negativity of their cell walls (Fig. 3a–h). For instance, the anionic trypan blue excluded by intact cell walls of live microorganisms did not stain native S. cerevisiae cells but traversed through the cell walls of dead ScGs imparting a blue stain (Fig. 3a and b, respectively), an observation reported elsewhere [55]. Differentiation of ScGs from native S. cerevisiae cells was also realized by toluidine blue, a basic dye having high affinity for nucleic acids [48, 84]. While native S. cerevisiae cells developed a deep blue toluidine stain, ScGs acquired a faint blue color, confirming evacuation of genetic materials (Fig. 3c and d, respectively).
Integrity and negativity of the cell wall of ScGs relative to native S. cerevisiae cells was affirmed by nigrosin and crystal violet staining. Nigrosin, a negatively charged acidic dye, stains the background, leaving bacterial cells unstained. Exclusion of nigrosin by both native S. cerevisiae and ScGs (Fig. 3e and f) verified integrity and negativity of the ghost cell walls similar to native Sc cells. However, pores produced during the generation of ghosts allowed some of the nigrosin dye to infiltrate the ghost cell wall. Negativity of the cell was further verified by the deep staining of native Sc cells by the cationic crystal violet dye known to interact with the negatively charged cell components, mainly cell walls and nucleic acids [79, 85]. The much lighter purple staining of ScGs implied lack of nucleic acids (Fig. 3g and h).
The morphology of ScGs in comparison with native S. cerevisiae cells was further verified by SEM and TEM (Fig. 4a–d) as relatively large 3D empty vesicles (~ 26 µm3 internal volume) devoid of cytoplasm and genetic material and exhibiting an intact negatively charged cell wall showing pores. Such characteristics indicated successful ghost formation [48, 55] and may have significant implications in their bioactivity and targeting properties. Conservation of the 3D structure and intactness of the cell wall of ScGs verified the preservation of the strong cell wall of yeast cells consisting of a reticulated meshwork of mannoproteins and fibrous beta-glucans and chitin along with a lipid bilayer. Such a highly structured cell wall may play a crucial role in promoting the structural integrity of the ghosts and their adhesive properties, particularly in cell–cell or cell-surface adhesion processes [86]. Moreover, a negative surface charge of the ScGs, similar to native cells, may drive the attractive forces in the encapsulation of bioactive agents [87].
As the microencapsulation of MRP by ScGs is a multifactorial process not documented to date, varying process parameters may complicate the selection of the optimal preparation conditions. Thus, a RSM with a three-level, four-factors BBD was used to optimize the loading variables to achieve maximal pigment loading expressed as loading efficiency %. Response surface methodologies were extensively reported as powerful empirical statistical models to localize the optimal levels of independent variables imposing significant consequences on the response (optimized bioproducts) such as α-amylase, alkaline protease, biopigments, and pectic oligosaccharides among others [88–91]. In addition, RSM helps unravel all forms of interactions among the tested independent variables which could be likely imposing significant influences on the process output (response). Herein, BBD was chosen for its efficiency of fewer experimental runs than full factorial designs and lack of combinations where all factors are simultaneously at their high or low levels, which helps avoid extreme experimental conditions [92]. According to optimization data (Table 1 and Fig. 5), the loading of MRP into pre-prepared ScGs by shaking 10 mg of the ghosts at 25 °C with a 300 mg/mL MRP solution in DMSO in a shaking incubator at 125 rpm for 90 min produced MRP-ScGs maximally loaded with the pigment (6.14 mg MRP/10 mg ScGs) achieving 61.4% loading efficiency. Effective drug loading can be attributed to the relatively large internal space of ScGs (~ 26 µm3) and possible non-covalent binding of MRP. This would allow the reduction of bacterial dose while securing enough drug for different applications [93].
Morphological characterization by light microscopy and TEM of the MRP-ScGs versus blank ScGs affirmed encapsulation of MRP by ScGs with no apparent change in the ghost envelope (Fig. 6). Efficient uptake of MRP by ScGs was confirmed by the red fluorescence detected by DIC confocal imaging throughout the internal space of the ScGs in addition to the relatively high LE% of MRP (61.4%). Encapsulation of hydrophobic bioactive agents by yeast cells relies on their uptake through the cell walls composed predominantly of polysaccharides, mannoproteins, and lipids [87]. The drug internalization process normally takes place via carrier-mediated transport by specific carrier proteins [94]. Conversely, the nonviable ScGs act as passive carriers which internalize drug molecules through the generated cell wall pores [93, 95] and by passive diffusion across the cell wall [96] which may be enhanced by electrostatic attraction with its negatively charged surface.
The release of MRP from ScGs was investigated in PBS (pH 7.4) containing 1.6% Tween 80, a medium designed to simulate physiological conditions and incorporate a surfactant as a release enhancer commonly used to enhance the delivery of hydrophobic drugs [48, 97]. Although MRP was soluble in a 1.6% Tween 80 solution in a preliminary solubility study, the release medium with a similar surfactant concentration was unable to effectively extract the pigment from ScGs. The pigment release followed a biphasic profile, remaining below 8% over 48 h (Fig. 7), suggesting strong binding of MRP to the ScGs. The rapid initial phase, corresponding to the release of MRP from or near the surface of the ghosts, was followed by a much slower phase. This pattern likely reflects the localization of hydrophobic MRP in a crystalline form within the hollow structure of the ghosts, surrounded by a resistant envelope that limits dissolution and release through pores.
The encapsulation and release of drugs or bioactive agents from bacterial ghosts are influenced by several factors, including the physicochemical properties of the drug, the type of ghost, drug cell wall binding via electrostatic interactions or hydrophobic bonding, and the composition and properties of the release medium [64, 87, 98]. Binding of MRP with the double carbohydrate wall/lipid membrane cell wall of ScGs, particularly via hydrophobic bonding, might explain the results obtained. Poor release of other hydrophobic agents such as doxorubicin [99], resveratrol [100] and prodigiosin [48] has been attributed to the interaction with lipophilic components of the bacterial cell or ghost envelope.
It is important to note that poor release of drugs encapsulated in bacterial ghosts does not necessarily hinder their bioactivity. Drug retention may actually prevent premature loss before reaching the target site, as long as drug release or uptake by target cells is facilitated through some mechanism. Studies have shown that the trafficking and targeting of stable drug–bacterial ghost combinations may involve fusion of the bacterial structure with the cell membranes of target cells, leading to the subsequent release of the cargo, rather than relying on extracellular delivery of the drug [48, 101, 102]. Additionally, spatiotemporal drug release from bacterial ghosts can be externally triggered, such as through photothermal energy [93].
The interaction of MRP-ScGs with A549 cells resulted in cytotoxic effects following cellular uptake (Fig. 8a and b), accompanied by interactions with various intracellular targets. Previous studies have reported that MRP induces apoptosis in gastric cancer cells by scavenging reactive oxygen species [103], which aligns with the molecular targets identified in our study, as several of them are enriched in complement pathways. Additionally, another study suggested that MRP exhibits a synergistic effect when combined with chemotherapeutic drugs such as paclitaxel and cyclophosphamide [104]. This synergy could be attributed to the predicted molecular targets involved in DNA repair, the mitogenic cascade, and epithelial-mesenchymal transition.
Taken together, these findings suggest that MRP could play a potential role as an adjunct in anti-cancer therapy. The enhanced cytotoxicity of MRP toward A549 lung cancer cells and its association with increased cellular uptake via S. cerevisiae microencapsulation further support the potential of ScGs as a drug carrier. Furthermore, the formation of a stable MRP-ScGs bio-microstructure may be advantageous for the application of MRP as a bioimaging agent or biocolorant.

Conclusion
This study demonstrates the successful development of a novel bio-microcapsule system by entrapping MRP within ScGs. The resulting MRP-ScGs formulation combines the therapeutic potential of MRP with the biocompatibility and structural stability of ScGs. Comprehensive morphological and microscopic analyses confirmed effective ghost cell evacuation and uniform MRP encapsulation. Process optimization using Box–Behnken design and response surface methodology achieved a high loading efficiency of over 61%. The encapsulated pigment exhibited a biphasic sustained release profile under physiological conditions, along with significantly enhanced cytotoxic activity against A549 lung cancer cells, attributed to improved cellular uptake. Moreover, in silico analysis suggested multiple molecular targets involved in key cellular processes, supporting the potential of MRP-ScGs in targeted drug delivery. Collectively, these findings highlight the promise of ScGs as a versatile and efficient carrier system for MRP, paving the way for its future application in drug delivery, therapeutic, bioimaging, and biocolorant applications. Nevertheless, MRP release kinetics, the precise mechanisms underlying cellular uptake and intracellular release of MRP from ScGs, as well as the behavior of the formulation in vivo, warrant further investigation.

Supplementary Information
Below is the link to the electronic supplementary material.