NRMJ_10_3_267-285

Introduction

According to the latest statistics provided by the United Nations, hunger is one of the most urgent global problems affecting more than 800 million people worldwide. The impoverished millions of people globally and the shortage of protein-rich food have triggered researchers to find replacements for the conventional expensive protein sources in animal feed such as fishmeal and soymeal (Escribano, 2016). Accordingly, the production of readily accessible protein products without a negative influence on human life is one of the main challenges in this context (Riches, 2002; Jach et al., 2022). Single cell protein (SCP) stands out as an appealing substitute for traditional protein sources in both human diet and animal feed, which can be produced efficiently using various cheap and abundant feedstock to tackle the worldwide food crisis (Tropea et al., 2022; Rasool et al., 2023; Sekoai et al., 2024). This sustainable protein source has the potential to improve food security by supplementing traditional feed sources and reducing reliance on resource-intensive protein production methods. Furthermore, many kinds of yeast and bacterial species such as lactic acid bacteria exhibit many other beneficial properties, including prebiotics, antioxidants, and immunomodulatory effects (Khadem et al., 2026). In addition, SCP can be obtained through the bioconversion of decomposable wastes, contributing to waste reduction and recycling efforts, which confers it superiority over the other available choices (Izah et al., 2019). Previous studies revealed the distinct advantages of microbial protein in various waste treatment practices (Yang et al., 2022; Koukoumaki et al., 2024).

Microbial proteins are sourced from a variety of microorganisms, including fungi, bacteria, yeasts, and algae, with yeasts being the most widely embraced among them (Bajić et al., 2022). This is primarily attributed to their ability to propagate easily using inexpensive raw materials and their simple harvesting process (Bekatorou et al., 2006). Yeasts also offer advantages such as the ability to grow at acidic pH and short growth times leading to a rapid biomass production (Chollom et al., 2017). Baker’s yeast Saccharomyces cerevisiae, a yeast species with distinctive features, plays a crucial role in the diverse biotechnological applications.

Single cell protein production is feasible in controlled environments, ensuring an efficient and scalable final product. The growth rate of microorganisms in a bioreactor may vary depending on the fermentation technique, growth conditions such as availability of nutrients in specific forms and ambiance factors i.e., temperature, aeration, pressure, pH, humidity level in the solid culture, and the presence of contaminants or the absence of competitors (Onyeaka et al., 2023). With respect to S. cerevisiae, multiple studies confirmed that a carbon source concentration and temperature affect the cellular growth decisively (Bertasini et al., 2022; Raita et al., 2022). Furthermore, S. cerevisiae is a facultative anaerobic microorganism and can live both in the presence and absence of oxygen, a feature making it more appropriate for industrial applications (Ishtar Snoek and Yde Steensma, 2007). Generally, effective aeration is a critical component for the yeast propagation process that prevents oxidative stress in cells via facilitating the oxygen availability. Oxygen levels in the medium are elevated by aeration and foster the aerobic metabolism, contributing to higher growth rate (Beugholt et al., 2023). Bertasini et al. (2022) in their experiments on the production of S. cerevisiae using the effluent of candies factory reported that aerobic conditions lead to a greater yeast propagation than anaerobic media (Bertasini et al., 2022). Another study demonstrated that under aerobic condition, the presence of oxygen in the medium resulted in increased ATP production, which enhanced S. cerevisiae metabolism and development (Boender et al., 2011). However, it should be noted that the importance of anaerobic conditions, when performed at the appropriate phase and for a defined duration, can be as remarkable as the aerobic conditions (Chopda et al., 2015). Firstly, anaerobic condition diminishes other potential aerobic microorganisms’ populations, thereby reducing the number of competitors in the medium (Ishtar-Snoek and Yde-Steensma, 2007). As a result, S. cerevisiae can expand more effectively in the absence of competitors and more ethanol will be produced through the fermentation process. Subsequently, accumulation of ethanol in the environment exerts toxic effects, leading to the elimination of the remained potential pathogens, while S. cerevisiae cells exhibit remarkable tolerance to elevated ethanol concentrations (Mazzoleni et al., 2015). This study, building on previous studies, aimed to exploit the advantages of both approaches and improve the process efficiency by integrating the two growth modes within a single system. The proposed method comprised a two-step process, initially conducted under anaerobic condition and subsequently transitioning to aerobic respiration after a certain period of time, when air was introduced into the system. A distinguished advantage of this approach is that both steps were performed in a single device specifically designed for this purpose under controlled conditions. In this novel device, aeration was provided via small nozzles embedded in the walls, allowing direct introduction of air into the medium instead of limiting it to the surface.

Moreover, the successful design of a microbial production process involves optimizing the various parameters affecting the process and the final output (Singh et al., 2017). Due to the time-consuming nature of the process, identifying the optimal levels of these critical parameters may be costly and necessitates numerous experiments. These challenges can be addressed by utilizing the response surface methodology (RSM), which is a statistical analysis tool and a validated method in optimizing the bioprocesses (Ferreira et al., 2023; Lestari et al., 2023). This reliable technique has been employed in many previous studies on SCP production by S. cerevisiae to achieve optimal operating conditions (Accettulli et al., 2023; Shahzad et al., 2024). The commonly applied designs in RSM include the central composite design (CCD) and the Box-Behnken design (BBD) (Said and Amin, 2015). Focusing on the efficiency, multiple previous studies indicated satisfactory results in using BBD for optimization of biomass yield in S. cerevisiae (Salem-Bekhit et al., 2023; Abdel-Fattah et al., 2024). By optimizing the biomass production by S. cerevisiae and its protein content using RSM with a Box-Behnken design in a new device, this study improved our understanding in the area of baker’s yeast biomass processing, with the aim of producing SCP for nutritional applications.

Materials and Methods

Microorganism culture and maintenance

The inactivated bakers’ yeast (S. cerevisiae) used in this study was commercialized by “AB Mauri Food Marketing Co., Ltd., China” and procured from a local supermarket of Kuala Lumpur, Malaysia. First, S. cerevisiae was immersed in a sterile solution of sugar water (1:4 w/w) overnight. In the next step, it was reactivated on petri plates of Yeast Peptone Dextrose agar medium (YPDA) composed of (g/l): yeast extract 10, peptone, dextrose 20 L, and agar 15 to verify obtaining of pure S. cerevisiae colonies (Dunuweera et al., 2021). Afterward, they developed pure colonies were maintained on plate count agar (PCA) (Sigma Aldrich, USA) plates and glycerol as stocks.

Experimental set-up

The yeast propagation was carried out in a pilot-scale device in batch mode under submerged fermentation without stirring. The device was designed and fabricated to control the environmental conditions based on the particular experimental method of this study, which t included both aerobic and anaerobic phases during the propagation process. It was equipped with a power supply, and comprised of an insulated vessel, a magnetic door, a temperature control system, and several nozzles (Figure 1). The insulated vessel and magnetic door were designed to prevent contamination and unintended air intrusion into the experimental medium. The temperature control system consisted of a thermostat-controlled heater and an external thermometer. The thermostat constantly monitored the medium temperature, and once the temperature fell below the predefined level (25, 30, and 35 oC), the thermostat instructed the heater to turn on. In contrast, after the temperature reached or surpasses the set point, the thermostat automatically deactivated the heating element. To ensure reliable and precise temperature control, an external thermometer was also utilized. The display unit was installed on the external surface of the device, while the sensor was directly in contact with the yeast production medium. The device was also equipped with an air pump connected to the nozzles through narrow hoses, enabling control over the air flow in the medium. It should be noted that the device was constructed from a fully insulated material and equipped with a heavy, solid magnetic door that ensured an airtight seal when closed. Consequently, these nozzles were the only pathway for air to enter the medium, which were activated according to the experimental design. Furthermore, the nozzles were integrated into the device design to enhance aeration through producing smaller air bubbles under water.

In this study, aeration duration was used as a proxy variable for dissolved oxygen (DO), providing a practical estimate of oxygen availability. All experiments consisted of two phases; an initial anaerobic phase and a subsequent aerobic phase. Each experiment had a total duration of 70 h that was set for 25-45, 35-35, and 45-25 h time intervals under anaerobic and aerobic conditions, respectively. To provide further explanation, there were three groups of experiments that all began under anaerobic condition for 25, 35, and 45 h, respectively, and then they continued under aerobic conditions for 45, 35, and 25 h in the same order given. In the first experimental phase the air pump was off and at the commencement of the second phase, it was turned on to change the medium from anaerobic to aerobic.

The total volume of the device was 90 l and the volume of medium was 35 l. The medium consisted of MnSO4, KH2PO4, ZnSO4, and MgSO4, each at a concentration of 1 g/l. Sucrose was the available carbon source that was mixed at different concentrations (10, 20, and 30 g/l) based on the experimental design. Ammonium sulfate was also added at the concentration of 0.3 g/l as a primary nitrogen source. The prepared medium was adjusted to pH 5.0 and sterilized at 120 oC for 15 min. Starter cultures were prepared using PCA plates and were incubated at 35 oC for 48 h. After incubation, S. cerevisiae colonies were/ transferred to 5 ml distilled water under aseptic conditions (Hezarjaribi et al., 2016). Then, 100 ml of the prepared growth medium were inoculated with 1 ml of S. cerevisiae suspension containing 1×108 CFU/ml and incubated at 35 oC for 48 h. The aforementioned medium was used as a seed culture for inoculating 2 L of the experimental growth medium, followed by incubation at 35 oC for 24 h. The main growth medium was inoculated using the latter culture (Arous et al., 2016). During all steps, the yeast’s cell concentration was kept constant to maintain an appropriate inoculation density in the final medium.

Determination of dry biomass yield

To determine the total dry biomass yield, the washed cells obtained from the incubation process were oven-dried (Memmert UN55 Natural Convection Lab Oven, Germany) to a constant weight at 50 oC for 16 h (Thiviya et al., 2022). The weight of produced biomass was measured using a standard precision balance (ViBRA ALE-503 Class II NTEP, Shinko Denshi Co., Ltd., Japan) with an accuracy of 0.001g.

Determination of protein content

The Kjeldah method was applied to measure the protein content of the experimental samples. The first step was digestion and conducted using concentrated sulfuric acid (H2SO4) and two copper catalyst tablets. A heat block (Behr K Series Kjeldahl Block Digestion System, Behr Labor-Technik, Germany) was used for this process conducted at 420 oC for 2 h. Then Kjeldahl distillation system (Behr Labor-Technik, Germany) was used for carrying out the distillation and titration steps, after cooling. In the distillation step, boric acid solution was used for capturing the produced ammonia gas. Finally, the titration was carried out using H2SO4 as standardized acid solution and the protein content of each sample was calculated using a conversion factor of 6.25 (Shahzad et al., 2024).

Statistical analysis

The experimental data were analyzed using the BBD design of RSM to evaluate the effects of three independent variables (i.e., temperature, aeration duration, and sucrose concentration) on S. cerevisiae growth activity and protein production. A quadratic polynomial model was fitted to the experimental results to define the correlation between the desired responses and the selected parameters. Analysis of variance (ANOVA) was employed to assess the statistical significance and adequacy of the developed models. Additionally, regression coefficients were used to determine the magnitude and direction of the effects of variables and their interactions on biomass and protein production. Response surface plots and contour plots were generated to visualize the interactions between variables and identify optimal conditions. Design-Expert software (version 13.0.5.0, StatEase Inc, Minneapolis, USA) was applied to perform all statistical analysis.

Results and Discussion

Experimental design

In this study, BBD that is one of the most popular experimental designs in RSM was employed to investigate and optimize the effective parameters in S. cerevisiae SCP production. Temperature (25, 30, 35 oC), aeration duration (25, 35, 45 h), and sucrose concentration (10, 20, 30 g/l) were chosen as independent variables, whereas the process responses set as the dependent variables were total biomass yield (R1) and the protein content (R2) (Table 1).

A total number of 15 experimental runs including three center points (applied for experimental error appraisal) were performed (Table 1). Each experiment was conducted in triplicate and the mean value was calculated. As the main step of the optimization analysis, a second-order polynomial model was established for each response while the experimental data was subjected to a stepwise regression analysis via applying a backward elimination subroutine (only the significant terms with p < 0.05 were picked). The analysis of variance (ANOVA) and residual analysis were specified and the correlation (R) and determination (R2) coefficients were assessed for each regression equation. The relationships between the responses and the experimental variables were visualized graphically by contour and 3-D response surface plots.

Table 1: Coded levels of independent variables, experimental design matrix using BBD and results.

Independent variables				Symbols	Level
					-1	0	+1
Temperature (oC)				A	25	30	35
Aeration duration (h)				B	25	35	45
Sucrose concentration (g/l)				C	10	20	30
Run	Variables			Responses
	Temperature (oC)	Aeration duration (h)	Sucrose Conc. (g/l)	Total Biomass (g/l)		Protein content (% w/w)
	Temperature (oC)	Aeration duration (h)	Sucrose Conc. (g/l)	Experimental	Predicted	Experimental	Predicted
1	25	25	20	10.05	9.90	25.75	25.59
2[a]	30	25	30	10.57	10.60	26.13	26.11
3	30	35	20	18.25	18.31	38.75	38.98
4	35	25	20	10.52	10.64	25.43	25.69
5	30	45	10	11.73	11.70	27.3	27.32
6	25	35	30	13.61	13.74	29.47	29.65
7[a]	25	35	10	10.89	11.04	26.48	26.72
8	30	45	30	13.08	13.07	28.78	28.86
9	35	45	20	13.23	13.39	28.88	29.04
10	30	35	20	18.39	18.31	39.15	38.98
11	30	35	20	18.29	18.31	39.05	38.98
12	35	35	30	13.57	13.42	29.19	28.95
13	25	45	20	11.75	11.63	27.35	27.09
14[a]	35	35	10	13.99	13.86	29.65	29.47
15	30	25	10	9.7	9.71	25.29	25.22

Where; [a]: Center point

The general equation of the quadratic model is demonstrated in Equation 1 (Nazari et al., 2017).

Where; Y = The response, β0 = The intercept coefficient, βi = The slope or linear effect of the factor Xi, βii = The quadratic effect of the factor Xi, βij = The interaction effect between the input factors Xi and Xj where ε is the residual term.

Statistical analysis and model fitting

Table 1 represents the experimental design matrix comprised of values and levels of independent variables, along with the values of total biomass yield and protein content as responses. The observed considerable changes in SCP production rate by S. cerevisiae for different combinations of variables indicated that they remarkably affected the yeast growth. According to the experimental results, the optimum values of temperature, aeration duration, and sucrose concentration were found to be 30 oC, 35 h, and 20 g/L, respectively, to achieve the maximum SCP production and protein content (Table 1). The following equations are the final developed models used to predict the values of total biomass yield and protein content with coded parameters:

Total Biomass Yield (g/L DW) = +18.31 + 0.6262 A + 1.12 B + 0.5650 C + 0.2525 AB – 0.7850 AC – 0.1200 BC – 2.59 A2 – 4.33 B2 – 2.71 C2 …(2)

Protein Content % (w/w) = +38.98 + 0.5125 A + 1.21 B + 0.6063 C + 0.4625 AB – 0.8625 AC + 0.1600 BC – 5.15 A2 – 6.98 B2 – 5.13 C2 …(3)

Where; A = Temperature, B = Aeration duration, C = Sucrose concentration

In these equations, the positive sign demonstrates the synergistic effect of the variables, whereas the negative ones indicate their opposed effect. The adequacy and statistical significance of the model were evaluated through the analysis of variance (ANOVA) (Celebi et al., 2007; El-Naggar and El-Ewasy, 2017) and the outcomes of ANOVA for the quadratic equation are presented in Table 2.

In RSM, ANOVA plays a crucial role in assessing the importance of the primary effects of the independent variables and the potential interactions among them, which entails fitting a regression model to the data and subsequently conducting F-tests to evaluate the significance of each term in the model. The ANOVA results affirmed the accuracy of the equation and true relationship existing between the response and the significant variables depicted in the equation. The values of F and p serve as indicators of the significance of the coefficient term, with higher F values and lower p values indicating a greater significance (Ramli et al., 2017). In ANOVA, regression coefficients have the essential function in determining the relation between the input variables (factors) and the response variable. They quantify the change in the response variable for a unit change in the corresponding input variable while fixing the other variables constant. The coefficient of determination (R²) defines the validity of the model in predicting the outcomes of the model or responses. In the context of regression analysis, R2 is a statistical indicator of how closely the regression line resembles the actual data. The lowest value of R2 is 0, which means that the points are not defined by the regression, whereas the highest possible value is 1 indicates that a regression line defines all points.

In this study, the coefficient of determination (R2) for both responses was close to 1 (0.9987 for both biomass yield and protein content), representing the desirable adjustment of the developed models to the experimental data. Additionally, the high values of adjusted-R2 for total biomass and protein content (0.9963 for both responses) confirmed that the models were highly significant. Since the quadratic models were fitted to a limited number of experimental runs, potential overfitting was assessed. A recorded difference of less than 0.2 between the predicted R2 and the adjusted R2 indicated that the predicted R2 was in proper agreement with the adjusted R2 for both responses. This reasonable agreement and the non-significant lack of fit tests demonstrated that the models adequately captured the underlying trends without overfitting. P-value indicated the probability of obtaining a result as extreme as or more extreme than what was actually observed, assuming that there was no impact or difference (null hypothesis). In a statistically significant test, p-value less than 0.05 means that the model terms are significant (Kim et al., 2003). As shown in Table 2, the p-value for most of the total biomass model terms (A, B, C, AB, AC, A², B², and C²), as well as the protein model terms (A, B, C, AB, AC, A², B², and C²) were lower than 0.05,

Table 2: Analysis of variance (ANOVA) for quadratic models for total biomass yield and protein content of Saccharomyces cerevisiae.

Source		Biomass				Source		Protein
	Sum of squares	df	Mean square	F-value	p-value	Source	Sum of squares	df	Mean square	F-value	p-value
Model	124.99	9	13.89	414.98	< 0.0001 (significant)	Model	347.59	9	38.62	416.18	< 0.0001 (significant)
Temperature (A)	3.14	1	3.14	93.76	0.0002	Temperature (A)	2.10	1	2.10	22.64	0.0051
Aeration duration (B)	10.01	1	10.01	299.20	< 0.0001	Aeration duration (B)	11.79	1	11.79	127.00	< 0.0001
Sucrose conc. (C)	2.55	1	2.55	76.31	< 0.0003	Sucrose conc. (C)	2.94	1	2.94	31.68	0.0025
AB	0.2550	1	0.2550	7.62	0.0398	AB	0.8556	1	0.8556	9.22	0.0289
AC	2.46	1	2.46	73.66	0.0004	AC	2.98	1	2.98	32.07	0.0024
BC	0.0567	1	0.0567	1.72	0.2465	BC	0.1024	1	0.1024	1.10	0.3416
A²	24.74	1	24.74	739.41	< 0.0001	A²	98.09	1	98.09	1057.00	< 0.0001
B²	69.35	1	69.35	2072.21	< 0.0001	B²	179.72	1	179.72	1936.66	< 0.0001
C²	27.04	1	27.04	808.06	< 0.0001	C²	97.23	1	97.23	1047.79	< 0.0001
Residual	0.1673	5	0.0335			Residual	0.4640	5	0.0928
lack of Fit	0.1569	3	0.0523	10.06	0.0918 (not significant)	Lack of Fit	0.3773	3	0.1258	2.90	0.2667 (not significant)
Pure Error	0.0104	2	0.0052			Pure Error	0.0867	2	0.0433
R2	0.9987					R2	0.9987
Adjusted-R2	0.9963					Adjusted-R2	0.9963
Predicted-R2	0.9798					Predicted-R2	0.9821
Adequate precision	57.6021					Adequate precision	55.3550
C.V. %	1.39					C.V. %	1.02
Total	125.15	14				Total	348.05	14

displaying that the regression equations illustrated the existing correlation between the response parameters and the variable parameters competently.

In the current study, adequate precision which measured the signal-to-noise ratio (S/N) was favorable (greater than four), recommending that the developed models can be used to navigate the design space and predict the responses (Isam et al., 2019). Furthermore, the validity and accuracy of the experimental design as well as the little deviation between experimental and predicted values was revealed by low coefficient of variance (CV) values (1.39 for total biomass yield and 1.02 for protein content) (Isam et al., 2019). This observation also indicated acceptable reproducibility of the experimental measurements. Additionally, a non-significant lack of fit (p < 0.05) for both responses revealed that the models adequately fitted the experimental data (Anuar et al., 2013). These results confirmed the reliability and robustness of the developed response surface models.

Model adequacy analysis

According to the data expanded out on the diagonal lines of the current study there was an obvious nonlinear relationship between the experimental variables, approving the selected quadratic models. In addition, the adequacy of the regression model could be assessed through analyzing diagnostic plots such as normal plot of residuals graph, predicted versus actual graph, residual versus run graph, and residuals versus predicted graph. The residual values appeared to be randomly scattered within the range while comprising low values. As shown in Figure 2A for total biomass yield and Figure 3A for protein content, the data points were close to the linear regression line without deviation, indicating a valid and normal data distribution on the normal plot of the residual graph. The points in the predicted versus actual graph of both responses followed a diagonal line, indicating that the predicted responses matched the observed values (Figures 2B, 3B). The graph of residuals versus

run shown in Figures 2C and 3C for total biomass yield and protein content, respectively, displayed that all data points were distributed randomly, elucidating the fitness of the models (Flaifel, 2020). The absence of obvious pattern and structure in residuals versus predicted graph is one of the most important criteria in determining the adequacy of a model (Anuar et al., 2013), which can be observed in the mentioned graph of this study, demonstrated in Figures 2D and 3D for total biomass yield and protein content, respectively.

Interaction between influential parameters

The experimental results provided substantial perspective on the interaction effects of the key variables on SCP production by S. cerevisiae and the protein content. The contour and 3-D response surface plots validated the statistical significant of the optimization strategies as well as the predictive accuracy of the regression models. The combined effects of key variables as well as critical regions of maximum efficiency were illustrated on contour plots (Figures 4 and 5).

The figures representing the growth rate of S. cerevisiae (total biomass) and the protein content of the produced biomass were a function of two experimental variables, while the third variable remained constant at the center point. The first contour plot of both responses (Figures 4A and 5A) showing the interaction between aeration duration and temperature indicated a critical threshold beyond which the response declined. This could be attributed to thermal stress or oxygen saturation effects, as S. cerevisiae growth and metabolism was remarkably influenced by the temperature and oxygen level of the growth culture. According to Takagi (2021), the thermal stress affected the metabolic processes and cellular functions of S. cerevisiae, leading to growth declines. Generally, the temperature between 28 oC to 33 oC is the preferred temperature range for Saccharomyces yeasts growth (Salvadó et al., 2011). Razzaq et al. (2020) reported that the maximum protein production by S. cerevisiae was observed at 30 oC, where further temperature elevation led to a decrease in final yield. In fact, the results of this study align with the previous findings where contour plots displayed a substantial enhancement in both biomass production and protein content at the temperature around 30 oC (Figures 4A, B and 5A, B).

Depending on the oxygen levels, S. cerevisiae expressed different growth behavior. For instance, its specific growth rate is considerably reduced in a microaerophilic condition (low oxygen levels of 2-10%) (Carbó et al., 2015). Conversely, the excessive oxygen level can cause detrimental effects (i.e., imbalanced ion distribution and cell leakiness) (Rintala, 2010). As presented in Figure 5A, C, the protein content of S. cerevisiae biomass was also influenced by the oxygen concentration of the growth culture. A previous study revealed that appropriate oxygen levels boosts a metabolic shift in S. cerevisiae that regulates the key genes involved in protein synthesis (Liu et al., 2024). As it has been depicted in Figures 4A, C and 5A, C, the maximum level of biomass yield and protein content was observed when the aeration duration was at an average level of 35 h.

The interaction between sucrose concentration and temperature (Figures 4B and 5B) was a key feature in representing the substrate utilization efficiency. As the contour plots demonstrate, the biomass production and protein content increased with higher sucrose concentration, but only when the concentration was around 20 g/l and within a specific thermal range. Beyond these ranges, excessive sucrose in the medium may inhibit the yeast function because of osmotic stress. Hezarjaribi et al. (2016) used sucrose as the main carbon source in their study for SCP production by S. cerevisiae and concluded that the enhancement of sucrose concentration increased the microbial protein production, in accordance with what was

observed in the current study by increasing the sucrose concentration from 10 g/l to 20 g/l. There was certain extent for this enhancement as further increase in the sucrose concentration of S. cerevisiae growth culture led to a reduction in biomass production and protein content. This result is in accordance with the previous studies conducted on S. cerevisiae, which reported a specific range of tolerable sucrose concentration for this yeast, attributable to sucrose accumulation in the medium and osmotic stress (Kara et al., 2017; Olivares-Marin et al., 2018).

Finally, the interaction between sucrose concentration and aeration duration is presented in Figures 4C and 5C. These contour plots reflected a distinguished interaction region where biomass yield and protein content were maximized respectively. Both of these graphical visualizations proposed an optimal balance between nutrient availability (sucrose concentration of 20 g/l) and oxygen transfer (aeration duration of 35 h) at the average levels of these experimental variables, ensuring an effective metabolic activity. This metabolic activity is essential for maintaining cellular energy production, as it regulates enzymatic reactions that drive ATP synthesis, substrate utilization, and overall metabolic flux (Lee et al., 2023). Adequate oxygen transfer supports the aerobic pathways and increase the process efficiency, while excessive or insufficient oxygen levels may disrupt the metabolic equilibrium, leading to reduced biomass production and protein content. Ribéreau-Gayon et al. (2015) also reported that while supplementing the yeast growth culture with sucrose as the main carbon source, a short period of aeration may lead to enhanced growth, notably after the first growth phase.

Effects of process variables on biomass yield and protein content

Saccharomyces cerevisiae growth and metabolism is influenced by the culture medium constituents and various environmental parameters. According to the experimental results of Box-Behnken design, the biomass yield and protein content ranged from 9.7 to 18.39 g/l and 25.29 to 39.15 % (w/w), respectively. The optimum values of both responses were obtained under the independent variables of 30 oC (temperature), 35 h (aeration duration), and 20 g/l (sucrose concentration). From the ANOVA table, there was a close correlation between the predicted and experimental results, where the maximum predicted values of biomass yield was 18.31 g/l and 38.98 % (w/w) for protein content (Table 1). This indicated the adequacy of the developed models. The correlation existing between each of the total biomass production and the protein content with the experimental variables was depicted in 3-D response graphs (Figures 6 and 7). These three-dimensional visualizations effectively demonstrated the orientation of each factor in influencing yeast growth and metabolism and facilitated the analysis of the developed quadratic models. Each plot represented the impact of the two tested variables within their examined ranges, while the third variable was constant at its optimal level.

The primary insight from the obtained graphs is that the optimum biomass and protein production were observed at the center point of the experimental domain of all the three tested variables, indicating that the range of variables was correctly defined. This reflected that the moderate levels of each factor synergistically contributed to the optimal responses, suggesting a balanced interplay rather than a linear or monotonic relationship. Thus, the suitability of the Box-Behnken design has been validated through placing the experimental runs at midpoints of edges (center) strategically to capture the interaction effects effectively.

As the convex surface of the 3-D plots indicates, the deviations from the medium levels in either direction (higher or lower) resulted in a decline in biomass and protein production. For instance, according to the first graph (Figures 6A and 7A), both the total biomass yield and the protein content peaked at intermediate values of temperature and aeration duration, while sucrose concentration was kept constant at the center level. Similarly, in the second graph (Figures 6B and 7B), the highest values of both responses were observed in the middle level of the sucrose concentration and temperature, while the aeration duration was kept constant at the center point. This suggests that an excessive or insufficient heat could limit the yeast growth and protein production, possibly due to the pivotal role of the temperature in metabolism and enzymatic activities of the S. cerevisiae. This aligns with the findings of another study performed on this yeast, where the temperatures of 32.3 oC and 45.4 oC have been reported as the highest optimum and maximum temperature for the baker’s yeast growth and protein production, respectively (Salvadó et al., 2011). The results of the current study also indicated that an optimal temperature of 30 oC was identified for both responses, while the growth and protein production continued by a temperature around 35 oC (i.e., the highest temperature examined in this study). On the other hand, several studies have reported a remarkable decrease in S. cerevisiae growth and protein production at a temperature below 30 oC (Feng et al., 2018; Lip et al., 2020). The optimal temperature of 30 oC most likely reflected the balance that existed between enzymatic activity and thermal stability in yeast metabolism. At higher temperatures, reduced yield may be attributed to protein denaturation and increased cellular stress (Hezarjaribi et al., 2016). From an industrial perspective, this temperature is advantageous as it minimizes the cooling costs while maintaining high productivity.

Regarding the influence of aeration duration on S. cerevisiae, as presented in Figures 6A, C and 7A, C, both the biomass and the protein production had maximized in the middle level of aeration duration. This could be attributed to the necessity of an adequate oxygen transfer for aerobic growth of S. cerevisiae and its metabolic stability. Insufficient oxygen availability affects the cellular processes and may lead to reduced

biomass and protein yield. In fact, S. cerevisiae requires a moderate level of oxygen for its optimal growth and protein production (Bisschops et al., 2015; Chopda et al., 2015). Various aeration tests (from micro-aerobic to aerobic cultures) declared that efficient aeration was critical for optimal productivity of protein and biomass production by the baker’s yeast in fed-batch processes (Bisschops et al., 2015). This may be assigned to the use of oxygen molecules in the components of the yeast cell membrane, i.e., unsaturated fatty acids and sterols. Furthermore, the chronological lifespan of S. cerevisiae is considerably affected by aeration, as inadequate aeration may lead to stress resistance in the stationary phase cultures, causing a reduced lifespan (Bisschops et al., 2015). Although S. cerevisiae as a facultative microorganism can perform fermentation and grow anaerobically, its biomass and protein yields are typically lower under anaerobic fermentation conditions compared to respiration. In this study, during the early experimental phase and in the absence of aeration, fermentation was the predominant metabolic pathway. Accordingly, glucose was metabolized via glycolysis and converted to ethanol (C2H5OH) and carbon dioxide (CO2), yielding only two ATP molecules per a glucose molecule. In the late phase, exposure to air triggered a metabolic shift in S. cerevisiae and activated several respiratory pathways, including the tricarboxylic acid (TCA) cycle and the oxidative phosphorylation, increasing ATP yield by 38 molecules per a glucose molecule. This metabolic transition may explain the observed differences in product yield under varying aeration conditions, as revealed by the optimization results of the current study (Bisschops et al., 2015; Mazzoleni et al., 2015).

Despite the fact that yeasts generally benefit from oxygen availability by supporting oxidative protein folding, but extreme aeration can cause damage in their intracellular structures such as DNA and disruption of various cellular functions. Prolonged exposure to high oxygen levels may negatively affect S. cerevisiae viability through oxidative stress. Enzyme inactivation, protein carbonylation, and disruption of some metabolic pathways are some of the consequences of this oxidative damage (Morano et al., 2012).

Likewise, the correlation between sucrose concentration and either temperature or aeration duration exhibited a dome-shaped response (Figures 6B, C and 7B, C), highlighting that an adequate carbon supply is crucial to optimize the biomass and the protein yield. Maintaining the carbon source at an appropriate level can avoid substrate inhibition or nutrient depletion. Thus, it directly influences the final processes of yields and metabolic efficiency. The optimum sucrose concentration for the growth of S. cerevisiae may differ based on the strain and the processing conditions (Shafaghat et al., 2010). In the current study, the sucrose concentration was examined in the range of 10 to 30 g/l and the highest value of biomass and protein production was observed at a concentration of 20 g/l. As the plots display, both responses increased with the enhancement of sucrose concentration up to this threshold (20 g/l), beyond which an inhibitory effects occurred possibly due to osmotic stress or substrate saturation. When the concentration of a carbon source in S. cerevisiae culture exceeds an optimal level, the resulting hyperosmotic stress may lead to the production of stress-response metabolites and subsequent water efflux and cellular dehydration. Furthermore, prolonged stress disrupts the nutrient uptake by inactivating the related transporters in the process. In addition, the biosynthetic pathways for certain amino acids may be downregulated, reflecting a cellular shift towards stress survival rather than growth (Marques et al., 2016).

Overall, the related 3-D plots of both responses confirmed the presence of remarkable quadratic effects, as captured by the second-order polynomial models. The peak observed at the center of these plots indicated that the response was sensitive to deviations from these medium levels, where either an increase or a decrease in any variable may lead to suboptimal biomass and protein production.

Conclusions and Recommendations

This study demonstrated that the propagation of S. cerevisiae under a specific combination of aerobic and anaerobic growth conditions can meaningfully influence the final yield and different quality criteria, such as its protein content. The findings indicate that an optimal SCP production by S. cerevisiae is a result of balanced correlation between temperature, aeration duration, and sucrose concentration. According to the coefficient of linear terms of ANOVA for both responses, the aeration duration was found to be the most influential parameter affecting the process responses. Thus, maintaining an appropriate oxygen level ensures improvement of bioprocess outputs and metabolic efficiency. Although the oxygen molecules are presented in various cellular structures and low oxygen availability may limit further growth and metabolic stability, high aeration duration can inhibit S. cerevisiae biomass and protein production due to oxidative stress.

While the results of the current study were satisfactory for the optimization process of S. cerevisiae production at pilot scale, reduced oxygen transfer rate (OTR) and lower volumetric mass transfer coefficients (kLa) can considerably affect the metabolic behavior of the baker’s yeast in larger bioreactors. Thus, further studies are recommended to evaluate oxygen transfer dynamics and process performance under industrial scale conditions, exploring the baker’s yeast deployment across diverse industrial, microbiological, and biotechnological contexts.

Acknowledgement

This work was supported by Fundumental Research Grant (03-01-18-1940FR) Universiti Putra Malaysia.

Novelty Statement

This study proposes an optimized protocol for Saccharomyces cerevisiae propagation in both aerobic and anaerobic conditions, leading to enhancement of metabolic efficiency and protein yield. The obtained results highlight the crucial role of controlled aeration in S. cerevisiae growth and are expected to benefit food and feed industry by providing a deeper understanding of the yeast’s performance during large-scale fermentation processes.

Author’s Contribution

Zahra Safdari: Conceptualization, investigation, methodology, data curation, formal analysis, validation, writing original draft.

Mohd Khairol Anuar Ariffin: Supervision, conceptualization, resource, methodology, validation, reviewing and editing, project administration

Mohamed Thariq Hameed Sultan: Conceptualization, reviewing and editing.

Eris Elianddy Supeni: Reviewing and editing.

Ethical approval

No ethical approval is required.

Funding source

This research did not receive funding from any scientific organization.

Generative AI and AI-assisted technology statement

No generative AI or AI-assisted technology was employed in this study.

Conflict of interests

The authors have declared no conflicts of interest.

References

Abdel-Fattah, Y., Sallam, L. and Diekmann, H., 2024. Box-behnken experimental design for optimization of biomass and ethanol yield in Saccharomyces cerevisiae. Future Perspect. Med. Pharm., 1(3): 9-14. https://doi.org/10.21608/fpmpeb.2024.391509

Accettulli, A., Sinigaglia, M. and Racioppo, A., 2023. Biomass production and use of Saccharomyces cerevisiae var. boulardii in a beverage for athletes. Biol. Life Sci. Forum, 26(1): 6. https://doi.org/10.3390/Foods2023-15108

Anuar, N., Adnan A.F.M., Saat, N., Aziz, N. and Mat Taha, R., 2013. Optimization of extraction parameters by using response surface methodology, purification, and identification of anthocyanin pigments in Melastoma malabathricum fruit. Sci World, pp. 1–10. https://doi.org/10.1155/2013/810547

Arous, F., Azabou, S., Jaouani, A., Zouari-Mechichi, H., Nasri, M. and Mechichi, T., 2016. Biosynthesis of single-cell biomass from olive mill wastewater by newly isolated yeasts. Environ. Sci. Pollut. Res., 23: 6783-6792. https://doi.org/10.1007/s11356-015-5924-2

Bajić, B., Vučurović, D., Vasić, Đ., Jevtić-Mučibabić, R. and Dodić, S., 2022. Biotechnological production of sustainable microbial proteins from agro-industrial residues and by-products. Foods, 12(1): 107. https://doi.org/10.3390/foods12010107

Bekatorou, A., Psarianos, C. and Koutinas, A. A., 2006. Production of food grade yeasts. Food Technol. Biotechnol., 44(3). Available online: https://hrcak.srce.hr/file/162096 (accessed on 12 May 2025).

Bertasini, D., Binati, R.L., Bolzonella, D. and Battista, F., 2022. Single cell proteins production from food processing effluents and digestate. Chemosphere, 296: 134076. https://doi.org/10.1016/j.chemosphere.2022.134076

Beugholt, A., Geier, D.U. and Becker, T., 2023. Improvement of Saccharomyces propagation performance through oxygen-enriched air and aeration parameter variation. Front. Chem. Eng., 5: 1193230. https://doi.org/10.3389/fceng.2023.1193230

Bisschops, M.M., Vos, T., Martínez-Moreno, R., Cortés, P.T., Pronk, J.T. and Daran-Lapujade, P., 2015. Oxygen availability strongly affects chronological lifespan and thermotolerance in batch cultures of Saccharomyces cerevisiae. Microb. Cell, 2(11): 429. https://doi.org/10.15698/mic2015.11.238

Boender L.G., Almering, M.J., Dijk, M., van Maris, A.J., de Winde, J.H., Pronk, J.T. and Daran-Lapujade, P., 2011. Extreme calorie restriction and energy source starvation in Saccharomyces cerevisiae represent distinct physiological states. BBA-Mol. Cell. Res., 1813(12): 2133-2144. https://doi.org/10.1016/j.bbamcr.2011.07.008

Carbó, R., Ginovart, M., Carta, A., Portell, X. and Del Valle, L.J., 2015. Effect of aerobic and microaerophilic culture in the growth dynamics of Saccharomyces cerevisiae and in training of quiescent and non-quiescent subpopulations. Arch. Microbiol., 197(8): 991-999. https://doi.org/10.1007/s00203-015-1136-x

Celebi, O., Üzüm, Ç., Shahwan, T. and Erten, H.N., 2007. A radiotracer study of the adsorption behavior of aqueous Ba2+ ions on nanoparticles of zero-valent iron. J. Hazard. Mater., 148(3): 761-767. https://doi.org/10.1016/j.jhazmat.2007.06.122

Chollom, P.F., Agbo, B.E., Doma, D.U., Okojokwu, J.O. and Yisa, A.G., 2017. Nutritional value of spent brewers’ yeast (Saccharomyces cerevisiae): A potential replacement for soya bean in poultry feed formulation.

Chopda, V.R., Rathore, A.S. and Gomes, J., 2015. Maximizing biomass concentration in baker’s yeast process by using a decoupled geometric controller for substrate and dissolved oxygen. Bioresour. Technol., 196: 160-168. https://doi.org/10.1016/j.biortech.2015.07.050

Dunuweera, A.N., Nikagolla, D.N. and Ranganathan, K., 2021. Fruit waste substrates to produce single‐cell proteins as alternative human food supplements and animal feeds using baker’s yeast (Saccharomyces cerevisiae). J. Food Qual., 2021(1): 9932762. https://doi.org/10.1155/2021/9932762

El-Naggar, N.E.A. and El-Ewasy, S.M., 2017. Bioproduction, characterization, anticancer and antioxidant activities of extracellular melanin pigment produced by newly isolated microbial cell factories Streptomyces glaucescens NEAE-H. Sci. Rep., 7(1): 1-19. https://doi.org/10.1038/srep42129

Escribano, A.J., 2016. Organic livestock farming–challenges, perspectives, and strategies to increase its contribution to the agrifood system’s sustainability: A review. Organic Farming. A Promising Way of Food Production, 1st ed.; Konvalina, P., Ed, 229-260. https://doi.org/10.5772/61272

Feng, L., Jia, H., Qin, Y., Song, Y., Tao, S. and Liu, Y., 2018. Rapid identification of major QTLS associated with near-freezing temperature tolerance in Saccharomyces cerevisiae. Front. Microbiol., 9: 2110. https://doi.org/10.3389/fmicb.2018.02110

Ferreira, N., Viana, T., Henriques, B., Tavares, D. S., Jacinto, J., Colónia, J. and Pereira, E., 2023. Application of response surface methodology and box–behnken design for the optimization of mercury removal by Ulva sp. J. Hazard. Mater., 445: 130405. https://doi.org/10.1016/j.jhazmat.2022.130405

Flaifel, M.H., 2020. An approach towards optimization appraisal of thermal conductivity of magnetic thermoplastic elastomeric nanocomposites using response surface methodology. Polymers 12: 1–16. https://doi.org/10.3390/polym12092030

Hezarjaribi, M., Ardestani, F. and Ghorbani, H.R., 2016. Single cell protein production by Saccharomyces cerevisiae using an optimized culture medium composition in a batch submerged bioprocess. Appl. Biochem. Biotechnol., 179: 1336-1345. https://doi.org/10.1007/s12010-016-2069-9

Isam, M., Baloo, L., Kutty, S. R. M. and Yavari, S., 2019. Optimisation and modelling of Pb (II) and Cu (II) biosorption onto red algae (Gracilaria changii) by using response surface methodology. Water, 11(11): 2325. https://doi.org/10.3390/w11112325

Ishtar Snoek, I.S. and Yde Steensma, H., 2007. Factors involved in anaerobic growth of Saccharomyces cerevisiae. Yeast, 24(1): 1-10. https://doi.org/10.1002/yea.1430

Islam-Shishir, M.R., Taip, F.S., Aziz, N.A., Talib, R.A. and Hossain Sarker, M.S., 2016. Optimization of spray drying parameters for pink guava powder using RSM. Food Sci. Biotechnol., 25: 461-468. https://doi.org/10.1007/s10068-016-0064-0

Izah, S.C., Enaregha, E.B. and Epidi, J.O., 2019. Vitamin content of Saccharomyces cerevisiae biomass cultured in cassava wastewater. MOJ Toxicol., 4(1): 42-45.

Jach, M.E., Serefko, A., Ziaja, M. and Kieliszek, M., 2022. Yeast protein as an easily accessible food source. Metabolites, 12(1): 63. https://doi.org/10.3390/metabo12010063

Kara Ali, M., Outili, N., Ait Kaki, A., Cherfia, R., Benhassine, S., Benaissa, A. and Kacem Chaouche, N., 2017. Optimization of baker’s yeast production on date extract using response surface methodology (RSM). Foods, 6(8): 64. https://doi.org/10.3390/foods6080064

Khadem, H., Boubakeur, B., Adnane, M., Meddah, B., Drabo, M. and Tirtouil, A., 2026. Effect of synthetic flavonoids on Lactobacillus rhamnosus aggregation and exopolysaccharide production. Nov. Res. Microbiol. J., 10(1): 01-14. https://doi.org/10.17582/journal.nrmj/2026/10.1.01.14

Kim, H.K., Kim, J.G., Cho, J.D., and Hong, J.W., 2003. Optimization and characterization of UV-curable adhesives for optical communications by response surface methodology. Polym. Test., 22: 899–906. https://doi.org/10.1016/S0142-9418(03)00038-2

Koukoumaki, D.I., Tsouko, E., Papanikolaou, S., Ioannou, Z., Diamantopoulou, P. and Sarris, D., 2024. Recent advances in the production of single cell protein from renewable resources and applications. Carbon Resour. Convers., 7(2): 100195. https://doi.org/10.1016/j.crcon.2023.07.004

Lee, Y.O., Do, S.H., Won, J.Y., Chin, Y.W., Chewaka, L.S., Park, B.R. and Kim, S.K., 2023. Inverse metabolic engineering for improving protein content in Saccharomyces cerevisiae. Biotechnol. J., 18(9): 2300014. https://doi.org/10.1002/biot.202300014

Lestari, E.N.E., Windarsih, A., Sunardi, S. and Nisa, K., 2023. Encapsulation of Channa striata albumin extract: Optimization by Box-Behnken design of response surface methodology. Int. Aquat. Res., 15(4): 333.

Lip, K.Y.F., García-Ríos, E., Costa, C.E., Guillamón, J.M., Domingues, L., Teixeira, J. and Gulik, W.M., 2020. Selection and subsequent physiological characterization of industrial Saccharomyces cerevisiae strains during continuous growth at sub-and-supra optimal temperatures. Biotechnol. Rep., 26: e00462. https://doi.org/10.1016/j.btre.2020.e00462

Liu, Y., Wu, Y., Lv, X., Li, K., Xiong, J., Liu, X. and Liu, Y., 2024. Improving cellular protein content of Saccharomyces cerevisiae based on adaptive evolution and flow cytometry-aided high throughput screening. J. Agric. Food Chem., 73(1): 706-717. https://doi.org/10.1021/acs.jafc.4c09632

Marques, W.L., Raghavendran, V., Stambuk, B.U. and Gombert, A.K., 2016. Sucrose and Saccharomyces cerevisiae: A relationship most sweet. FEMS Yeast Res., 16(1): fov107. https://doi.org/10.1093/femsyr/fov107

Mazzoleni, S., Landi, C., Cartenì, F., de Alteriis, E., Giannino, F., Paciello, L. and Parascandola, P., 2015. A novel process-based model of microbial growth: Self-inhibition in Saccharomyces cerevisiae aerobic fed-batch cultures. Microb. Cell Fact., 14: 1-14. https://doi.org/10.1186/s12934-015-0295-4

Morano, K.A., Grant, C.M. and Moye-Rowley, W.S., 2012. The response to heat shock and oxidative stress in Saccharomyces cerevisiae. Genetics, 190(4): 1157-1195. https://doi.org/10.1534/genetics.111.128033

Nazari, L., Yuan, Z., Ray, M.B. and Xu, C.C., 2017. Co-conversion of waste activated sludge and sawdust through hydrothermal liquefaction: optimization of reaction parameters using response surface methodology. Appl. Energy., 203: 1-10. https://doi.org/10.1016/j.apenergy.2017.06.009

Olivares-Marin, I.K., Madrigal-Perez, L.A., Canizal-Garcia, M., García-Almendárez, B.E., González-Hernández, J.C. and Regalado-Gonzalez, C., 2018. Interactions between carbon and nitrogen sources depend on RIM15 and determine fermentative or respiratory growth in Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol., 102: 4535-4548. https://doi.org/10.1007/s00253-018-8951-3

Onyeaka, H., Nwaiwu, O., Obileke, K., Miri, T. and Al‐Sharify, Z.T., 2023. Global nutritional challenges of reformulated food: A review. Food Sci. Nutr., 11(6): 2483-2499. https://doi.org/10.1002/fsn3.3286

Raita, S., Kusnere, Z., Spalvins, K. and Blumberga, D., 2022. Optimization of yeast cultivation factors for improved SCP production. Environ. Clim. Technol., 26(1): 848-861. https://doi.org/10.2478/rtuect-2022-0064

Ramli, N.A., Rahman, R.A., Ngadi, N. and Samah, R.A., 2017. Optimisation of fermentation conditions for isobutanol production by Saccharomyces cerevisiae using response surface methodology. Chem. Eng. Trans., 56: 301-306. https://doi.org/10.1007/s11694-020-00498-x

Rasool, K., Hussain, S., Shahzad, A., Miran, W., Mahmoud, K.A., Ali, N. and Almomani, F., 2023. Comprehensive insights into sustainable conversion of agricultural and food waste into microbial protein for animal feed production. Rev. Environ. Sci. Bio., 22(2): 527-562.

Razzaq, Z.U., Khan, M.K.I., Maan, A.A. and Rahman, S.U., 2020. Characterization of single cell protein from Saccharomyces cerevisiae for nutritional, functional and antioxidant properties. J. Food Meas. Charact., 14: 2520-2528. https://doi.org/10.1007/s11694-020-00498-x

Ribéreau-Gayon, P., Larue, F. and Chaumet, P., 2015. The effect of addition of sucrose and aeration to grape must on growth and metabolic activity of Saccharomyces cerevisiae. Visit J. Grapevine Res., 26(4): 208.

Riches, G., 2002. Food banks and food security: welfare reform, human rights and social policy. Lessons from Canada? Soc. Policy Admin., 36(6): 648-663. https://doi.org/10.1111/1467-9515.00309

Rintala, E., 2010. Effects of oxygen provision on the physiology of baker’s yeast Saccharomyces cerevisiae. PhD diss. Espoo, Finland: University of Helsinki, Department of Biological and Environmental Sciences. VTT Publications.

Said, K.A.M. and Amin, M.A.M., 2015. Overview on the response surface methodology (RSM) in extraction processes. J. Appl. Sci. Process Eng., 2(1). https://doi.org/10.33736/jaspe.161.2015

Salem-Bekhit, M.M., Riad, O.K.M., Selim, H.M.R.M., Tohamy, S.T.K., Taha, E.I., Al-Suwayeh, S.A. and Shazly, G.A., 2023. Box–Behnken Design for Assessing the Efficiency of Aflatoxin M1 Detoxification in Milk Using Lactobacillus rhamnosus and Saccharomyces cerevisiae. Life, 13(8): 1667. https://doi.org/10.3390/life13081667

Salvadó, Z., Arroyo-López, F.N., Guillamón, J.M., Salazar, G., Querol, A. and Barrio, E., 2011. Temperature adaptation markedly determines evolution within the genus Saccharomyces. Appl. Environ. Microbiol., 77(7): 2292-2302. https://doi.org/10.1128/AEM.01861-10

Sekoai, P.T., Roets-Dlamini, Y., O’Brien, F., Ramchuran, S. and Chunilall, V., 2024. Valorization of food waste into single-cell protein: an innovative technological strategy for sustainable protein production. Microorganisms, 12(1): 166. https://doi.org/10.3390/microorganisms12010166

Shafaghat, H., Najafpour, G.D., Rezaei, S.P. and Sharifzadeh, M., 2010. Optimal growth of Saccharomyces cerevisiae (PTCC 24860) on pretreated molasses for ethanol production: Application of response surface methodology. Chem. Ind. Chem. Eng. Q., 16(2): 199-206. https://doi.org/10.2298/CICEQ100201029S

Shahzad, H.M.A., Asim, Z., Mahmoud, K.A., Abdelhadi, O.M., Almomani, F. and Rasool, K., 2024. Optimizing cultural conditions and pretreatment for high-value single-cell protein from vegetable waste. Process Saf. Environ. Prot., 189: 685-692. https://doi.org/10.1016/j.psep.2024.06.139

Singh, V., Haque, S., Niwas, R., Pasupuleti, M. and Tripathi, C.K.M., 2017. Strategies for fermentation medium optimization: An in-depth review. Front. Microbiol., 7: 227613. https://doi.org/10.3389/fmicb.2016.02087

Takagi, H., 2021. Molecular mechanisms and highly functional development for stress tolerance of the yeast Saccharomyces cerevisiae. Biosci. Biotechnol. Biochem., 85(5): 1017-1037. https://doi.org/10.1093/bbb/zbab022

Thiviya, P., Gamage, A., Kapilan, R., Merah, O. and Madhujith, T., 2022. Single cell protein production using different fruit waste: A review. Separations, 9(7): 178. https://doi.org/10.3390/separations9070178

Tropea, A., Ferracane, A., Albergamo, A., Potortì, A.G., Lo Turco, V. and Di Bella, G., 2022. Single cell protein production through multi food-waste substrate fermentation. Fermentation, 8(3): 91. https://doi.org/10.3390/fermentation8030091

Yang, R., Chen, Z., Hu, P., Zhang, S. and Luo, G., 2022. Two-stage fermentation enhanced single-cell protein production by Yarrowia lipolytica from food waste. Bioresour. Technol., 361: 127677. https://doi.org/10.1016/j.biortech.2022.127677