Assessment of yeast physiology during industrial‐scale brewing practices using the redox‐sensitive dye resazurin

Beer refermentation in bottles is an industrial process utilized by breweries where yeast and fermentable extract are added to green beer. The beer is refermented for a minimum of 2 weeks before distribution, with the physiological state of the yeast a critical factor for successful refermentation. Ideally, fresh yeast that is propagated from a dedicated propagation plant should be used for refermentation in bottles. Here, we explored the applicability of the fluorescent and redox‐sensitive dye, resazurin, to assess cellular metabolism in yeast and its ability to differentiate between growth stages. We applied this assay, with other markers of yeast physiology, to evaluate yeast quality during a full‐scale industrial propagation. Resazurin allowed the discrimination between the different growth phases in yeast and afforded a more in‐depth understanding of yeast metabolism during propagation. This assay can be used to optimize the yeast propagation process and cropping time to improve beer quality.

The development of staling flavors in beer are influenced by the production, packaging, and storage parameters . Beer staling can be minimized using cooled transport and storage and reducing the presence of oxygen during brewing and bottling Wauters et al., 2021). Alternative measures to reduce beer ageing is bottle refermentation (or bottle conditioning). Bottle refermentation is a process that involves secondary fermentation through the addition of fermentable extract (priming sugar) and yeast to the bottle (Wauters et al., 2021). The practice of bottle refermentation is a traditional process used to naturally carbonate the beer, provide flavor stability, and potentially decelerate the aging process (Saison et al., 2010(Saison et al., , 2011. Yeast influences the flavor stability by acting as an oxygen scavenger to decrease oxidative reactions and reducing aldehydes and ketones to their less flavor-active alcohol counterparts during refermentation (Saison et al., 2010;Vanderhaegen et al., 2003). It is imperative that the yeast maintains a healthy state for an extended period after refermentation is complete, as yeast autolysis leads to the formation of negative beer attributes such as poor foam stability and decreased shelf-life (Marconi et al., 2016). It has been shown that the longevity of yeast during bottle refermentation depends on the yeast strain (Qin & Lu, 2006), the priming sugar (Vanbeneden et al., 2006), and the strategy to produce the yeast (Van Landschoot et al., 2004).
The preferred method of yeast production for bottle refermentation is the use of a propagation plant, as it allows the preadaptation of yeast to conditions they will encounter in bottle refermentation, such as low pH, high ethanol, and nutrient availability (Rogers et al., 2016) and the lower risk of microbial contamination (Boulton & Quain, 2006). Typically, brewery propagations are a batch process that involve the volumetric scale-up of yeast from laboratory agar slants (Lodolo et al., 2008). The medium used for brewery propagations is usually wort as it is cost-effective and provides the necessary nutrients for yeast to grow. Despite the tanks being aerated during propagation, the metabolic flux in yeast is driven towards fermentation pathways due to the high sugar content in wort, while the oxygen only supports lipid synthesis (Moutsoglou & Dearden, 2020). After reaching the desired yeast count, the yeast is pitched into a vessel for primary beer fermentation or bottles for refermentation. It has been suggested that the growth phase of yeast for seeding should result in a minimal lag phase during refermentation to prevent prolonged exposure of the beer to oxygen and oxidative reactions (Derdelinckx et al., 1992). Despite these findings, the characterization of yeast physiology during a brewery propagation before bottle refermentation at an industrial scale remains scarce.
Many yeast biomarkers have been investigated to determine the propensity of yeast to perform optimally (Kwolek-Mirek & Zadrag-Tecza, 2014). Current approaches employed by breweries for the evaluation of yeast physiology involve staining techniques in combination with direct microscopy. Methylene blue is classified as the "industry standard" for the determination of live and dead cells (Painting & Kirsop, 1990). It is a cell membrane permeable dye that penetrates both cell types, where live cells reduce the dye into a colorless product, but dead cells remain blue. Alternative dyes, such as trypan blue, are dependent on the integrity of the cell membrane and only enter dead cells with a compromised membrane (McGahon et al., 1995). The sole use of methylene blue or trypan blue as an indicator of yeast physiology is insufficient (Layfield & Sheppard, 2015), as the brewing process exposes yeast cells to environmental changes such as varying oxygen levels or pH, which can impact different physiological parameters and the quality of the assay (Gibson et al., 2007). Evaluating the quality of the yeast population and their performance capabilities in different brewing processes requires the use of multiple complementary assays (cell number, pH, and the intracellular components trehalose and glycogen) (Layfield & Sheppard, 2015;Tapia et al., 2015). Specifically, it has been shown that the accumulation or utilization of intracellular glycogen can provide an overview of the nutritional status of the cell, with glycogen accumulation occurring in yeast upon the limitation of carbon, nitrogen, sulfur, or phosphorus (Lillie & Pringle, 1980). Likewise, trehalose accumulation occurs in response to multiple environmental stresses including nutrient deprivation (Lillie & Pringle, 1980), desiccation (Tapia & Koshland, 2014), heat shock (Pan et al., 2019), oxidative stress (Herdeiro et al., 2006), and ethanol toxicity (Mansure et al., 1994). There are numerous assays commercially available that exploit different cellular processes to measure cellular viability and vitality (Crouch et al., 1993;Kuhn et al., 2003;Levitz et al., 1985). The BacTiter-Glo Microbial Cell Viability Assay has been used to estimate cell viability in bacteria and yeast (Doll et al., 2016;Z. Wang, Wang, et al., 2019). This assay relies on the extraction of adenosine triphosphate (ATP) and its subsequent reaction with luciferin in the presence of Mg 2+ cations to produce a bioluminescent signal (Crouch et al., 1993). Despite the sensitivity of monitoring ATP content in cells, it might be cost-inhibitory.
Fluorescence can also be used to determine the metabolic activity of yeast cells that do not have compromised cellular membranes. For example, fluorescence in conjunction with resazurin (7-hydroxy-10-oxidophenoxazin-10-ium-3-one) has been used as a rapid, cost-efficient, and nontoxic alternative for monitoring cell metabolism for mammalian cells (O'Brien et al., 2000). Resazurin is a blue dye, that when added to live cells, is converted to the reduced pink form (Scheme 1) (Uzarski et al., 2017). The reduction of resazurin to resorufin is mediated by multiple intracellular enzymes (diaphorases) and can be monitored using colorimetry or fluorimetry (O'Brien et al., 2000). Resorufin can be reduced further to the nonfluorescent product, dihydroresorufin (Chen et al., 2018). The reduction of Take-away • Resazurin can discriminate between different growth phases in yeast.
• Resazurin permits deeper understanding of yeast physiology during a brewery propagation.
• Yeast metabolic activity was decreased at the end of an industrial-scale propagation.
resazurin can be used to indicate impairment or changes to cellular metabolism (Czekanska, 2011).
In this study, a cost-effective technique to assess the reductive capacity of yeast within a laboratory setting and compare its reliability to the BacTiter-Glo assay is described. The resazurin assay was used to assess the reductive capacity of yeast during a full-scale industrial propagation. Together with measurements of yeast quality (methylene blue and storage carbohydrates), it was possible to successfully differentiate between lag, logarithmic, and stationary growth phases using resazurin that preceded any detectable changes in viability. This method has the potential to be applied in other food and beverage contexts, where a high-quality product is required.

| Chemicals and yeasts
Acetic acid (for molecular biology), amyloglucosidase from  Corning. Samples were analyzed using a VICTOR 3 1420 Multilabel Counter (PerkinElmer) and measured every 30 min for 24 h. The absorbance of wells containing YPD without yeast cells was subtracted from sample absorbances. Specific gravity was measured using an EasyDens device (Anton Paar).

| Industrial propagation
Yeast from agar slants were streaked onto WLN agar and incubated at 27°C for 3 days. The plate was stored at 4°C until required. A single colony was inoculated into 1 mL of sterilized wort and incubated at room temperature for 24 h with shaking. The yeast culture was serially subcultured to obtain a 1 L culture. The culture was placed into a Carlsberg flask containing 9 L of sterilized wort and incubated at room temperature for 24 h with continuous aeration.
The yeast was transferred to a 10 hL CONTI PROP tank (ESAU & HUEBER) containing sterilized wort at 14°P and incubated at 14°C for 24 h. Yeast was subsequently transferred to a 60 hL CONTI PROP tank containing 40 hL of sterilized wort with an initial gravity of 14°P.
Propagation was performed at 14°C with 13 min homogenization, 7 min aeration, and 20 min rest for a period of 47.5 h.

| Cell counts
Yeast cell counts were performed in duplicate using a hemocytometer.
Samples were diluted in an appropriate amount of PBS buffer and mixed 1:1 with trypan blue (0.2% final concentration). For brewery counts, methylene blue was dissolved in 2% (w/v) trisodium citrate and filtered to obtain a 0.01% (w/v) solution. Sample duplicates were diluted 1:10 with methylene blue solution and counted with a hemocytometer.

| Glycogen and trehalose analysis
Glycogen and trehalose analysis was performed as previously described with minor modifications (Shi et al., 2010). Yeast cells (1 × 10 7 cells) were harvested from the propagator by centrifugation S C H E M E 1 The conversion of resazurin to resorufin and dihydroresorufin. Resazurin (weakly fluorescent, blue) is converted to resorufin (highly fluorescent, pink) by reducing enzymes within the cell. Resorufin can be further reduced to dihydroresorufin (nonfluorescent, colorless). Samples were centrifuged at 20,000g for 10 min and the supernatant analyzed using the glucose (GO) assay kit in a 96-well plate format.
The absorbance was measured at 540 nm using a Multiskan FC (Thermo Fisher Scientific) and SkanIt Software (ver 4.1).

| Alcohol content, apparent extract, and pH
Propagation samples were filtered through a 0.4 µm filter and sonicated for 10 min before the analysis of alcohol content and apparent extract (AE). Alcohol content (%) and AE (%) were determined using an Alcolyzer: Beer Analyzing System (Anton Paar).
The pH was determined using a pH meter from Mettler Toledo.
These measurements were performed as a single replicate. Excitation and emission were performed at 570 nm (±10 nm) and 590 nm (±10 nm), respectively, with automatic gain set at 1000.

| Evaluation of resazurin concentration
The stock solution was diluted in PBS to give 10, 20, and 40 µM resazurin before addition of 1 × 10 6 or 2 × 10 6 yeast cells in triplicate.
The plates were incubated with the various resazurin concentrations with a final volume of 0.2 mL from 1 to 4 h in the dark.

| Evaluation of cell number and incubation time with resazurin
Approximately 0.125 × 10 6 to 3 × 10 6 cells were resuspended in PBS containing 40 µM resazurin in triplicate. The samples were incubated at room temperature in the dark and the fluorescence monitored between 0.5 and 4 h.

| Interaction of brewing media with the resazurin assay
To elucidate the potential impact of culture medium differences (pH and ethanol content), we compared the response of resazurin with PBS, beer, and YPD culture medium (Figure 2). While PBS in the presence of resazurin produced a consistent background fluorescence, YPD showed a higher background fluorescence that increased with longer incubation times. Compared to resazurin in PBS, the quenching of fluorescence in beer containing resazurin was most likely due to the low pH (Bueno et al., 2002). The fluorescence also increased with longer incubation times but remained below that of resazurin in PBS until 4 h. As there were two factors influencing the background readings of resazurin with beer, subsequent brewery samples were washed in PBS and resuspended in the desired resazurin concentration to ensure consistent readings.

| Evaluation of cell density and incubation time for quicker resorufin production
Resazurin is considered nontoxic to mammalian cells (Ansar Ahmed et al., 1994). To ensure resazurin was nontoxic to yeast cells, ale yeast was grown in YPD medium, and 1 × 10 6 cells incubated with or without 100 µM resazurin. The cell viability was evaluated using the trypan blue exclusion method at 1 and 4 h postresazurin addition. As cell viability remained the same irrespective of the addition of resazurin, it demonstrated 100 µM resazurin was nontoxic to yeast cells up to 4 h (Supporting Information: Figure S1).
A rapid and inexpensive assay that determines yeast physiology is essential in a brewery setting, as corrective action can be performed before the pitching of yeast. For example, culture-based methods require incubation periods of 24-48 h. Therefore, sufficient resorufin production within a short incubation time is required for the assay. To ensure that resazurin was not a limiting factor, the dye was added at 10, 20, or 40 µM to yeast cells and resorufin production measured from 1 to 4 h (Supporting Information: Figure S2). At 10 and 20 µM resazurin, the rate of resorufin production began to slow within 3 h. A consistent reduction was observed for 4 h incubation with 40 µM resazurin, indicating that this concentration of resazurin was not a limiting factor and suitable for subsequent experiments.
As the fluorescence signal of resazurin with beer was most likely affected by pH (pH~4) and compounds in beer that causes the formation of resorufin, it was not suitable for the assessment of yeast metabolism during brewing processes. Further, as YPD may potentially alter the metabolism of the yeast obtained from brewery processes, we assessed the suitability of incubating cells obtained from laboratory cultures in PBS containing resazurin. Yeast cells were isolated at the logarithmic phase of growth and incubated with 40 µM resazurin containing PBS (Figure 3).
The fluorescence was higher as the number of cells increased, irrespective of the incubation time. This was expected as a higher number of cells F I G U R E 1 Experimental workflow for the evaluation of yeast metabolism during brewery processes. Yeast cells (1 × 10 6 ) were collected, counted, washed with PBS, and seeded into a 96-well flat-bottom plate in PBS containing 40 µM resazurin. Samples were incubated at room temperature for 1 h and fluorescence measured using a multimode microplate reader. Created with BioRender.com. PBS, phosphate-buffered saline.
F I G U R E 2 Effect of medium on resorufin production over incubation time. PBS, Beer, and YPD medium in the presence of 40 µM resazurin was incubated at room temperature for up to 4 h. Data shown as mean fluorescence of triplicate wells and error bars represent standard deviation. PBS, phosphate-buffered saline; YPD, yeast extract peptone dextrose.
should produce an increased amount of resorufin. This has been demonstrated with previous studies that have assessed the formation of resorufin from resazurin in mammalian cells (Gong et al., 2020). A linear increase of fluorescence was observed at cell densities of 1 × 10 6 to 3 × 10 6 from 0.5 to 4 h, indicating fluorescence formation can directly reflect cell vitality. Additionally, the variability of the fluorescence measurements was generally small (within a coefficient of variation of 20%) for all cell numbers and incubation times (Supporting Information: Figure S3). This showed that the measurements were consistent across the observed range of fluorescence values and cell numbers.
To demonstrate the reliability of the resazurin assay, we performed the BacTiter-Glo Microbial Cell Viability Assay in parallel. Yeast cells were isolated at the logarithmic phase of growth and resuspended in PBS before the addition of BacTiter-Glo reagent. A linear relationship between F I G U R E 3 Evaluation of cell number on resazurin reduction in PBS. Ale yeast was grown in YPD medium at room temperature. 0.125 × 10 6 to 3 × 10 6 yeast cells were obtained by centrifugation, washed with PBS, and resuspended with PBS containing 40 µM resazurin. Samples were incubated in the dark for (a) 0.5 h, (b) 1.5 h, (c) 2.5 h, (d) 3.5 h, and (e) 4 h prior to measurements. Data shown as the mean fluorescence of triplicate wells and error bars represent standard deviation. PBS, phosphate-buffered saline; YPD, yeast extract peptone dextrose. luminescence and cell density from 0.0625 × 10 6 to 1 × 10 6 was observed (Supporting Information: Figure S4). This was expected as it has been demonstrated that ATP content relates directly to the number of cells present in culture (Crouch et al., 1993).
To ensure consistency in subsequent experiments, we utilized a cell density of 1 × 10 6 , a resazurin concentration of 40 µM, and a 1 h incubation time as this enabled a rapid determination of cell vitality within the linear range. It is possible that the optimal resazurin concentration could be above 40 µM, but a rapid increase in fluorescence within a short incubation time was preferred to enable implementation in a brewery setting.

| Resazurin can indicate differences in metabolism of yeast at different stages of growth
Since the resazurin assay provides an indication of the reductive capacity of yeast, it may be possible to differentiate cellular growth phases during the brewery process. Typically, yeast cells grown in glucose-rich media in aerobic conditions have distinct growth phases including the lag and logarithmic phases, diauxic shift, and stationary phase (Werner-Washburne et al., 1993).
Yeast grown in YPD showed a lag phase from 0 to 3 h, followed by a logarithmic phase from 3 to 12 h, and then entered stationary phase (Figure 4a). The specific gravity (SG) corresponds to the density of dissolved sugars in the media. The sugars were consumed by the yeast during logarithmic phase, which correlated to a decrease in SG.
The SG did not change during stationary phase as the yeast was not in active growth.
Yeast cells (1 × 10 6 cells) were obtained at the lag (1 and 3 h), logarithmic (5.5 and 8 h), and stationary phases (24 h). Interestingly, the fluorescence of yeast cells in the early lag phase (1 h) was higher than all other time points (Figure 4b). However, the fluorescence observed during the late lag (3 h) and early logarithmic (5.5 h) phases could not be easily differentiated from each other, but both were significantly higher than late logarithmic phase (8 h), while the fluorescence of stationary phase (24 h) cells was lower than lag or logarithmic phase yeast cells. The decline in fluorescence over time was most likely due to a decrease in the redox potential of the yeast cells. It has been reported that the rapid decline of redox potential correlated to the faster yeast growth and reached a minimum at stationary phase (Lin et al., 2010).
The ATP content of the yeast cells was compared to resorufin production. A lower ATP content during early lag phase (1 h) was observed (Figure 4c). The ATP content increased at the late lag phase (3 h) and showed similar levels throughout the logarithmic growth phases (5.5 and 8 h) before decreasing at stationary phase (24 h). It has been shown that ATP content increases during cellular growth and rapidly decreases after glucose exhaustion (Sato et al., 2000). The different trends observed between resorufin fluorescence and ATP luminescence F I G U R E 4 Evaluation of the resazurin assay to differentiate between yeast cell growth phases. (a) Yeast growth was measured by monitoring the optical density at 600 nm (n = 4) and the specific gravity of YPD was measured (n = 2) up to 24 h. Data shown as mean value with error bars representing standard deviation. Yeast (1 × 10 6 cells) obtained from a laboratory culture at the lag, logarithmic, and stationary phases were resuspended in PBS containing either (b) 40 µM resazurin and incubated for 1 h or (c) ATP bioluminescence reagent and incubated for 5 min. Data shown as the mean fluorescence of at least n = 3 replicates with error bars representing standard deviation. Statistical analyses were performed using t-tests: *p ≤ 0.05, **p ≤ 0.005, ***p ≤ 0.0005. ATP, adenosine triphosphate; PBS, phosphate-buffered saline; YPD, yeast extract peptone dextrose. However, we have shown that resazurin was able to better differentiate between different stages of yeast growth.
3.5 | Application of the resazurin assay to evaluate an industrial-scale propagation process Next, we evaluated the applicability of the resazurin assay to monitor changes in yeast metabolism during a brewery propagation. Yeast samples were obtained from an industrial-scale propagator at regular intervals and changes to yeast metabolism were monitored using resazurin, storage carbohydrates, yeast cell counts, and methylene blue assays.
The viability of the yeast cells remained above 95% for the duration of propagation (Figure 5a). The formation of resorufin was consistent between the samples collected at 0 and 23.5 h (Figure 5b), but a decrease in fluorescence was observed in the 47.5 h samples collected at the end of propagation. These results indicate that metabolism decreased towards the end of propagation and suggests the yeast may have entered stationary phase. This was also observed for laboratory-grown yeast cells in stationary phase (Figure 4b).
It was expected that the yeast would be driven towards fermentative metabolism during propagation, despite being supplied with oxygen at regular intervals as the concentration of sugars in the wort were high (Moutsoglou & Dearden, 2020). Limiting factors in the wort such as nutrient deprivation (assimilable nitrogen or zinc ions), the production of ethanol and carbon dioxide, and a high cell density results in slower yeast growth and entry into the stationary phase (Hulse, 2003).
To gain a more detailed insight of yeast metabolism during propagation, we evaluated the cellular levels of the storage carbohydrates glycogen and trehalose. The monitoring of storage F I G U R E 6 Monitoring of different parameters during industrial-scale propagation. Ale brewing yeast cells (1 × 10 7 cells) were obtained from a propagator at different times before seeding for bottle refermentation. Extracted carbohydrates were digested with amyloglucosidase or trehalase to determine (a) glycogen or (b) trehalose concentrations, respectively. Data shown as the mean absorbance of triplicate wells and error bars represent standard deviation. Yeast obtained from the propagator were mixed with methylene blue and counted using a hemocytometer. Data shown as the mean count of replicate samples with error bars representing standard deviation. (c) Propagation samples were filtered through a 0.4 µm filter and degassed before the measurement of alcohol concentration (v/v,%), apparent extract (w/v,%), and pH. carbohydrates provides complementary insights to the resazurin assay regarding cellular growth and metabolism.
The concentration of glycogen was 23.4 ± 0.4 µg/mL at the start of propagation (Figure 6a). Glycogen began to accumulate immediately after it was transferred into the propagator and reached 67.7 ± 1.2 µg/mL after 23.5 h and continued to increase to 137.8 ± 1.8 µg/mL at the end of propagation (47.5 h). The immediate accumulation of glycogen during propagation was expected as the yeast was transferred into the propagation tank during logarithmic growth such that no lag phase was observed. The accumulation of glycogen during logarithmic growth has been shown to occur during fermentation on wort and is in response to the depletion of essential nutrients, such as glucose (Ouain et al., 1981).
The amount of glycogen present in the yeast before seeding for bottle conditioning is important as it has been reported that yeast with lower glycogen content will perform less well (Quain, 1988). Therefore, poor glycogen content could result in a slower refermentation rate and a longer lag phase during bottle conditioning. This is undesirable as it can lead to the presence of unwanted flavors (Derdelinckx et al., 1992;Hulse, 2003).
Trehalose levels were low at the start of propagation (0.54 ± 0.1 µg/ mL) and remained relatively low at 7.6 ± 0.04 µg/mL until 23.5 h ( Figure 6b). Nonetheless, the amount of trehalose rapidly increased from 13.7 ± 0.2 µg/mL at 29.5 h to 51.7 ± 1.1 µg/mL at 47.5 h. This accumulation of trehalose towards the end of propagation was expected as trehalose synthesis does not usually occur in logarithmic phase unless the yeast cells are exposed to stress (Eleutherio et al., 2015). As the amount of yeast cells in the propagator began to plateau between 29.5 and 47.5 h, along with the AE and alcohol content (Figure 6c), it suggested that trehalose synthesis was initiated by energy limitations and in preparation towards a different growth stage. This has been shown to occur in yeasts during the shift from logarithmic to stationary phase under wine-making conditions (Novo et al., 2005).
In summary, the lower metabolic activity observed with resazurin and the concomitant increases in glycogen and trehalose indicate that the yeast reached stationary phase after 47.5 h. As resazurin can detect changes in metabolic activity, it can provide breweries with a useful method to rapidly detect the onset of stationary phase before the degradation of glycogen and trehalose. This is important during propagation as selecting an optimal cropping time before storage carbohydrate degradation occurred would promote consistent bottle refermentation.
In this study, we demonstrated the application of resazurin to detect changes in the metabolic activity of yeast under laboratory and industrial settings. Although cell number, incubation time, resazurin concentration, and media type influence assay measurements, they can be successfully used for the assessment of cell vitality, provided appropriate control experiments are performed to determine their influence on fluorescence.
After evaluation of these parameters, a cell density, incubation time, and resazurin concentration showed a linear relationship with fluorescence production that allowed an evaluation of changes in yeast vitality within a rapid timeframe.
We successfully showed that a decrease in metabolic activity using resazurin preceded any discernable decrease in storage carbohydrates and viability during propagation. This is important as propagation time and the cropping of yeast for use in downstream production processes can directly impact product quality. This rapid method can be implemented by breweries to monitor changes in yeast metabolic activity in combination with other techniques to assess yeast quality.