Effects of yeast genetic diversity and microbial interactions on dough rise phenotype

This is the journal article that I wrote reflecting my work at the Kruglyak Lab this summer. All data plots were coded using R.

Effects of yeast genetic diversity and microbial interactions on dough rise phenotype

Erin Li, Chantle Swichkow, Leonid Kruglyak

Abstract

The process of fermentation spans over many centuries and across different cultures and usages. Yeast is a single-celled organism that is an essential component in the process of fermentation. However, there have been limited studies done on the cultivation of yeast in the form of microbial communities. Bread baking is an ancient process and starts in the form of a sourdough starter, which includes flour, water, and yeast. In natural fermentation practices, yeast is found in the presence of bacteria. In commercial bakery settings, yeast has been artificially selected to maximize volatile compounds and is often isolated so that it is found in the form of pure yeast.

In this study, we aim to investigate the microbial interactions between yeast and bacteria in the form of sourdough bread. We create dough-rise assays to compare the dough-rise ability of different strains of S. cerevisiae and how they are affected by the presence of bacteria. The presence of bacteria had variable effects on the ability of yeast to create dough rise. Upon further study, we discovered that dough rise in the form of a sourdough starter does not necessarily correlate to dough rise in the form of physical bread. These differences may be explained by the yeasts’ ability to ferment in different environments, such as the unique environment created in the oven. 

Introduction

Fermentation is an ancient process that has been used by many different cultures around the world. Yeast is a critical ingredient in fermented products. Yeast species S. cerevisiae has been domesticated for a wide variety of uses, including for different fermented drinks and food products (Bigey et al. 2021). To create sourdough bread, it is necessary to include environmental microbes– yeast and bacteria– in the dough. The origin of starters is categorized into commercial, referring to bread that is mass-produced in a professional kitchen, versus artisanal, which is often made by home bakers. Commercial starters consist of commercial yeast that has properties of faster dough-rise rates, while artisanal starters often contain wild yeast that are less predictable in rise-times. Starters can also include multiple strains of yeast that are a combination of commercial and artisanal strains (Bigey et al. 2021). Artisanal and commercial sourdough starters possess differences in environments, production practices, and ingredients. These differences are factors of interest to scientists because of their potential to impact sourdough community structure, which describes the relative abundance and coexistence patterns of different species of yeast and bacteria (Landis et al. 2021). The production of organic acids and extracellular enzymes by different strains of yeast also impacts bread flavor, texture, shelf-stability, and nutrition (Arendt et al. 2007, Du Vuyst et al. 2016, Gobbetti et al. 2014, Hansen and Schieberle 2005, Slim-ur-Rehman et al. 2006 as cited in Landis et al. 2021). 

Model yeast species S. cerevisiae has been domesticated for a wide variety of uses including for different fermented drinks and food products. Depending on the origin and usage of the yeast strain, it can be artificially selected for its particular usage. For example, commercial strains are likely to have faster dough rise rates to fit the needs of bakery environments while natural strains could be slower to rise but may have stronger aromas. 

Different strains of yeast have been selected to best fit its usage in different contexts. Domestication of yeasts for bread production may lead to a selection of yeast optimized for maltose catabolism, which is how yeast strains break down sugar, an ability necessary for fermentation. Through the process of fermentation, carbon dioxide gas is released, leading to trapped air between the dough and, ultimately, dough rise. 

Genetic variations such as differences in ploidy, single nucleotide polymorphisms, and copy number variations could affect a particular strain’s ability to produce carbon dioxide gas (Peter et al. 2018). Genetic variations can also affect a yeast strain’s production of amino acids. The specific amino acids as well as the amount produced by a yeast strain will impact the bacteria’s consumption of the amino acids. Amino acids can also affect the pH of the starter environment, affecting a bacteria’s survival as some species don’t have the ability to tolerate more acidic or basic environments (Ruiz et al. 2021). 

Since artisanal starters contain wild strains of yeast, it is often more likely that they will be less predictable since they haven’t been artificially selected as commercial strains have been. Therefore, wild strains of yeast are more likely to create unique environments, such as extreme pHs, which might contribute to the aroma of the sourdough starter. By studying a variety of strains from different ecological origins, we can compare the differences between artisanal and commercial strains in their dough rise behaviors.

Lactic acid bacteria are commonly found in fermented food and have been observed to have variability in individual starters. Lactic acid bacteria are known to play a role in the variation of volatile compounds, which leads to variations in dough rise rate and maximum dough rise. (Landis et al. 2021) Interactions between yeast and lactic acid bacteria can be due to genetic variation within individual yeast strains having certain mutations that cause the particular strain to have different microbial interactions. 

A certain strain’s interactions with bacteria in a starter environment can also contribute to dough rise. In past studies, it has been observed that yeasts, lactic acid bacteria, and acetic acid bacteria in the starter produce CO2, causing the dough to rise (Landis et al. 2021). In this research, we aimed to study the role of interaction between yeast and bacteria in dough rise by measuring the phenotype of dough rise in sourdough starters made with different strains. For this experiment, we studied the lactic acid bacteria species L. plantarum’s interactions with different strains of S. cerevisiae. L. plantarum was chosen for its perceived interactions with yeast in previous studies, particularly its dependence on yeast secretion of metabolites for survival (Pomonarova et al. 2017). We studied their ability to produce the dough rise phenotype, analyzing the differences both between starter cultures containing different strains of S. cerevisiae and starter cultures containing strains in combination with L. plantarum. 

From these experiments, we observed a variable effect of the presence of bacteria on the different strains of sourdough starters. It was noticed that some yeast strains rise better with the presence of bacteria, while in others, dough rise was decreased when bacteria were present. These findings could be due to the unique isolations of each strain, their traditional usages, and whether they are commonly found in the presence of bacteria.

Results 

To study different S. Cerevisiae strains and their interactions with L. plantarum, a species of lactic acid bacteria commonly found in food, we created dough rise assays. We created the dough rise assays using two sets of sourdough starters– one consisting of just yeast, and the other including yeast and bacteria. We used starters to study the variations between different strains of S. cerevisiae consisting of flour, water, and yeast strain (Figure 1). To study the microbial interactions and effects of lactic acid bacteria, we created starters consisting of flour, water, the yeast strain, and L. plantarum (Figure 2). We analyzed microbial interactions between S. Cerevisiae strains and L. plantarum by observing the differences in dough rise between different strain and bacteria combinations. From the dough rise assays, we observed that there were particular strains that seemed to have more apparent effects from the presence of L. plantarum than others.  

Figure 1: Variation in dough rise exists between strains

We repeated the dough rise assays 8 times and measured the rise after 24 hours. The y-axis represents the dough rise after 24 hours and is measured in centimeters. The center line on the box plot indicates the mean, the box shows the first and third quartiles of the dough rise, and the dots represent the outliers. Dough rise was measured for control experiments of sourdough starters with just the yeast strain. Different strains with various ecological origins and usages have different effects on dough rise.

Figure 2: Strains interact with bacteria differently

Dough rise was measured for treatment experiments of sourdough starters with both yeast strain and bacteria. Some strains have the ability to coexist with bacteria and even have mutualistic relationships, while others are negatively affected. 

Figure 3: Boxplot of control and with bacteria treatment starters

Dough rise was measured after 25 hours and recorded for all starters. There are variable effects of the presence of bacteria on different strains. For some strains, the presence of bacteria caused the dough rise to decrease, but in others, the dough rise increased. 

Baking Assays

To further study the microbial interactions and dough rise mechanisms behind three specific strains– CBF, BTC, and BMR– we conducted baking assays to see how the differences from the dough rise assays would manifest in the form of sourdough bread. We created starters with and without bacteria for each strain and baked them in the oven to observe how the process of creating actual sourdough bread might differ from the results we observed from the dough rise assays. In the form of sourdough bread, the differences in dough rise varied from our observations in the dough rise assays. CBF with bacteria treatment significantly had less dough rise in bread but did not have as much observable difference in the form of sourdough starter. BMR matched sourdough starter observations and wasn’t particularly affected by the bacteria treatment. BTC also differed from the sourdough starter experiment as it had significantly negative effects from bacteria treatment, but, in the form of bread, with and without bacteria treatment appear to be approximately similar (Figure 4). 

Figure 4: Differences in microbial interactions between yeast and bacteria as observed in the form of sourdough bread

We conducted baking assays 4 times. When the starters were baked into sourdough bread, the differences in dough rise between treatments varied. This shows that even if a strain seems to be significantly affected by bacteria treatment, it may not result in physical differences in actual bread.

Discussion

Overview of dough rise assays

In this study, we aimed to investigate how microbial interactions affected the ability for yeast strains to create dough rise. To observe how the presence of bacteria, specifically L. plantarum,  would affect a yeast strain’s ability to create dough rise, we conducted two sets of dough-rise assays using 30 strains of S. cerevisiae– one with just the yeast strain, and the other with both yeast and bacteria. In the starters with only the yeast strain, we aimed to visualize the variability in dough rise between strains of S. cerevisiae. In the starters including both yeast and bacteria, we aimed to observe how the strain’s ability to create dough rise would be affected by the presence of bacteria.

From the results of the dough-rise assays, we observed variability in the effect of bacterial presence on the dough rise. For some strains, the presence of bacteria increased the amount of dough rise. Whereas for other strains, the dough rise decreased when bacteria was added. 

To further investigate how a strain’s isolation, ecological origins, and geographical origins might influence the microbial interactions in the form of a sourdough starter, we narrowed the selection of 30 strains to 3 strains, each with unique interactions with the presence of bacteria. The different environments yeast strains are found in has selection pressures the strain needs to adapt to. For example, for commercial strains, the environment has required the strain to more efficiently break down maltose, which leads to more dough rise (Lahue et al. 2020). Our selection of 3 strains included CBF, a bakery levain strain, BMR, which is isolated from chickens, and BTC, the most common commercial bakery strain. In our observations of the dough rise assays, CBF created the least dough rise overall, both on its own and in the presence of bacteria. BMR produced the same level of dough rise in both sets of the dough rise assays, leading us to believe that it was not affected by the presence of bacteria. BTC created the most dough rise overall; however, when the presence of bacteria was added, the dough rise seemed to decrease significantly. 

Overview of baking assays

Next, we aimed to see how the differences in dough rise between these three specific strains would manifest in the form of physical sourdough bread. To compare the strain’s ability to create dough rise in the form of sourdough bread, we conducted baking assays, where we created starters with and without bacteria for each strain and baked them in the oven to observe how the process of creating actual sourdough bread might differ from the results we observed from the dough rise assays. In the form of sourdough bread, the differences in dough rise varied from our observations in the dough rise assays.

For CBF, the lowest-rising strain in the form of the sourdough starter, we observed a significant increase in dough rise when it was baked. BMR did not seem to be affected much in the presence of bacteria during the dough rise assays, and we noticed similar observations during the baking assays. BTC had the most observed dough rise in the control treatment, but in the presence of bacteria, the dough rise decreased significantly during the dough rise assays. In the baking assays, however, the bacteria treatment did not seem to have a significant effect on the dough rise.

Baking properties and microbial interactions affect dough rise in sourdough bread

The differences in dough rise from the observations made during the dough rise assays compared to the baking assays might be explained by the specific settings and bread-baking environments. When conducting our baking assay experiment, we baked the bread using the steam setting for the first 10 minutes, which might have affected the strains’ ability to create dough rise. In past studies, it has been observed that loaf volume is significantly affected by the early stages of baking, such as temperature and humidity (Baker et al. 1939 as cited in Fan et al. 1999). In this experiment, we had not expected such a strong impact from the baking conditions. From this, we were able to learn that dough rise studies are vastly different from and can not necessarily predict dough rise in physical bread.

To bake the bread, we set the temperature to 425 ℉, which could potentially lead to a rupture in the dough and release of gas. The high temperature would affect our baking results in that it would cause the dough to gain gas bubbles and, therefore, have an increase in dough rise. High temperatures can lead a closed cell space, which is usually dense and includes the tough texture of the gluten, to become an open cell sponge. Open cell sponges are more light and airy, which can lead to a sharp release of gas and, therefore, gain in dough rise (Gan, Ellis & Schofield, 1995 as cited in Fan et al. 1999). To create a more accurate experimental design, it will be necessary to grow the cultures under the same conditions by placing them in an incubator and increasing the humidity. By creating a more standard experimental environment, it is expected that results should be standardized between baking and dough rise assays.

The way a certain microbial community interacts with changes in temperature could also be due to genetic variation and the environment the strain has been selected to survive in.  It is common that commercial strains have been selected to produce more dough rise and begin the process of fermentation more quickly, whereas artisanal bakery strains have larger cultures with more population in the microbial community (Lahue et al. 2020). By baking the bread, it can be possible to study the viability of different strains in the presence of heat. In future experiments, more in-depth studies will be conducted on yeasts’ ability to create dough rise in different environments. In a broader biological context, the findings of these experiments can be used to determine which yeast strains possess the ability to create dough rise in different environments. 

Methods

Dough Rise Assays

Flour

We used a 1:1 Ratio of King Arthur- Unbleached All-Purpose Flour 100% Unbleached Hard Red Wheat and Bob’s Red Mill- 100% Stone Ground Whole Wheat Flour Whole Grain Hard Red Wheat to create the flour mixture used for the dough rise assays. We autoclaved the flour using the gravity cycle for 20 minutes. The dough master mix comprised a 4:3 ratio of autoclaved flour and sterile MilliQ water. 

Overnight Cultures

Bacterial cultures were grown in MRS Broth. MRS broth consisted of 51 g MRS (Sigma-Aldrich) and 1 mL Tween 80 (Sigma-Aldrich) dissolved in 1 liter of distilled water. The solution was autoclaved at 121oC for 15 minutes, then incubated for 3 days at 35oC. Bacterial cultures were made by adding frozen glycerol stock of L. planetarium using a wooden dowel to 14 mL of MRS broth, then letting the culture incubate overnight at 30oC. 

Yeast cultures were grown in Sourdough Media (SDM). Sourdough Media consisted of 24 g Wheat peptone (Sigma-Aldrich), 4 g Di-potassium hydrogen phosphate (Sigma-Aldrich), 4 g Potassium dihydrogen phosphate (Sigma-Aldrich), 0.2 g Magnesium sulfate heptahydrate (Sigma-Aldrich), 0.05g Manganese sulfate (Sigma-Aldrich), 1mL Tween 80 (Sigma-Aldrich), Citric acid monohydrate (Sigma-Aldrich) (added until pH reached 4.5), 15g Glucose (Sigma-Aldrich), 35 g Maltose (Sigma-Aldrich), 0.2 mg Thiamine (Sigma-Aldrich), 0.2 mg Folic acid (Sigma-Aldrich), 0.2 g Nicotinamide (Sigma-Aldrich), 0.2 mg Pantothenic acid (Sigma-Aldrich), 0.2 mg pyridoxal-phosphate (Sigma-Aldrich), 0.2 mg cyanocobalamin (Sigma-Aldrich). The solution was autoclaved on a gravity cycle for 15 minutes. Then, 375 mL of sterile water to bring the solution to 1 L. Yeast cultures were made by adding a single colony of a particular yeast strain into 2 mL of SDM using a wooden dowel. The cultures were placed on a shaking incubator at 30oC overnight. 

We created 1:20 dilutions of the cultures and autoclaved Milli-Q water to be read in the spectrophotometer. Using the ODs from the reading, we calculated the amount of yeast culture and water necessary to reach 0.1 OD in 500 mL of solution. The remaining volume of the 500 mL was filled with autoclaved Milli-Q water. We also created samples that included both yeast and bacteria. In these samples, we added the amount of yeast to have an OD of 0.1 and the amount of bacteria to have an OD of 0.1. The remaining volume of the 500 mL was filled with autoclaved Milli-Q water. 

Once the yeast and yeast+bacteria solutions were created, we combined them with the dough master mix. Each sample was created using 3.5 g of the dough mixture and 500 uL of the yeast or yeast+bacteria solution. The mixture was combined using a wooden dowel and then spun down in the centrifuge at 1400 rpm for 30 seconds. The samples were then incubated at room temperature (25oC) overnight and measured at 21 and 25 hours. 

Baking Assays

Three strains (CBF, BMR, and BTC) were used for baking experiments. An overnight culture of each strain and the bacteria (LPI) was made. The samples were diluted (1:20)  using sterile water and then placed into the spectrophotometer. After the OD was read, the amount of each strain necessary was calculated by taking all 2mL of the strain with the lowest OD. The other strains were calculated to match the OD of the lowest strain. The calculated amount was transferred into microcentrifuge tubes, which were spun down in the centrifuge for 5 min at 500 g. The supernatant was removed so that only the pellet remained.

Each yeast strain was resuspended in 1.1 mL of sterile water, and the bacteria was resuspended in 1.8 mL of water. A 1:1 ratio of flour (360 g)  and water (360 g) was used to create the dough mixture. The dough was mixed by hand until gluten could be observed and the dough could be stretched. The dough was separated into 6 bowls with 120 g of dough in each bowl. Each bowl received 500 uL of a particular yeast strain or 500 uL of yeast and 500 uL of bacteria. The bacteria and yeast were mixed into the dough using a plastic knife. The dough was placed into molds (9.17 x 8.42 x 0.9 inches) and left to rise overnight at room temperature. The dough was then baked using an ANOVA oven at 425oF. They were baked using 100% steam for the first 10 min, then steam was turned off, and they baked for another 6 minutes.

Data Analysis 

PRISM 10 was used to model the data. 

R studio and ggplot2 were used to create data plots.

Works Cited

Albasir MOS, Alyassin M, Campbell GM. 2022. Development of Bread Dough by Sheeting: Effects of Sheeting Regime, Bran Level and Bran Particle Size. Foods. 11(15):2300. doi:10.3390/foods11152300. [accessed 2023 Nov 13]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9367850/.

Bigey F, Segond D, Friedrich A, Guezenec S, Bourgais A, Huyghe L, Agier N, Nidelet T, Sicard D. 2021. Evidence for Two Main Domestication Trajectories in Saccharomyces cerevisiae Linked to Distinct Bread-Making Processes. Current Biology. 31(4):722-732.e5. doi:10.1016/j.cub.2020.11.016. [accessed 2023 Nov 13]. https://linkinghub.elsevier.com/retrieve/pii/S0960982220316912.

Ehrenreich IM, Gerke JP, Kruglyak L. 2009. Genetic dissection of complex traits in yeast: insights from studies of gene expression and other phenotypes in the BYxRM cross. Cold Spring Harb Symp Quant Biol. 74:145–153. doi:10.1101/sqb.2009.74.013.

Fan J, Mitchell JR, Blanshard JMV. 1999. A model for the oven rise of dough during baking. Journal of Food Engineering. 41(2):69–77. doi:10.1016/S0260-8774(99)00070-9. [accessed 2023 Nov 13]. https://www.sciencedirect.com/science/article/pii/S0260877499000709.

Landis EA, Oliverio AM, McKenney EA, Nichols LM, Kfoury N, Biango-Daniels M, Shell LK, Madden AA, Shapiro L, Sakunala S, et al. 2021. The diversity and function of sourdough starter microbiomes. Weigel D, Mitri S, Adams R, editors. eLife. 10:e61644. doi:10.7554/eLife.61644. [accessed 2023 Nov 13]. https://doi.org/10.7554/eLife.61644.

Lahue C, Madden AA, Dunn RR, Smukowski Heil C. 2020. History and Domestication of Saccharomyces cerevisiae in Bread Baking. Front Genet. 11:584718. doi:10.3389/fgene.2020.584718. [accessed 2023 Nov 17]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7686800/.

Ponomarova O, Gabrielli N, Sévin DC, Mülleder M, Zirngibl K, Bulyha K, Andrejev S, Kafkia E, Typas A, Sauer U, et al. 2017. Yeast Creates a Niche for Symbiotic Lactic Acid Bacteria through Nitrogen Overflow. Cell Syst. 5(4):345-357.e6. doi:10.1016/j.cels.2017.09.002.

Punia Bangar S, Whiteside WS, Singh A, Özogul F, Gupta A, Gahlawat SK. 2022. Properties, preparation methods, and application of sour starches in the food. Trends in Food Science & Technology. 121:44–58. doi:10.1016/j.tifs.2022.01.029. [accessed 2023 Nov 13]. https://www.sciencedirect.com/science/article/pii/S0924224422000371.

Ruiz J, de Celis M, Diaz-Colunga J, Vila JC, Benitez-Dominguez B, Vicente J, Santos A, Sanchez A, Belda I. 2023. Predictability of the community-function landscape in wine yeast ecosystems. Molecular Systems Biology. 19(9):e11613. doi:10.15252/msb.202311613. [accessed 2023 Nov 13]. https://www.embopress.org/doi/full/10.15252/msb.202311613.