Sunburst has been engineered to produce elevated levels of three different fruity ester molecules: ethyl-hexanoate, ethyl-octanoate, and ethyl-decanoate. When combined in just the right amounts, these three esters synergize to create a complex fruity flavor profile that is strongly pineapple, with secondary notes of cherry, citrus, and stone fruit. On this page, we’ll summarize the years of research that went into developing Sunburst, and we’ll give you a solid understanding of how this strain creates such strong pineapple flavors during fermentation.
Our starting point: What are the key flavor molecules in a pineapple?
To create Sunburst, the first question we had to answer was simple: What flavor molecules make a pineapple taste like a pineapple? To answer this question, we conducted sensory experiments in which we spiked beer with a variety of flavor molecules commonly found in pineapples. Through these experiments we found that a combination of several different ethyl esters – ethyl hexanoate, ethyl octanoate, and ethyl decanoate – make up the majority of the flavor of a pineapple. When combined in just the right proportions, these three molecules interact to create a tropical fruit medley with ripe pineapple as the primary sensory note.
Figure 2. Pineapple flavor is largely a combination of ethyl esters.
The knowledge that these ethyl esters are major drivers of pineapple flavor was a key starting point for out engineering work. To create a pineapple brewing strain, we would need to engineer a strain that produced high levels and specific amounts of these pineapple flavor molecules: ethyl hexanoate, ethyl octanoate, and ethyl decanoate.
Our engineering strategy
It turns out that during beer fermentation, most yeast strains already produce small amounts of these pineapple ethyl esters. With ordinary strains however, the levels of these esters that are produced are so low that they rarely have a significant impact on the taste of beer. In order to develop a strain that created strong pineapple flavors, we knew that we would need to substantially increase the ability of yeast to produce these pineapple flavor molecules. Our strategy for accomplishing this was straightforward: We would use the same biosynthesis pathway that yeast already uses to make small amounts of ethyl esters, but we would modify the enzymes involved in the pathway so that far more of the pineapple ethyl esters would be produced. The ethyl ester biosynthesis pathway that we targeted is shown in the figure below.
Figure 3. Red arrows show the two key enzymatic reactions leading to the biosynthesis of pineapple flavored ethyl esters in yeast.
As shown in the figure above, yeast production of pineapple ethyl esters requires two enzymatic reactions that branch off from the main fermentation pathway. In the first reaction, chemical intermediates of the fermentation pathway are converted into acyl-CoA precursors. This reaction is catalyzed by an enzyme called Fatty Acid Synthetase, or FAS. In the second reaction, acyl-CoA precursors are converted into ethyl esters by enzymes called Alcohol Acyl-transferases, or AATs.
The reason that yeast strains normally produce such low levels of pineapple ethyl esters is that their FAS and AAT enzymes are not very good at catalyzing these ester-forming reactions: The FAS enzyme produces only small amounts of acyl-CoA precursors, and yeast AAT enzymes do a poor job of converting these precursors into ethyl esters. In the following sections, we’ll describe how we modified a brewing yeast to dramatically improve its ability to catalyze these reactions, with the result being far greater production of pineapple ethyl ester molecules.
Initial engineering work: Our discovery of better AAT enzymes
We began our strain engineering by focusing on the second reaction in the yeast ethyl ester biosynthesis pathway: the conversion of acyl-CoA precursors into ethyl esters. This reaction is catalyzed by one or more of the AAT enzymes that yeast is known to express. Our first attempt to improve efficiency of this pathway involved increasing the expression level of these yeast AATs. We hypothesized that “over-expressing” these AAT enzymes might result in more acyl-CoA precursors being converted into ethyl esters. To test this, we created several engineered strains that over-expressed the known yeast AAT enzymes: EEB1, EHT1, or ATF1. We next performed small-scale fermentations with these strains and measured their total ethyl ester production relative to the un-modified parent strain. As shown in the left panel of Figure 4 below, over-expression of these yeast AATs had only a small positive effect on total ethyl ester production. Even over-expression of the most active yeast AAT, EEB1, resulted in only a 2-fold increase in ethyl ester production – not nearly enough to create meaningful fruity or pineapple flavors.
Figure 4. Total ethyl ester production by strains engineered for high-level expression of the indicated AAT enzymes. Ethyl ester production levels are shown relative to the non-engineered parent strain.
These data indicated that yeast AAT enzymes are not active enough to biosynthesize high concentrations of pineapple ethyl esters. Many other species aside from yeast also have AAT enzymes however, and we hypothesized that some of these non-yeast AATs might be more efficient than the yeast ones. To test this, we searched through enzyme databases and collected a set of nine different AAT enzymes from a variety of species like peach, tomato, and several non-yeast fungi. We collected the DNA sequences (genes) encoding these enzymes, expressed these genes in a brewing yeast strain, then measured the total ethyl esters that each AAT-expressing strain produced. As shown in the right panel of Figure 4 above, several of these non-yeast AATs turned out to be highly active, and they produced up to 6-fold more ethyl ester than the parent strain! This discovery was a major step towards developing a pineapple flavor strain. We chose the best of these enzymes, BY-2, to use as the AAT in our later strain development research.
Engineering work continued: Increasing precursor levels
After discovery of the highly active AAT enzyme BY-2 , our next step was to increase the levels the acyl-CoA precursors that are produced as a by-product of yeast fermentation. We hypothesized that if we paired BY-2 expression with increased acyl-CoA precursor levels, the result would be a large increase in production of pineapple ethyl esters. To accomplish this, we made two mutations to the yeast FAS enzyme that catalyzes formation of acyl-CoA precursors. We then created a strain that paired expression of this modified FAS enzyme with expression of BY-2. Finally, we set up fermentations to assess the production of ethyl esters by this new strain. The figure below shows the levels of all three pineapple ethyl esters molecules that we measured after these fermentations.
Figure 5. Production of ethyl hexanoate, ethyl octanoate, and ethyl decanoate by the parent and engineered strains. Ethyl ester production levels are shown relative to the parent strain.
Combining BY-2 expression with increased precursor levels resulted in a major boost in ethyl ester production compared to the strain expressing BY-2 alone. This was most notable for ethyl hexanoate, which was increased 3-fold over the strain expressing BY-2 alone, and >20-fold over the parent strain! This was very encouraging, as we knew that in pineapples, ethyl hexanoate plays a more prominant role in the flavor profile than ethyl octanoate or ethyl decanoate.
As we expected, beer brewed with this enhanced ethyl ester producing strain had strong aroma notes of pineapple, along with secondary notes of citrus and stone fruit (Figure 6, left panel). Unfortunately, the tasting notes from this beer were not uniformly positive. In addition to these fruity flavors, our tasters also described a bitter off-flavor that detracted from the overall quality of the beer. Subsequent metabolite profiling of the beer revealed the likely culprit of this off-flavor. After analyzing the flavor molecules found in the beer, we found that the concentration of fatty acids were 3-fold higher in beer brewed with the enginered strain compared to beer brewed with the parent strain (Figure 6, right panel). As fatty acids are known to impart bitter flavor notes, these were very likely responsible for the off- flavors described by our tasting panel.
Figure 6. Left panel: Sensory profiles of beer brewed with the engineered strain and the non-engineered parent strain. The engineered strain expresses BY-2 and was modified for increased acyl-CoA precursor levels. Right panel: Fatty acid levels in beer brewed by the engineered and parent strains.
Final engineering work: Removing off-flavors
Further engineering work to eliminate these off-flavors required a substantial effort from our strain development team. First, we moved expression of BY-2 and the modified FAS enzyme into a Chico brewing strain that produced lower levels of fatty acids compared to our original engineered strain. Next, we reduced the levels of fatty acids that this strain produced even further by identifying and reducing expression levels of the enzymes responsible for fatty acid formation. Together, these efforts resulted in a pineapple flavor strain that produced similar levels of fatty acids to it’s non-engineered parent, and that produced beer with no bitter off-flavors (Figure 7). Finally, we re-ran our analysis of flavor molecules produced by this strain during to confirm that no additional off-flavors or unwanted by-products were being created.
Figure 7. Left panel: Levels of pineapple ethyl esters and fatty acids in beer brewed with Sunburst or the non-engineered parent strain. Levels of all flavor molecules are shown relative to the parent strain. Right panel: Sensory profiles of beer brewed with Sunburst or the parent strain.
