Unlocking the Causes of Haze
While lagers and clear IPAs seem to be enjoying a resurgence in popularity, hazy IPAs continue to dominate beer menus at home and in the taproom. Hazy IPAs came to prominence around the time I co-founded Omega Yeast Labs in 2013. While the science wasn’t understood at the time, brewers found certain yeast strains more reliably produced the desired extreme, orange juice-like turbidity. Sales of our British V strain (known from other providers as London Ale III, London Fog, Juice, and Foggy London) grew steadily along with the growth in popularity of hazy IPA. Adjuncts like oats and wheat are also used to boost haze potential, but the common thread among the best examples was the yeast strain. What is it about the yeast strain that made it so good at creating this haze? Was it possible that certain strains had a “genetic predisposition” to creating haze?
From penicillin to the physics of microwave ovens, many scientific discoveries stem from a fortuitous observation not even connected to the goal of the immediate experiment. Our discovery of the “haze gene” in yeast is no different. Our R&D Director, Dr. Laura Burns, and research scientist, Keith Lacy, were conducting flask-based experiments to study hop creep based on dry hop timing. Hop creep refers to the release of dextrin-degrading enzymes (amylases) that break down long carbohydrate chains to release fermentable sugars.
The experimental setup was simple — wort with a specific gravity of 1.060 was prepared from 2-row base malt and distributed to flasks. Two popular strains were chosen to control for strain-dependent observations — West Coast Ale I (i.e., Chico) and British Ale V. Each flask was inoculated with yeast followed by a dry hop at the same time the yeast was pitched in flask one; day one into fermentation in flask two; day two in flask three; etc. through day seven. The goal of the experiment was to determine if earlier dry hop timing accelerated the drop in gravity caused by amylase enzymes in hops, thereby speeding up hop creep. While we were able to make interesting observations about dry hop timing and hop creep, the more exciting observation was related to haze in the final product. Visually, the flasks ranged from super bright to massively hazy. For British V, the later the dry hop, the hazier the beer. For West Coast Ale I, very little haze was seen in any flask.
This simple experimental setup to study hop creep became known as our “haze assay” and has been wielded to make dozens of observations relating to strain variation, dry hop timing, dry hop dose rate, and dry hop variety that will be discussed in this article. We used the assay to characterize our strain collection into haze-positive (like British V) and haze-neutral (like West Coast Ale I) strains.
The haze assay gave us an opportunity to conduct some genetic experiments to hopefully identify the genetic component of haze. Ultimately, we wanted to know the mechanism behind the haze-positive phenotype (trait). We generated a hybrid strain by crossing British V to a haze-neutral wine strain (yes, certain strains can undergo a sexual cycle to make new strains). To our surprise, the resulting hybrids were split between haze-positive and haze-neutral (in genetics terms 2:2 segregation). This pattern of inheritance suggested that the haze-positive phenotype was dominant and linked to a specific gene, and from there we set off on our quest to find the haze gene. This seems pretty straightforward, but it was way more complicated than we had anticipated.
We next took one of the haze-positive hybrids and “backcrossed” it with the original haze-neutral wine strain. And again, we tested the resulting hybrid yeasts for haze status in our haze assay. We repeated this six more times and by doing so, we “diluted out” the British V genes — except for the gene responsible for the haze trait. Finally, we compared the whole genome DNA sequences for the hazy hybrids from the last backcross to the original British V and wine strain genomes. By doing so, we were able to locate a gene from British V in the last backcross that wasn’t present in the wine strain or the non-hazy backcrosses.
Luckily, we have a full-time bioinformatics scientist on staff because this project went down a deep hole of complicated genetics. Using modern technology, including long-read sequencing data, led us to a gene with an unknown function, which we named HZY1. Excitingly, we saw extreme variation from strain to strain in the type of HZY1 gene that a strain had. The HZY1 gene seems to be prone to mutations that involve repetitive DNA resulting in versions (alleles) of the gene that are anywhere from 580 base pairs (bp) to >2,500 bp. Within any given strain, we found up to five different versions of HZY1! Consequently, it isn’t a straightforward answer for which version causes haze, but we found a strong correlation between longer versions of the HZY1 gene and the haze-positive phenotype. However, even our haze-neutral strains contribute to small degrees of haze (think the small amounts of haze in most dry-hopped beers). When the HZY1 gene is disrupted in every brewing strain we have tested, the resulting dry-hopped beers are significantly less hazy.
While uncovering the HZY1 gene was exciting, it didn’t answer the question: How does yeast cause haze in beer? We still don’t know for certain, and experiments continue. We do, however, have some good hints about how this gene could physically lead to haze from the DNA sequence itself. The HZY1 gene encodes a secreted glycoprotein (i.e., a protein with sugar molecules attached to it that is sent outside the cell) that is anchored to the cell surface; very similar to the FLO genes that encode the cell surface proteins that lead to flocculation.
The measurements of haze with different Omega yeast strains before and after the HZY1 gene is removed.
The HZY1 gene is heavily glycosylated and we can detect the HZY1 glycoprotein in the haze particles (~500 nm colloids) that form after dry hopping with haze-positive yeast. The particles are actually visible under the microscope! At this point, we can confidently say that stable haze shouldn’t rely on yeast cells in suspension, but yeast-derived HZY1 glycoprotein will do the trick. In other words, flocculation is completely unrelated to haze status. A common misconception has been that you need non-flocculent yeast to make hazy IPA. If you take one thing away from reading this, let it be that you don’t need non-flocculent yeast to make hazy IPA. There are flocculent yeast that make hazy IPA (British V is quite flocculent) and there are non-flocculent yeast that will not produce the haze brewers seek. While it’s true that non-flocculent yeast will make hazy beer, it’s not the stable, colloidal haze that is common in the best hazy IPA — eventually those yeast will drop to the bottom and the beer will be clear.
If you’re reading this and thinking to yourself, “I don’t make hazy beer and I don’t care about any of this,” our haze research isn’t all about making hazy beer. It has helped us understand a lot about avoiding haze, too. The brewing strains you use have a version of HZY1 that either leads to a small amount of haze or a large amount of haze. If you want to make a dry-hopped beer that is brilliantly bright, chances are good that getting rid of HZY1 in your house strain will make that easier. Knowing the haze status of your strain will get you closer to the outcome you desire for your beer. In other words, it is an uphill battle to make clear dry-hopped IPA with the British V strain. It is inherently prone to making hazy beer due to its genetic makeup.
We still have more to learn, but we ultimately hope to make the execution of hazy and non-hazy beer more of a science and to provide brewers with more tips, yeast options, and predictability when designing their IPAs.
Other Haze Factors
We have discovered many factors that impact haze by altering variables in our haze assay. The guidance provided by these observations can be used by the homebrewer to create haze in beer when they want it and avoid it when they don’t. After months of testing, several important factors influencing haze began to stand out:
1. Yeast Strain
When we looked at the rest of our strain collection, we found that British V was not unique and there were a handful of brewing strains that produced significant haze. In our flask fermentations, yeast that produced >200 NTUs (nephelometric turbidity units) with a late fermentation dry hop at 8 g/L (about 2 lbs./bbl or 5 oz. per 5 gallons) were termed “haze-positive” (blue in Chart 2). There is extremely useful information in this chart. You wouldn’t want to bang your head against the wall trying to make hazy IPA with Chico, for example. You might be able to get mildly hazy beer by utilizing adjuncts like oats or wheat, but you’re not going to get orange juice-level haze with Chico. Conversely, if you’re attempting a modern take on a crystal clear Kölsch-style beer by including a dry hop and Kolsch I (OYL-017) is your strain of choice, it is going to be exceedingly difficult to get it clear because Kolsch I is genetically prone to haze in combination with dry hopping. Don’t fight genetics!
2. Dry Hop Timing
Our experiments have shown that dry hopping close to the end or after fermentation is best for inducing yeast-derived haze. Late dry hop = more haze. The good thing about this is that it allows you to harvest yeast before dry hopping without fear of losing haze potential. Harvesting yeast after dry hopping can cause lots of viability problems.
Conversely, dry hopping in the first 24 hours leads to less haze and can even prevent haze from forming with later dry hop additions. This is one of the most interesting and confounding observations from our research. To this day, we don’t know the mechanism, but a very small dry hop (e.g., ½ oz. or less per 5 gallons, or 14 g/19L) at the same time as you pitch your yeast can help make your beer clear. This even works with haze-positive yeast! We’ve had professional brewery customers implement this technique in recipes where they have had difficulty achieving clear beer and it has helped tremendously. Try it the next time you’re making a clear lager.
3. Dry Hop Dose Rate
Testing dry hop rates of 1.3, 2.5, 5, 7.5, 10 oz./5 gallons (37, 71, 142, 204, 284 g/19 L), we found a linear correlation to the amount of haze formed: The heavier the dry hop, the more the haze. Keep in mind though, hazy is hazy. By eye you might have a hard time telling the difference between skim milk and whole milk — both are pretty opaque! There are diminishing returns in aroma and flavor quality with higher dry-hop loads and the same is true for haze.
4. Hop Variety
This part of the research is surprising and likely provides some clues into how dry hopping is leading to the yeast-dependent haze. What component of the hops is providing the signal or is part of the haze particle? We don’t currently know why, but different hop varieties lead to different amounts of haze with various combinations of haze-positive and haze-neutral yeast (see Chart 3). Some varieties of hops, like Enigma® and Sabro®, provide a decent amount of haze even when used with a haze-neutral yeast. Some varieties, like Galaxy®, provide huge amounts of haze when used with haze-positive yeast strains but provide very little haze when used with haze-neutral strains. Hop suppliers might be onto something with hop products that stabilize haze and we are keeping a close eye on their findings.
Conclusions
As history has shown again and again, you never know where you’re going to end up when you track down a scientific observation. A lesson I learned as a graduate student was to never conduct experiments with blinders on. You might be setting up an experiment to examine a particular problem, but if there are unexpected observations, you might be led somewhere far more fruitful if you follow the unexpected path.
As you can now appreciate, the opaque haze that we have come to love and expect in hazy IPA is largely a result of the particular type of HZY1 gene that a yeast strain possesses. That’s not the whole story, however. In tracking down the HZY1 gene, we made observations that are useful to brewers when formulating a new beer. One must consider the strain, the dry-hop timing, the dry-hop amount, and the hop variety when considering the amount of haze desired in the final product. All of these variables have a profound effect on the outcome.
We know a lot about how to produce stable haze, but does haze contribute to flavor, mouthfeel, or aroma? We’ve now conducted multiple experiments with breweries using a version of British V that has the HZY1 gene removed to make beer along with the traditional version of British V in parallel. While it’s simple to visually distinguish the beers, tasters have statistically been unable to distinguish them when the beers are served in opaque cups. This doesn’t necessarily diminish the importance of haze in beer because we do “drink with our eyes,” but it is amusing nonetheless that brewers spend so much time and energy making a beer clear and consumers have expectations associated with haze, but none of it matters strictly from a taste or aroma perspective.
Where do we go from here? Nature has provided us with a large amount of variation when it comes to dry hop-induced haze and the HZY1 gene, but what if we want to make hazy beer with Chico yeast? Now that we know the genetic basis of haze, it is possible to move a haze-inducing version of HZY1 into Chico for stable haze while maintaining all the traits we love about it, including minimal esters and good hop aroma expression. What if we want the fruitiness and soft mouthfeel of British V, but the clarity of a West Coast IPA? We can, and have, knocked out the HZY1 gene from British V in a new strain now available called DayBreak-V (OYL-408). This provides exciting opportunities to merge styles. Soft, yeast-derived fruity esters in a crystal clear dry hopped beer merges the traits of West Coast IPA and hazy IPA to create something fresh and exciting. Brewers, including those of you reading this, will probably discover new applications for these modified strains that we’re not even contemplating — and that’s why yeast scientists do what we do!
