No Ifs or Bots. How Alcohol, 48,000 plates, and Stressed Forks Served Up Increased Mutation Rates
A Flame-Side Chat, with Dr Karin Voordeckers
Recently, I described an article published by the group of Prof Kevin Verstrepen. Their work sought to investigate the molecular events that lead to enhanced mutation rates in the presence of Alcohol. I am pleased to report on a Flame-Side Chat with Karin Voordeckers, one of the team members who led this work.
Besides her work addressing the biology of yeast and cancer, Karin and her colleagues organise a (free) online course entitled “Beer: The Science of Brewing”, due to start in May. You will read about it in the last part of our chat. Mark your calendar for this (I have) and let’s get on with this chat!
Edgar: Work leading to this current publication, published in PLoS Genetics, sought to evolve strains that become more tolerant to ethanol. In this paper, the narrative centres around cancer biology. What prompted you to shift your focus from the biotechnological aspects of beer brewing to cancer biology?
Karin: The goal of our study that was published in PLoS Genetics was twofold:
Firstly, we wanted to see how repeatable evolution was. If we independently grow (genetically identical cells) do they become ethanol tolerant equally fast and through the same mechanism? If we play the tape of evolution multiple times, do we each time get the same result?
The second goal was to find out which changes in the DNA were responsible for the increased ethanol tolerance. Funnily enough, we don’t really know what makes ethanol kill cells, and also not how cells can become more tolerant. In addition to answering these questions, we figured we may also learn how we could engineer cells to become more resistant to alcohol, which would be a gamechanger for bioethanol production.
When we sequenced the DNA of our evolved strains to answer these questions, we realized that these strains had many more mutations in their genome than what would be expected. This, of course, made our work more difficult – if you only have let’s say 5 mutations in a genome, it is easier to identify which ones are really responsible for your trait of interest (in our case ethanol tolerance); and which ones are not so important. So, you can imagine we cursed a lot when we saw how many mutations our strains had!
For our first study (in PLoS Genetics), we really tried to identify the mutations responsible for the increased ethanol tolerance, and sort of decided to not try to investigate WHY they had so many mutations.
But of course, we are scientists and we don’t like unresolved questions ☺, so after we had identified the causative mutations for ethanol tolerance, we decided to the tackle the next question and investigate why there were so many mutations in the cells that had been exposed to ethanol for prolonged periods of time.
Edgar: You have found that alcohol exposure raises mutation rates. How does this rate differ from mutation rates induced by other treatments, traditionally used for mutagenesis?
Karin: Well, a lot of course depends on the dose (amount of mutagen, alcohol in this case) you use; how long the exposure is maintained, and the specific yeast used. In yeast, some commonly used mutagens include UV, EMS and MMS. Exposure to MMS (which also causes replication stress btw) can in some cases lead to an almost 50-fold increase in mutation rate – so much higher than what we see with ethanol.
Edgar: As alcohol is a lot safer to work with at more rudimentary setups, one reader wondered whether we should be isolating wild yeast in 6% + alcohol to both try and drive some mutations and select strains suited for brewing more quickly?
Karin: The key thing to remember is that while ethanol increases mutation rate, mutations still occur randomly across the genome, and they occur gradually over time. So, short exposure to ethanol will in most cases not result in mutations that make them more ethanol tolerant or more suited for brewing. Evolution takes time, and you need a sufficient number of organisms in the population to get a few that eventually get better. Ethanol creates random mutations and then selection acts on these mutations. Specific mutations that give your cells an advantage (increased fitness) will allow cells to grow better/faster in the environment they are in. This is exactly what happened in our PLoS Genetics paper, where we managed to evolve strains that were more ethanol tolerant (since we grew large populations of cells for several years in gradually increasing levels of ethanol; only cells with mutations that made them more ethanol tolerant could grow/survive and were hence selected for).
So, if you would like to select for mutations that make a strain more suited for brewing, you would need to keep growing it in an environment that mimics brewing conditions. And, big caveat, the brewing phenotype you want, needs to be something you can easily select for – which cannot always be done in a high-throughput way. For example, let’s say that you want to create a brewing strain that produces more of a specific flavor compound, such as isoamyl acetate (banana flavor). This is not something that can be done easily by simply growing your cells for a long time in the brewing conditions you want to use – since mutations that result in cells that produce more of this banana flavor do not increase the cell’s fitness.
Edgar: Can you think of phenotypes researchers & brewers might use with mutagenesis to improve traits other than ethanol tolerance?
Karin: Well, in principle you should be able to improve any phenotype you are interested in using mutagenesis – but it greatly helps if you have a phenotype that you can easily select for. Take the example of banana flavour I mentioned above. In principle it is possible to get a yeast strain that produces more banana flavor by using mutagenesis – but to identify that yeast would involve setting up may fermentations with all the different mutants you created; and then smelling all of them to identify the good ones. Problem is that if is often looking for one good mutant in a pool of millions that did not change, or even got worse.
If you are interested in phenotypes that can be (relatively) easily screened for; this includes growth at different temperatures, the ability of yeast cells to consume different sugars, and resistance to toxic chemicals like ethanol.
Edgar: Your work invokes a mechanism in which cellular stress (protein folding), translates to replication stress. To what extent do you think this pathway is specific to alcohol? Would other triggers of (proteotoxic stress and) replicative stress lead to the same outcome (enhanced mutation rates)? If so, would this open up new routes for intentional mutagenesis and/or a better understanding of (the onset of) cancer?
Karin: This is definitely not specific to ethanol, there are many studies out there that have shown that causing proteotoxic stress/replication stress also affect mutation rates. In fact, there are several studies showing that replication stress causes genome instability and ultimately cancer in higher eukaryotes.
One of the current ‘hot’ topics in the cancer field is to indeed try to better understand the onset and specific causes of cancer – and this can, for example, be done by looking at so-called mutational signatures. (sequence and context of specific mutations in cancer) This can help elucidate the cause of cancer. For example, there are specific mutational signatures that have been linked to UV exposure (in skin cancer), smoking (in lung cancer), and alcohol (in oesophagal cancers). In other words, by looking at the type of mutations found in cancer cells, researchers can now start to unravel what the causes of these mutations (and sometimes also cancer) were.
Edgar: There will be lots of brewers and biologists that look at your findings and suppose that your observations are (not) specific to yeast. What is your view on either mindset?
Karin: We have collaborated with a lab looking at the effects of ethanol on E.coli (a bacterium) (eLife, 6:e22939.) and they also see that ethanol increases mutation rates. (although the mechanism could be a bit different here).
The fact that ethanol causes proteotoxic and replication stress, and that these stresses have been shown to also cause an increased mutation rate in higher eukaryotes indicates that our observations are not just yeast-specific. The cellular machinery affected by ethanol in yeast are also present in mammals and humans.
Edgar: Your work implicates error-prone polymerases in alcohol-related mutations in S. cerevisiae. To what extent are these polymerases conserved amongst Eukaryotes? Are these part of an expanded family in some organisms and not others?
Karin: Translesion polymerases (sometimes also called error-prone polymerases) can be divided into different subfamilies and are conserved amongst eukaryotes. In fact, TLS polymerases are found in organisms throughout all three domains of life. Most TLS polymerases are members of the Y family of DNA polymerases (199), a unique class of DNA polymerases with specialized structures optimized to allow replication on damaged DNA substrates and, in some cases, to promote mutagenic DNA synthesis.
In our study, we looked at Rev1, Rev3 and Rad 30. Rev1 and Rad30 belong to the Y family of TLS polymerases, and Rev3 to one of the other families. Interestingly, and related to your previous question, a recent study reported error-prone polymerase-associated mutational spectra in alcohol-related tumours 17. In the latter case, tumour samples displayed a mutational spectrum characteristic for PolH (encoded by RAD30 in S. cerevisiae), whereas our data implicate PolZ as the primary source of alcohol-related mutations in S. cerevisiae.
Edgar: What were the major bottlenecks in your research? Were there any aspects/observations you would have liked to look into a little more?
Karin: The major bottleneck was how cumbersome the experiments were to determine mutation rates. This experiment involves a lot of manual labour – for example, the data shown in some of the ‘simple’ bar graphs of the paper involved pouring > 1500 agar plates (per bar graph – just count how many bar graphs are in the paper ☺), plating cultures on them, and then counting the colonies on all these agar plates!
Our lab has recently received funding to buy a robot that could automate most of these steps – the team would have been very grateful if that robot would have already been there for our experiments! ☺
Edgar: What is the next important, open question that you are now addressing?
Karin: We now have PhD student, shared with a lab at the university hospital, who is looking more into detail into the effects of alcohol on mammalian cells. So that is very exciting! I must say though that working with mammalian cells made us realize even more how easy it is to work with yeast cells! Mammalian cell lines, for example, grow much slower; so every experiment takes much longer before you have results. This becomes especially problematic for these experiments, since the mutagenic effect of alcohol really only becomes visible for extended exposure… And while we do believe that some of our results translate to the effect of alcohol on human cells, all of us are still enjoying a beer now and again. It’s really all about moderation!
Edgar: I am sure that my readers would like to learn more about your online course. Given your experience in practising brew science and teaching its principles, how do you think brewers and homebrewers will benefit from the course and understanding the science of brewing beer?
Our online course is aimed at students and professional brewers who are interested in getting an introduction to the science underlying beer brewing. The course will explain HOW and WHY all ingredients and steps of the brewing process lead to the different beers we know today. At the end of the course, people will understand the basics of the beer brewing process; and why different beers have different flavors and aromas. Students will learn how the key ingredients, different biochemical reactions and process parameters influence the taste and aroma of beer.
Apart from making brewers and beer enthusiasts better understand the brewing process and the science of beer, the course also proves to be a fantastic vehicle for teaching people quite complex chemistry, biochemistry, and engineering aspects. By applying all these aspects to beer, students somehow stay more interested ☺. If you find the subject interesting, all those enzymes and thermodynamic reactions suddenly become a lot more interesting, especially if you realize how they influence the final beer! We certainly assume that the students have a basic scientific knowledge. For example, we no longer explain how DNA works or what a protein is.
We have opted for a layered approach, where we start from basic principles and then explain the chemical reactions involved. For example, we first explain what happens when you roast malt, and then the science geeks among us can click even further to delve deeper into the science. The advantage of a MOOC is that you can go through everything at your own pace, you can go to the discussion forum for extra feedback and you can also largely decide for yourself how deep you want to delve into a particular (scientific) aspect of brewing.
In addition, we also have the great advantage that many top-level guest speakers are involved in the MOOC. In addition to the knowledge we have here at KU Leuven, authorities in the research field such as Prof. Charlie Bamforth (UC Davis) and Tom Shellhammer (Oregon University) are great assets for such a course, and top industry people such as Anne -Françoise Pypaert (Brewer Orval), Prof Hedwig neven (main brewer of Duvel-Moortgat), and Dr David De Schutter (Global VP Innovation & Technology Development AB InBev) show how this science is applied in the professional field.
So far, in the first year after its release there have already been more than 10 000 people worldwide that have followed our MOOC; and quite frankly we have been overwhelmed with the positive response so far! We have put an enormous amount of effort in this MOOC and it was really gratifying for us to receive so much positive reactions and enthusiasm; and to see all the interactions on the discussion forum – ranging from people having questions about some scientific concepts covered in our MOOC, to people sharing brew tips and tricks with each other! One student even went so far as to write: “I cannot imagine this course won't become a bible for all brewers worldwide”. That definitely made us blush!
Edgar: Thanks so much Karin for providing more insights about your work. It was great to get a picture (almost literally!) of all the hard work that goes into a published piece of work, not to mention to learn about the MOOC, a course that meets such an important need for brewers and breweries.
Enjoyed this article? Consider supporting us by subscribing to this newsletter and share this article far and wide.
Best wishes,
Edgar, The Beerologist at ExtrAnalytics.