iBiology

Bringing the World's Best Biology to You

Home » Research Talks » Bioengineering

Metabolic Engineering and Synthetic Biology of Yeast

00:00:11.21 So, my name is Jens Nielsen and I'm
00:00:13.13 going to talk to you today about metabolic
00:00:15.00 engineering and synthetic biology of yeast.
00:00:17.02 And to start off, let me put into context
00:00:20.23 and tell a little bit about cell factories and their
00:00:23.22 use. So this is a quite old concept that we are using
00:00:27.04 cell factories in the production of beer, wine, bread,
00:00:30.06 and yogurt, as traditional products. But also
00:00:33.13 a number of pharmaceuticals are being produced using
00:00:35.16 cell factories, antibiotics, hormones, and anti-cancer drugs.
00:00:39.06 And if we look into biopharmaceuticals, to date they
00:00:43.00 produce more than 300 of these. And they have sales exceeding
00:00:46.11 $100 billion U.S. dollars. But more recently, also there's been
00:00:51.01 more focus on using cell factories for the production of fuels
00:00:53.14 and chemicals. Bioethanol is a classical example.
00:00:56.17 But also citric acid going into soft drinks, and more recently
00:00:59.20 1,3 propanediol, which is used for polymer production
00:01:04.11 that goes for example into carpets and other types of fabrics.
00:01:07.03 If we look at this biorefinery concept, which is very
00:01:12.21 much about production of chemicals and fuels, the idea
00:01:16.01 is to use plant feedstocks as a feedstock for the synthesis
00:01:22.00 and production of chemicals and fuels. And of course,
00:01:25.10 this has to go and be integrated in global sustainable
00:01:30.04 society, where we also use solar, wind, and other types of
00:01:34.00 sustainable energy sources. But particularly, the cell factories
00:01:38.13 can provide liquid transportation fuels, but also different
00:01:42.16 types of biochemicals. So in the biorefinery concept,
00:01:46.15 what you do is you have biomass that you are converting
00:01:50.15 -- can come from different sources, you convert to sugars
00:01:53.14 and then the sugars are used in a microbial fermentation,
00:01:57.20 where you then produce the fuels and the chemicals here.
00:02:01.14 And what is characteristic in the bioreactor is, you have
00:02:06.05 a cell factory here that you use for this bioconversion
00:02:10.12 process. And what we normally do is, we do metabolic
00:02:14.01 engineering of this cell factory in order for it to change its
00:02:17.12 metabolism such that it can produce these different fuels
00:02:20.04 and chemicals. If we look a little bit about the value chain
00:02:23.20 in the conversion process. So again, we have up here
00:02:28.04 the pre-treatment, the fermentation, then you come
00:02:31.07 to purification, and formulation, the traditional processes
00:02:33.24 in the chemical industry. But what is characteristic here is
00:02:37.04 that the development of the cell factory that is used
00:02:40.05 in the biotransformation is really the research intensive
00:02:43.07 part. And total cost for developing a cell factory is
00:02:46.15 typically in the range of $50 billion U.S. dollars --
00:02:50.04 sorry, $50 million U.S. dollars. And there's a trend
00:02:54.08 therefore a trend towards using partnerships where
00:02:57.04 companies are sharing risk and capital investments.
00:03:00.22 And so, some typical examples are illustrated here,
00:03:03.24 BP and Dupont partnered up to make isobutanol,
00:03:07.01 Total and Amyris partnered up on making farnesane,
00:03:10.04 that can be used as jet fuels and for heavy trucks,
00:03:14.01 but also 1,4 butanediol, a partnership between
00:03:18.08 genomatic and BSF for production of 1,4 butanediol,
00:03:21.24 which is used for the production of spandex and other
00:03:23.23 types of polymers. So, briefly, an introduction about
00:03:28.16 metabolic engineering. So, yeast cells naturally produce
00:03:32.04 ethanol and CO2 and more yeast cells, as illustrated here.
00:03:36.00 And these are of course the traits that are used in the
00:03:38.22 production of beer, bread, and wine, but also bioethanol
00:03:42.02 production. And so, what do we do when we want to use
00:03:45.05 this yeast cell for production of other chemicals is we
00:03:48.23 take use of this very efficient glycolytic pathway that yeast has
00:03:54.19 in the conversion of glucose to ethanol. But we want to tap
00:03:57.20 off the stream here and insert, typically a pathway by taking
00:04:02.03 genes from another organism, it can be from a bacteria,
00:04:05.02 or it can be from plant cells or whatever. What kind of
00:04:07.05 product we want to produce. Well, let me just illustrate it here.
00:04:10.06 Let's say we want to produce a compound that can be used
00:04:13.14 as a biodiesel, we would tap it off from the main glycolytic
00:04:17.14 pathway here, and insert this pathway here, and we will produce
00:04:21.05 some of this secreted by the cells. But as you can see,
00:04:25.04 a very small amount of this is produced when we first insert
00:04:29.00 this pathway in yeast, because yeast has evolved of course
00:04:31.10 to be efficiently converting the glucose into ethanol.
00:04:34.04 So the challenge we're facing is to redirect or rewire the
00:04:38.04 metabolism such that we begin to produce more of the biodiesel
00:04:41.02 and basically reduce the production of ethanol. And this is
00:04:45.00 basically the essence of metabolic engineering, where we need to
00:04:47.23 understand and get insight into the yeast metabolism, in order
00:04:51.21 to do this. So therefore, we also often see that it's quite easy
00:04:57.05 to make a proof of principle strain here. But what is much
00:05:01.03 more difficult is to make a final strain here that meets certain
00:05:04.09 requirements in titer yield and productivity over here.
00:05:08.01 And this final strain here, we have to obtain in order to
00:05:11.12 have an economically viable process. And so the development
00:05:14.19 from this proof of principle strain here to this final strain
00:05:18.22 here is what is costly. It takes time, 3-6 years, and several
00:05:23.23 person -- hundreds of person years. So, what is really what we're
00:05:28.01 looking for is enabling technologies, novel technologies that can
00:05:31.10 reduce this development time such that we can bring new products
00:05:35.11 and new technologies onto the market faster.
00:05:38.10 So, but if we look at the challenge on this. So if we look at a
00:05:43.07 yeast cell, it has about 1500 metabolic reactions. So these
00:05:46.23 are associated with about 900 genes. If we take a human cell,
00:05:50.17 it's more, it's about 8000 metabolic reactions associated with
00:05:53.17 3000 reactions. So metabolism is quite complex and diverse.
00:05:57.24 And so, in order to study and get insight into this, and particularly
00:06:02.04 engineer it, we need a number of different tools at hand.
00:06:06.00 So that's why we often talk about the metabolic engineering
00:06:08.15 cycle, or also often called the test, build, design cycle.
00:06:13.12 So, we often would start with some kind of design criteria
00:06:18.00 here, we would implement these in a strain by strain construction,
00:06:22.02 we would then characterize this in a fermentation process here,
00:06:25.07 we would try to design a process that has industrial-like conditions,
00:06:29.15 and we can then use a number of tools from systems biology
00:06:33.16 to characterize and phenotype the strain, and that can then be used
00:06:37.21 for further design. And so we often have this cyclic operation in
00:06:42.04 development. And we may have to go through this cycle many
00:06:44.24 times, in order to develop the strain to become applicable
00:06:48.24 for an industrial based process. This part over here we normally
00:06:54.01 call systems biology, so this is here, we have gained a lot of
00:06:57.00 experience and knowledge from the development of systems biology.
00:07:00.00 And the strain construction is often also referred to as synthetic
00:07:03.16 biology, and particularly, many tools in genome editing coming from
00:07:07.08 the synthetic biology field have impacted metabolic engineering
00:07:10.05 significantly. So, synthetic biology is a broad discipline but it
00:07:14.19 interacts closely with metabolic engineering in certain applications.
00:07:19.04 And particularly in the design and implementation of novel cell factories
00:07:23.04 and their properties. Synthetic biology has provided a number of
00:07:26.18 new tools, as we discussed recently in this commentary here
00:07:31.09 on the emergence and the interactions between synthetic biology
00:07:36.01 and metabolic engineering. So if we look at yeast as a
00:07:40.07 cell factory, why is that attractive to use? Well, first of all,
00:07:45.02 there are some advantages. It's extremely well-characterized,
00:07:48.12 we have a lot of information, it's genetically tractable,
00:07:51.06 it's what is characterized as generally regarded as safe,
00:07:54.11 that means the products produced in yeast are relatively
00:07:57.10 easy to get on the market, and it's also very robust,
00:08:01.07 so that means it is very easy to implement for industrial settings.
00:08:04.19 And so this is also why it's already used for production of
00:08:08.06 a number of different products. Bioethanol, but also
00:08:11.00 many higher value products, including also pharmaceutical
00:08:15.03 proteins. And there's also a number of ongoing development
00:08:19.01 of yeast cell factories for production of fuels and commodity
00:08:21.23 chemicals and so on. So yeast is to a large extent, one of the
00:08:26.05 preferred cell factories in industry and these are some of the
00:08:29.20 reasons for that. So, in my lab, we're therefore working on
00:08:35.01 developing yeast as a platform organism to produce a range
00:08:38.13 of different chemicals here. They range from biofuels to commodity
00:08:42.05 chemicals, and fine chemicals, and proteins. And in the
00:08:46.14 process here, we are using and developing also new synthetic
00:08:51.01 biology tools, as well as systems biology tools to phenotype
00:08:54.02 the yeast strains here. And I'm going to give a couple of
00:08:57.13 examples on production of some of these compounds listed
00:09:00.16 here to illustrate these principles here.
00:09:05.03 So, one general comment first is that often when we are going to
00:09:10.04 produce different types of products, as illustrated here, it can
00:09:13.17 be 1-butanol, it can be isoprenoids derived from plants,
00:09:16.23 for example, sterols, polyketides, polyphenols, alkanes,
00:09:20.23 alkenes, we often see that they are derived from the same
00:09:24.19 pathway intermediate, in this case here, acetyl-CoA,
00:09:28.14 which is a key central metabolite in metabolism.
00:09:32.20 And so, often the challenge in actually overproducing
00:09:36.02 and efficiently producing all of these products here boils down
00:09:40.02 to having an efficient conversion of glucose here down to acetyl-CoA.
00:09:44.23 And so, here we are challenged in yeast by the fact that acetyl-CoA
00:09:49.19 is in different compartments, it's in the nucleus, it's in the peroxisome,
00:09:53.01 it's also in the mitochondria, as illustrated here. So, what we need
00:09:57.00 to do is if we want to implement these pathways in the cytosol,
00:10:00.01 we need to make sure that the glucose is converted efficiently
00:10:03.13 to acetyl-CoA here. And this is also why we talk about platform
00:10:07.04 cell factories, because if we have an efficient cell factory for
00:10:10.22 conversion of glucose to acetyl-CoA, we can use that cell factory
00:10:14.09 for production of a range of different products here. Where we insert
00:10:17.17 the corresponding pathways. So, let me give a couple of examples
00:10:22.16 of acetyl-CoA derived products. The first one is 3-hydroxypropionic
00:10:25.19 acid, this can be derived in a simple two step pathway, acetyl CoA
00:10:32.03 goes to malonyl-CoA that then goes further to 3-hydroxypropionic
00:10:35.07 acid. This chemical is interesting because it can be converted into
00:10:39.19 acrylates, which are used as super absorbent polymers, for example
00:10:43.06 diapers, but find a number of different applications as a polymer.
00:10:48.03 The market is quite large, as illustrated here, and there are several
00:10:50.17 companies that are interested in producing this. So we worked
00:10:54.10 on engineering expression, first of all, the enzyme for malonyl-CoA
00:10:58.21 to 3-hydroxypropionic acid, but then we demonstrated that
00:11:02.05 by optimizing the production of acetyl-CoA, the precursor
00:11:06.18 up here, we could increase significantly the production of
00:11:10.06 3-HP, as illustrated here. So this was about a 10-fold improvement
00:11:13.19 that was obtained by providing more efficiently the precursor
00:11:17.10 upstream of this pathway here. We also worked on another
00:11:23.24 that is also used as a feedstock for production of polymers.
00:11:30.02 This is succinic acid, there's also a lot of interest in this
00:11:33.03 in the field today. Succinic acid can be used for de-icing of
00:11:37.02 airplanes, but can also be used as a feedstock for production
00:11:40.19 of biopolymers like spandex. And so what we did was we
00:11:44.20 did metabolic engineering of yeast as well as combining
00:11:48.09 this with adaptive laboratory evolution and with this process
00:11:51.13 we see an improved production of succinic acid by 80-fold.
00:11:55.08 Another product -- the opportunity to produce the polymers directly
00:12:02.21 in yeast, so what we did here was we took a pathway from
00:12:06.16 bacteria and expressed it in yeast to produce this
00:12:09.19 polyhydroxybutyrate, which is a biodegradable polymer.
00:12:13.00 There's much interest in using that, Coca Cola is introducing
00:12:17.09 this for their Coke bottles. And what we did was, that again
00:12:21.20 illustrated here, by insertion of the pathway in yeast,
00:12:25.00 we produce relatively low levels, as seen over there on the reference.
00:12:29.01 But when we then engineered again this acetyl-CoA supply
00:12:32.09 we can boost the production 80-fold inside the cell.
00:12:35.24 We begin to see accumulation here, as the white dot illustrated here,
00:12:39.24 of this polymer inside the cell. Another class of compounds
00:12:45.16 that there's much interest in is isoprenoids, and particularly,
00:12:50.11 one subclass of these, sesquiterpenes. Some of these
00:12:54.06 can find a number of applications. Here is listed
00:12:57.08 four compounds that are of interest to be used for perfumes
00:13:00.10 or in perfume ingredients, and this is work we did in collaboration
00:13:04.02 with a Swiss company, Firmenich. And so what we did was
00:13:07.19 we inserted the plant genes for production of these
00:13:11.02 pathways -- of these metabolites in yeast. And at
00:13:15.04 the same time, we worked on up-engineering of the upstream
00:13:18.08 metabolism, such that we could efficiently produce these
00:13:22.04 different perfume ingredients. And here I'm just going to illustrate
00:13:25.00 briefly, for santalene. So by expression of the santalene
00:13:31.00 synthase directly in yeast, as you can see over there,
00:13:34.03 that we have a very low level of production. But by engineering
00:13:38.07 the upstream pathway by targeting a number of different
00:13:41.23 steps, but also optimization of the fermentation process,
00:13:44.19 we could more than 10-fold improve this and actually reach
00:13:48.02 a titer that was sufficient to push this forward for beginning to
00:13:53.00 have commercial production. Finally, we -- to mention also,
00:13:59.01 we also worked on another pathway derived from acetyl-CoA,
00:14:01.15 n-Butanol production. So here, you can see we took a pathway
00:14:05.22 from bacteria again. And yeast naturally does not produce
00:14:09.18 1-butanol, so we expressed these enzymes in yeast.
00:14:12.12 It gives initially very low level production, but again through
00:14:16.16 engineering of the acetyl-CoA metabolism, we could increase
00:14:19.19 the production about 5 to 6-fold. And butanol is interesting to
00:14:24.15 be used both as a chemical building block, but it can also be used
00:14:27.06 as an alternative biofuel. Biodiesel, there's also much interest in
00:14:33.13 producing biodiesel. We've taken different approaches to this
00:14:36.09 One is to produce what is very similar to what is used in
00:14:41.00 biodiesels today, these are esters. So what you do today is
00:14:44.14 that you extract plant oils and you do a re-esterification,
00:14:49.11 as illustrated up here, with methanol typically, so you produce
00:14:52.07 fatty acid methyl esters. We expressed enzymes in yeast
00:14:57.01 so we can take fatty acids, fatty acyl-CoAs, esterize them
00:15:01.12 with ethanol here, that yeast naturally produce, and then produce
00:15:04.19 the fatty acid ethyl ester that is completely compatible
00:15:08.04 with the current biodiesel used. And again, we illustrated here
00:15:12.05 by engineering, yeast central metabolism here, we could
00:15:15.12 increase the production of these fatty acyl ethyl esters
00:15:18.09 by more than 5-fold. What there's really interest in is
00:15:23.06 to produce alkanes, because these are completely compatible
00:15:26.21 with the current infrastructures. So what we have also been
00:15:30.20 working on is to convert the fatty acyl-CoAs directly to alkanes
00:15:34.21 and these can then be used as diesel substitute and be directly
00:15:39.10 blended into both heavy diesel and also into the jet fuels.
00:15:45.05 A couple of other examples of products we worked on,
00:15:50.23 we also worked on production of resveratrol, this is an antioxidant
00:15:55.23 ingredient in plants, for example in grape. It's clinically proven to be
00:16:00.03 beneficial for a number of different human diseases, cancer and
00:16:03.14 type 2 diabetes treatment, for example. But it's also used
00:16:05.21 in cosmetics. And this is a pathway that is derived from aromatics,
00:16:11.02 and we express then the plant genes in yeast and we could then
00:16:16.12 initially produce some proof of principle levels here. But by
00:16:20.04 pathway optimization, particularly upstream of phenylalanine
00:16:25.04 biosynthesis, we could significantly improve the production
00:16:28.01 as illustrated here. And so this is an example also of how we
00:16:31.15 can insert complex plant pathways in yeast and use that
00:16:36.01 -- the yeast thereby to more efficiently produce these compounds.
00:16:39.02 We also worked on human insulin. Yeast is already used
00:16:43.18 for commercial production of human insulin by Novo Nordisk,
00:16:47.00 which is the largest insulin producer in the world. It's also
00:16:51.01 used for production of many other biopharmaceutical proteins.
00:16:54.17 But we were interested to see if we could make an efficient
00:16:57.02 expression system for expression of insulin, but also other
00:17:00.09 secreted proteins in yeast. And so we evaluated the number
00:17:03.14 of different expression systems, as illustrated here. And then
00:17:07.11 we could identify a very efficient expression system where we
00:17:11.05 could 20-fold improve the secretion of the insulin out of the
00:17:14.10 yeast cells, and actually reach titers close to 100 milligrams
00:17:19.13 per liter, which is quite significant and is close to what you
00:17:22.19 would require for industrial scale production. We also work on
00:17:27.02 production of human hemoglobin, so this is a protein that
00:17:32.16 has a heme group inserted into it, and if we can therefore
00:17:36.02 not have it secreted, so it's accumulating inside the cell.
00:17:38.12 But there's much interest in producing human hemoglobin
00:17:41.12 as a blood substitute to be used, for example, in transfusion.
00:17:45.10 And hereby, you could prevent disease transfer and hemoglobin
00:17:50.18 is a key ingredient in blood transfusion because of its oxygen
00:17:55.03 carrying capabilities. What we did was we combined expression
00:17:58.22 of different forms of the protein of the human hemoglobin,
00:18:03.13 but we also upregulated the heme biosynthetic pathway in yeast.
00:18:07.13 And hereby, we could increase the accumulation more than 3-fold
00:18:11.15 of hemoglobin inside the yeast cell. So let me end off talking
00:18:17.05 a little bit about somewhat different approach. So we were interested
00:18:21.19 in having yeast cells that could grow at higher temperatures
00:18:26.04 and there is much interest in that in the bioethanol industry,
00:18:31.01 because then you could reduce the cost of cooling, you can also
00:18:35.09 have a more stable process, for example, combining with
00:18:37.24 enzymes that like to operate at higher temperatures. So
00:18:41.22 the problem however, is that it's difficult for yeast to grow
00:18:44.12 above 35C, it has very, very slow growth. And so one could
00:18:49.19 maybe begin to look at how the heat shock response functions
00:18:54.02 in yeast, and engineer this. But the problem is the heat shock
00:18:58.04 response is something that the cell has evolved to cope with
00:19:01.22 sudden temperature stresses. And basically, cope with this stress
00:19:07.12 in order to survive when the temperature later on drops again.
00:19:11.00 So we therefore took another approach, where we did so-called
00:19:15.20 directed adaptive laboratory evolution. So we grew yeast
00:19:20.00 over a series of shake flasks and many generations at 40 degrees,
00:19:24.08 39.5 to be more specific. And after about 600 generations
00:19:31.00 we then isolated clones that were able to grow better at
00:19:35.00 this higher temperature. As illustrated here, we did this in 3
00:19:40.01 shake flask series. We isolated three clones from each of these
00:19:43.11 flasks, and as shown here, that all of these 9 clones here,
00:19:46.19 they grow about twice the rate of the wild type strains at this
00:19:50.23 elevated temperature. This we also characterized these strains
00:19:54.17 in bioreactors, and as shown here, they grow faster at this temperature
00:19:58.02 here. And they also have a higher glucose uptake rate.
00:20:02.08 So, we then -- sorry, we're then interested to see what happened
00:20:08.12 actually in these strains. Which genetic mutations accumulated
00:20:11.09 in order to accommodate for this improved growth at higher
00:20:14.20 temperatures. So we did genome sequencing, we found
00:20:18.03 that in some of the clones that we isolated, there was a duplication
00:20:21.24 in part of chromosome 3, which is one of the smaller chromosomes
00:20:26.03 in yeast. And we can see here that there are certain segments
00:20:30.02 that were duplicated here. One could of course speculate whether this
00:20:33.03 had anything to do with the adaptability and ability to grow at
00:20:37.14 higher temperatures. However, we also noticed that there was
00:20:41.20 accumulation of point mutations in several genes, as indicated
00:20:45.02 over here. And particularly, in one gene was identified to have
00:20:49.24 a point mutation in the clones that we isolated. And that was ERG3,
00:20:55.05 so we of course were zooming in to look into ERG3 further.
00:20:58.19 We did see also accumulation of point mutations in ATP3,
00:21:03.21 so we also looked further into that. And when we inserted the
00:21:07.15 point mutation into ATP3, you can see here in this M22 strain
00:21:11.07 here, we can see that that is not really what brings the
00:21:15.03 phenotype is not conferring increased growth at high
00:21:19.00 temperatures. But when we introduce point mutations in
00:21:22.02 the ERG3 that we identified by mutation, we can see in the M7
00:21:25.16 strain here, that we basically have improved growth. Almost
00:21:30.12 as good as those strains that we evolved, and much better
00:21:33.20 definitely than the wild type. So what is the mutation in ERG3
00:21:37.09 bring to the cell? Well, what we found was that there is a complete
00:21:41.03 shift in its sterol content. You can see here, the wild type mainly
00:21:45.19 is accumulating ergosterol, which is the typical sterol
00:21:49.01 of yeast. Whereas, the evolved mutant is accumulating fucosterol,
00:21:54.05 which is illustrated here. It's a bended sterol compared with
00:21:57.14 ergosterol, which is a more flat sterol. And also when we introduce
00:22:02.01 the point mutation in ERG3, we can also see that we get this
00:22:06.01 fucosterol here. So, we are confident that basically changing
00:22:09.17 the sterol composition is what brings this phenotype and the
00:22:12.21 ability to grow at an elevated temperature. And it's kind of
00:22:15.18 remarkable that just a single point mutation actually is sufficient
00:22:19.11 to bring this adaptability of yeast to grow at higher temperatures.
00:22:23.02 This performance, or this ability to grow at higher temperatures has
00:22:29.03 significant translational impact that's illustrated over here.
00:22:32.19 Because if you grow at higher temperatures, you can see that
00:22:35.12 the wild type is producing less ethanol than the evolved mutants,
00:22:39.15 as illustrated here. So with that, let me just acknowledge
00:22:44.06 everyone who has been working on this. Here's a list of the people
00:22:48.22 and current members and alumni that have been working on this,
00:22:51.07 and also collaborators. And of course, also the funders.
00:22:53.11 And thank you for your attention.

This Talk
Speaker: Jens Nielsen
Audience:
  • Researcher
Recorded: June 2015
More Talks in Synthetic Biology
All Talks in Synthetic Biology
Share

Talk Overview

Dr. Jens Nielsen introduces the idea that cells can act as microbial factories for the sustainable production of diverse products. Dr. Nielsen explains that the goal of metabolic engineering is to alter a cell’s metabolism to produce a desired product. He gives examples of successful implementation of these concepts in yeast, including acetyl-CoA overproduction and synthesis of polyhydroxybutyrate (for biodegradable plastics), n-butanol (a biofuel) and resveratrol (a cancer and diabetes drug). He concludes by highlighting a project to improve the thermotolerance of yeast to facilitate industrial applications.

Speaker Bio

Jens Nielsen

Jens Nielsen

Jens Nielsen received his PhD in Biochemical Engineering from the Danish Technical University. Currently, he is a professor and head of the Division of Systems and Synthetic Biology at Chalmers University of Technology in Gothenburg, Sweden. His research focuses on understanding metabolic processes in humans and microbes along with developing chemical and protein production pathways… Continue Reading

More Talks in Bioengineering

Reader Interactions

Leave a Reply

Your email address will not be published.