Engineering Microbes to Solve Global Challenges

Top

00:00:07.00 Hi, my name is Jay Keasling.
00:00:08.26 I'm from the Joint BioEnergy Institute,
00:00:11.00 the University of California at Berkeley,
00:00:13.04 and Lawrence Berkeley National Laboratory.
00:00:14.22 And today I'm going to be talking about engineering
00:00:16.22 microbes to solve global challenges.
00:00:18.28 It's probably no surprise to anyone that we derive
00:00:22.29 most of our fuels, chemicals, and even drugs from
00:00:26.09 petroleum. We pump it up from underground, we refine it
00:00:29.11 in a refinery, and then from that refinery, we produce all of
00:00:33.29 the fuels, chemicals, and drugs that we use on a day to day
00:00:37.01 basis. Now, if you just look at a typical barrel of oil.
00:00:41.11 Roughly, 80-85% of that barrel of oil goes to producing
00:00:45.11 all of the fuel that we use. In the U.S. that's on the order
00:00:49.09 of 225 billion gallons of transportation fuels every year.
00:00:53.27 The other 15-20% goes into producing all of the other
00:00:58.15 products, from carpet fiber to paint to plastic bottles, to even
00:01:03.00 some of the drugs that we ingest into our body.
00:01:05.24 Now, we'd like to have an economy that's not just
00:01:09.04 based on petroleum, but that's also based on renewable
00:01:12.10 feedstocks. Things like sugarcane, corn, and even cellulose
00:01:17.19 biomass in the form of grasses. Now if you were a farmer,
00:01:22.26 and you were to grow those various crops depending on where
00:01:26.00 you are in the world, you'd send them to a bio-refinery. And that
00:01:29.10 bio-refinery, just like a petroleum refinery, would split
00:01:32.27 those molecules apart, those sugar containing compounds apart,
00:01:36.13 and transform them into fuels, into chemicals, and even
00:01:40.11 into drugs. If you burn those fuels in your car, you'd still
00:01:44.06 be producing CO2 like you do when you produce
00:01:47.11 and burn petroleum based fuels, but that carbon dioxide
00:01:50.24 would be taken back up by the plant and incorporated
00:01:54.06 into the plant containing sugars, that then would be
00:01:58.04 refined into fuel. And in this sense, you'd have a carbon
00:02:01.19 neutral cycle of producing and consuming carbon dioxide.
00:02:06.10 Now if we just look at the volume and the value
00:02:11.20 of these molecules, at the very top of the volume and the bottom
00:02:16.12 of the value are the fuels. They have the lowest
00:02:19.20 value of many of the molecules that we use on a regular
00:02:23.00 basis. And yet, they have the highest volume. Remember I
00:02:26.26 told you that's about 225 billion gallons of transportation
00:02:30.20 fuel in the U.S. alone on an annual basis.
00:02:33.25 At the top of the values, and at the bottom of the
00:02:37.10 volume are the drugs we use. Extremely high value
00:02:40.22 molecules produced in small volumes. And then
00:02:44.11 in the middle are all of the other molecules and products
00:02:47.06 that we use on a day-to-day basis, like carpet on the
00:02:50.25 floor, the plastic bottles, the paints on the wall, mouthwash,
00:02:54.07 perfume, etc ... Now when we think about flexibility for
00:02:59.23 substitution, we know that with our drugs we don't want
00:03:02.23 anything that's just kind of like the drug we use but
00:03:05.28 exactly like the drug we know will cure a particular
00:03:09.28 disease. But we often think that there's a great deal more
00:03:13.28 substitution we can have for things like carpet fiber,
00:03:17.05 for mouthwash, and even more for fuels. However, that's
00:03:20.24 not really the case. Our fuels are regulated, what we put
00:03:24.24 in our tank is highly regulated, because that's what gets
00:03:28.09 exhausted into the atmosphere. And that ensures that our
00:03:32.08 planes stay up at the altitude. Same case for plastic bottles,
00:03:36.16 and carpeting, and paint. We need to know that those
00:03:39.28 products will work the same as they've always worked
00:03:42.25 regardless of whether they're produced from petroleum
00:03:46.09 or from carbon that comes from plants. Now today, I'm going
00:03:51.11 to be talking about engineering microbes, namely yeast
00:03:55.09 and E. coli, to produce a variety of these products.
00:03:58.15 And to produce products that behave or are identical
00:04:01.27 to the products that we would normally get from petroleum.
00:04:05.04 And I'm going to start off with a story about drugs.
00:04:08.29 And in this case, the story about artemisinin, a drug that
00:04:12.22 cures malaria. This guy, his name is George, I met him when
00:04:18.27 I traveled to Kisumu, Kenya. Kisumu is on Lake Victoria
00:04:24.15 and it's about sea level. And it has a very high incidence rate
00:04:30.02 of malaria. In fact, 80% of the population has malaria.
00:04:34.12 When I visited Kisumu that day, it was very warm,
00:04:38.02 but yet George came into this clinic wearing a long
00:04:41.12 coat and pants, and he had a runny nose. And it was pretty
00:04:44.01 clear that he wasn't feeling well. This is the clinic that he
00:04:47.15 came into, it's a government run clinic, and although it's
00:04:50.16 difficult to see, there are children that are lined up outside
00:04:53.18 the clinic waiting to get in. This at the top is a picture of
00:04:59.07 the typical home, a home that George might live in. It has
00:05:01.26 mud walls, a steel roof, but you could tell that the steel isn't
00:05:06.17 in very good shape. The nurse saw George and she knew that
00:05:10.19 he probably had malaria. So she pulled out a brand new
00:05:13.23 test kit that she just had access to. She pricked George's
00:05:17.29 finger, put a drop of blood at the bottom, and then
00:05:21.07 followed that with a couple of drops of buffer. A few seconds
00:05:24.13 later, the blood ran up the gradient. The two lines that
00:05:28.24 appeared, the top line is the control line to say that the kit
00:05:31.24 worked, and the bottom line by the T is the test line.
00:05:35.24 And that indicated that George had malaria.
00:05:38.12 The nurse then took George and weighed him, and
00:05:41.24 from his weight, got a package of Coartem
00:05:44.28 which contains artemisinin, an antimalarial drug.
00:05:48.15 She gave him the first tablet and then she instructed
00:05:51.15 his brother to give him a tablet 12 hours later, the third
00:05:56.24 tablet 12 hours after that, the fourth tablet 12 hours
00:05:59.28 after that, and so on. She repeated the instructions and
00:06:03.08 sent them on their way. And then she turned to me and
00:06:05.18 she said, "Jay, after the third tablet, George is going to be
00:06:08.29 feeling great. And his mother's going to take the rest of
00:06:12.00 those tablets and tuck them away in a drawer, because
00:06:14.04 the next time he gets malaria, she may not have access
00:06:17.08 to this free Coartem that she was able to get."
00:06:20.07 And because they're living on less than $1 a day,
00:06:23.04 there's no way she could afford the Coartem or any other
00:06:26.27 anti-malarial drug that would be sold in the private pharmacies.
00:06:32.04 A little bit about malaria. Roughly 3 billion people
00:06:35.18 on the planet at any one time, are at risk of getting malaria.
00:06:39.08 Roughly 250 million people are infected with the malaria
00:06:43.25 plasmod, the causative agent, called plasmodium.
00:06:47.28 And at any one time, roughly 500,000 to 1 million
00:06:52.12 people die of malaria, 2/3 of them are children under the age
00:06:56.10 of 5. There's a treatment, a cure for malaria, it's called
00:07:01.11 artemisinin. And it comes from this plant Artemisia annua.
00:07:04.07 And it has a great history, it goes back to 168 B.C.
00:07:07.25 where it was first described in the writings for being used
00:07:11.29 for treating hemorrhoids. And then a few hundred years later,
00:07:15.13 there are writings of it being used to treat fevers.
00:07:18.07 Presumably fevers due to malaria. It was largely
00:07:21.17 forgotten, but in the '60s and '70s, China was fighting
00:07:27.25 in Vietnam and malaria was rampant in Vietnam.
00:07:31.15 And the drugs they currently had available, based on quinone
00:07:34.18 were no longer effective. So Mao sent his medical core out
00:07:38.25 to find a cure for malaria. This woman here, Youyou Tu,
00:07:43.14 was one of the people who went out and led the medical
00:07:46.28 core. She pored through the ancient literature and finally
00:07:50.28 she found a description, shown here, for using
00:07:55.12 sweet wormwood, making a tea of the wormwood,
00:08:00.03 extracting the drug, and patients drinking it.
00:08:04.04 They purified the active ingredient, found out how
00:08:06.27 best to extract it and then by 1972, had the active
00:08:12.00 ingredient isolated and described. In 2004, the World Health
00:08:17.02 Organization recommended artemisinin-based combination
00:08:19.22 therapy as the drug of choice for treating malaria.
00:08:22.24 So, what's the challenge? We've got this great cure for
00:08:27.07 malaria. It's an ancient Chinese therapy now been westernized.
00:08:30.29 What's the problem? Well, I alluded to a few of these
00:08:34.02 challenges earlier. The first one is availability. Remember
00:08:38.07 George's mother would give him 3 tablets, he'd be feeling
00:08:41.08 great, and then she'd tuck the rest away because it
00:08:44.20 wouldn't be available to them. And, if it were available to them
00:08:47.22 the cost would be too high. There are also challenges around
00:08:51.14 quality of the drugs. There are makers of artemisinin
00:08:55.12 therapies that don't put as much artemisinin in them as
00:08:58.18 they should or any at all, and try to sell them as artemisinin
00:09:01.20 drugs. And finally, there's the challenge due to resistance.
00:09:05.17 In many -- in a few parts of the world, where they've been
00:09:09.22 using artemisinin the most, resistance has -- plasmodium
00:09:14.08 has become resistant to artemisinin. And this has become
00:09:17.02 a big challenge. This is mainly occurring in regions around
00:09:21.09 Vietnam, where artemisinin was first used.
00:09:24.20 Now, this is the process by which we get artemisinin.
00:09:29.12 We grow plants on a plantation, we purify the artemisinin
00:09:33.25 from those plants, and then convert it chemically into the
00:09:37.05 derivatives that are used in artemisinin combination therapies.
00:09:40.05 I visited one of these plantations and actually saw
00:09:43.09 this in action. Now, the artemisinin is produced in
00:09:47.05 small trichomes, where it's held, that are on the leaf
00:09:50.19 of the plant. These trichomes store the artemisinin and when
00:09:54.13 you rub the plant, you can get the artemisinin off.
00:09:57.01 When you make a tea of the plant, these trichomes
00:09:59.07 burst and the artemisinin comes out. Now the way we
00:10:03.20 get artemisinin is that the seeds are planted in
00:10:09.01 small greenhouses, then once the plants grow up
00:10:12.10 to be about six inches to a foot tall, they're taken out
00:10:14.27 into the fields, they're planted by hand, where they grow
00:10:18.00 up over a growing season, and they grow to be about
00:10:20.10 8 feet tall. After they've grown up, they are then harvested,
00:10:24.24 the plants are chopped, generally by hand, and then laid out
00:10:29.00 on tarps to dry in the sun. The artemisinin is contained in
00:10:34.04 the leaves of the plants. So then they beat the plants
00:10:36.17 to get the leaves off, they collect all of those leaves
00:10:39.08 that are on the tarps, run them through sieves to get only
00:10:42.08 the smallest leaves, and then they take it into the factory
00:10:46.02 and extract it. Now, I mentioned that price and availability
00:10:50.03 were a challenge. In 2004, artemisinin was chosen as the drug
00:10:54.15 of choice for treating malaria. And by 2005, the price had shot
00:10:58.15 way up to over $1100 a kilogram. Farmers saw that they
00:11:03.01 could make money from producing artemisinin, so they
00:11:05.25 started producing lots of it. And shown in the blue bars
00:11:08.23 are the amount of artemisinin. And by 2007, they had
00:11:13.19 overproduced it so much so that the price fell to around
00:11:17.15 $200/kilogram. Farmers then stopped planting artemisinin,
00:11:21.29 there was a shortage by 2009, and the price started shooting
00:11:24.23 back up. So you see these swings in price and availability.
00:11:28.09 The other challenge with artemisinin is the long time
00:11:31.22 it takes to produce it. You can see that by the difference
00:11:35.11 in the peaks of the price and the availability. It's two years.
00:11:39.16 It takes anywhere from 18 months to 2 years for pharmaceutical
00:11:43.10 companies to decide how much artemisinin they need,
00:11:45.26 to then have contracts with farms to grow it, to get the proper
00:11:51.00 seeds that produce the most, for it to go through a growing
00:11:54.02 season, for the farmers to harvest it and get it back
00:11:56.19 into the factory where it's extracted, and then turned
00:12:00.00 into the pharmaceutical companies, where they then
00:12:03.02 make the derivatives that are produced and used in
00:12:05.24 artemisinin combination therapies. So that time window
00:12:09.20 of 18-24 months is also a challenge.
00:12:13.10 Now, what else could farmers plant if they didn't plant
00:12:17.07 artemisinin? Well, they could plant many food crops
00:12:19.17 and shown here are the prices of artemisinin compared
00:12:23.04 with food crops in China and Vietnam. Early in
00:12:27.07 the use of artemisinin in 2004, many farmers in Vietnam
00:12:31.29 started planting artemisinin. But those farmers went out
00:12:35.15 of business, since large plantations in China began
00:12:39.13 to sprout up. Now most of the artemisinin is produced
00:12:42.16 from plants grown in China. I mentioned that resistance
00:12:48.10 is rising to artemisinin. You can see here that several
00:12:52.25 countries allowed sale and production of mono-therapies.
00:12:57.00 Mono-therapies, because there's only one drug in them,
00:13:00.23 actually encourage resistance. If you have more than one
00:13:04.16 drug that fights the plasmodium with multiple modes
00:13:07.27 of action, then you have a chance of containing that
00:13:10.18 resistance. And as I mentioned earlier, there's been
00:13:13.10 resistance documented on the Cambodia-Thailand
00:13:16.23 border near Vietnam. So, several years ago, our
00:13:21.04 goal was to engineer a microbe to produce artemisinin
00:13:24.07 from an inexpensive renewable resource that would allow
00:13:26.25 us to stabilize and reduce the price. Stabilize and increase
00:13:30.14 supply. Reduce production time. And control resistance
00:13:34.03 by controlling the supply. This was our process.
00:13:37.27 We were going to replace the plant with microbes, either
00:13:41.20 E. coli or yeast. Rather than produce artemisinin, we
00:13:44.15 proposed producing artemisinic acid. Artemisinic acid is
00:13:48.00 less toxic than artemisinin to yeast and E. coli.
00:13:51.18 But there was a readily known conversion process that
00:13:54.24 would allow us to convert artemisinic acid into
00:13:57.02 the same derivatives that are currently produce from
00:13:59.04 artemisinin. So then it could be substituted directly
00:14:01.21 in the drug pipeline. This is the strategy we would use.
00:14:07.13 We would engineer a microbe to produce artemisinic acid,
00:14:10.15 we would then hope to get that artemisinic acid
00:14:12.29 out of the microbe, some way through a purification
00:14:15.29 process. And then chemically convert it into the derivatives.
00:14:19.04 We started with E. coli as our first chassis.
00:14:22.29 E. coli is an organism that we know a lot about, we have
00:14:25.21 a lot of genetic tools for engineering it. And so that
00:14:29.00 became our first chassis. E. coli produces farnesyl
00:14:32.21 pyrophosphate, a precursor to artemisinin.
00:14:35.04 And so that gave us the first step in that biosynthetic
00:14:38.13 pathway. Unfortunately, E. coli produces very little
00:14:41.19 farnesyl pyrophosphate, so we brought in heterologous
00:14:46.08 pathway from another organism, yeast. In yeast this
00:14:49.22 pathway produces a relative of cholesterol called
00:14:52.25 ergosterol, that's needed for the growth of the yeast.
00:14:56.22 We used this pathway to produce farnesyl pyrophosphate.
00:15:00.01 It tapped into acetyl-CoA, which E. coli has plentiful amounts
00:15:04.29 of. The next step in the process was to find the
00:15:10.28 enzyme that converts farnesyl pyrophosphate into
00:15:13.19 amorphadiene, the first committed precursor in the
00:15:16.18 artemisinin biosynthetic pathway. Luckily for us, that gene
00:15:20.14 had been cloned and all we thought we needed to do
00:15:24.25 was get access to the gene. Unfortunately, the person who'd
00:15:28.02 cloned that gene wouldn't give us access, so we had to
00:15:30.13 synthesize the gene directly. We also developed a number
00:15:35.03 of synthetic biology tools that would allow us to increase
00:15:37.29 the flux through this biosynthetic pathway. We developed
00:15:41.23 scaffolds that would allow us to attach the enzymes
00:15:44.11 in the biosynthetic pathway to a synthetic protein scaffold,
00:15:47.20 and thereby collect all the enzymes together so that they
00:15:50.28 could increase the flux through the pathway. We developed
00:15:54.14 regulators that could sense intermediates in the pathway,
00:15:58.10 like farnesyl pyrophosphate, which are toxic to organisms
00:16:01.14 like E. coli if they accumulate at too high a level.
00:16:04.13 And then those regulators would regulate the biosynthetic
00:16:08.10 pathway. We also developed a type of debugging tool
00:16:11.21 that would allow us to debug the inside of a cell and
00:16:14.16 determine which genes were turned on, which genes were
00:16:17.10 turned off, and what might be the cause of any toxicity
00:16:20.00 or any lack of production that we witnessed.
00:16:23.08 Through the use of those tools, through synthesis
00:16:27.13 of genes, we were able to increase the production
00:16:29.21 to about 10 million fold from where we started. And this
00:16:34.17 was work that was done both in my laboratory at UC
00:16:37.03 Berkeley, as well as in Amyris, a company that was started
00:16:39.15 out of my laboratory. Now, the next steps in the biosynthetic
00:16:44.06 pathway involve cytochrome P450s that would put an oxygen
00:16:48.04 on that amorphadiene. Unfortunately for us at the time,
00:16:51.27 the genes hadn't been cloned and hadn't been
00:16:54.13 described. But we know that it was going to involve
00:16:56.25 multiple steps and potentially multiple cytochrome P450s.
00:17:00.14 We were worried. E. coli is not a great host for expressing
00:17:05.01 these enzymes. So, the first thing we did was went about,
00:17:08.17 we went about building a yeast that would be a reagent
00:17:12.14 for us. An organism where we would engineer the production
00:17:15.16 of amorphadiene, and then use it to find the P450s that
00:17:19.16 might be involved in the biosynthetic pathways to convert
00:17:22.06 amorphadiene into artemisinic acid. We increased
00:17:26.17 the expression of the enzymes in the mevalonate pathway,
00:17:29.12 that same pathway that we imported into E. coli
00:17:32.03 to produce farnesyl pyrophosphate, and by upregulating
00:17:35.10 the genes in that biosynthetic pathway, we were able
00:17:37.21 to produce sufficient farnesyl pyrophosphate and then
00:17:41.10 amorphadiene, when we introduced amorphadiene synthase.
00:17:44.29 We could use that organism as a reagent to find the enzymes.
00:17:48.23 But we still had to go into artemisia and find the genes
00:17:53.01 in that last part of that pathway. Well, as I mentioned earlier,
00:17:57.25 artemisinin is produced in trichomes on the leaves.
00:18:00.19 And there are basal cells that produce that artemisinin.
00:18:04.19 They must be enriched in the enzymes that produce the
00:18:08.01 artemisinin, if that's where it's produced, and therefore
00:18:10.06 they'd be enriched in the transcripts that encode those
00:18:13.20 enzymes. So we went in looking for those transcripts.
00:18:17.05 We developed a method to purify the trichomes off the
00:18:20.18 leaf, which was a very difficult process. And then we made
00:18:24.23 cDNA libraries of those trichomes and looked for RNAs
00:18:28.20 that might be important for that biosynthetic pathway.
00:18:32.08 But it'd be like looking for a needle in a haystack.
00:18:36.00 So, a very creative postdoc in my laboratory said,
00:18:39.21 well, the artemisinin biosynthetic pathway is it looks
00:18:43.11 a lot like biosynthetic pathways that are in lettuce
00:18:47.03 and sunflower. If you look, the molecules are very
00:18:50.07 similar, and because the molecules look similar,
00:18:53.06 that must mean the enzymes are similar.
00:18:55.25 So he then sorted through a cDNA library that we had
00:18:59.09 available to us, of sunflower and lettuce. Looked for
00:19:02.17 P450s that were unusual, and compared those to the
00:19:06.11 P450s that we got in the cDNA library that we got from
00:19:09.12 Artemisia annua. And about that time, a bunch of
00:19:13.23 miracles started happening. The first miracle was that
00:19:16.13 the first enzyme we tried was the right enzyme.
00:19:19.27 The second miracle was that rather than just
00:19:22.17 catalyzing the first step in the biosynthetic pathway,
00:19:25.04 it catalyzed all three steps and the yeast produced
00:19:28.12 artemisinic acid. The third miracle was really where
00:19:33.03 the artemisinic acid ended up. We couldn't find it
00:19:36.24 inside the cells, we couldn't find it in the broth,
00:19:39.04 in the shake flasks. It turns out that artemisinic acid
00:19:42.04 is not as toxic as artemisinin to yeast, but toxic
00:19:44.20 enough that the yeast excreted it out of the cell,
00:19:48.06 turned on pumps to pump it out selectively,
00:19:51.00 so that it wouldn't be inside the cell and toxic
00:19:53.12 to the yeast. And because the pH of the fermentation
00:19:57.08 broth was relatively low, the artemisinic acid
00:20:01.04 precipitated on the yeast membrane. As we
00:20:05.10 increased the production of artemisinic acid,
00:20:08.16 it actually comes out of solution in the fermentation
00:20:12.17 broth. This is a picture contributed by Amyris
00:20:16.01 when they took their highly engineered yeast,
00:20:19.17 it actually precipitates, secretes the artemisinic
00:20:22.16 acid outside the cell, and it precipitates in the
00:20:24.27 fermentation broth. And this gives you the perfect
00:20:27.11 purification. Now, the yeast that we delivered
00:20:31.14 to Amyris was not ideal. It produced artemisinic
00:20:36.24 acid, but not at a high level, and it grew slowly
00:20:38.28 because of the toxicity. And through engineering
00:20:41.23 the P450, as well as the CPR, and controlling
00:20:46.02 the electron transfer between the CPR and the
00:20:48.22 P450, Amyris was able to increase the production
00:20:52.08 and give us a host that's highly productive.
00:20:56.08 That host is now being used by Sanofi, producing
00:21:01.18 artemisinic acid in large scale in Eastern Europe.
00:21:05.16 That artemisinic acid is then converted into
00:21:09.15 artemisinin, using a light catalyzed process.
00:21:12.05 Much like we think occurs in the plant, where
00:21:15.01 light catalyzes the conversion of artemisinic acid
00:21:18.03 to artemisinin. And this is the full-scale facility
00:21:21.03 that Sanofi built in Garessio, Italy, where they're
00:21:24.02 producing the artemisinin. That artemisinin,
00:21:27.07 shown here pallets of it, is then shipped to
00:21:31.15 Morocco, where it's tableted and there is the final
00:21:34.11 product, right down there. 3 million treatments
00:21:37.21 had been delivered to Africa as of about January,
00:21:42.10 by May of this year, 16 million treatments had been
00:21:46.17 shipped to Africa. And Sanofi will have the capacity to
00:21:50.06 produce between 100-150 million treatments on an
00:21:53.22 annual basis. That's roughly one half of the world's
00:21:56.27 needs. Now, what I've told you is a story about
00:22:01.01 producing artemisinin in yeast and E. coli.
00:22:04.10 There are many other products that we might want
00:22:06.10 to produce. Things like fuels. Now, as I mentioned
00:22:09.20 earlier, if you produce fuels using engineered organisms
00:22:13.21 and those fuels are derived from renewable resources,
00:22:18.09 like sugars from biomass, the carbon that you put into
00:22:21.29 the atmosphere that the plants took up in producing that
00:22:25.14 biomass. So you have this possibility of producing carbon
00:22:28.13 neutral fuels. Well, we had an organism engineered
00:22:33.13 that could produce hydrocarbons. Now ethanol is one
00:22:37.29 of the fuels that's most widely used, and it's commercially
00:22:40.20 produced in the U.S. and supplemented into gasoline.
00:22:44.00 But it can't substitute for diesel and jet fuel.
00:22:47.18 We need a fuel that behaves exactly like the fuels
00:22:50.14 we get from petroleum. Well, we had in our hands
00:22:54.08 a host that would produce hydrocarbons, and we thought
00:22:56.25 we could just convert that into a fuel producing host.
00:23:00.11 Here are some of the fuels that could possibly be
00:23:04.04 produced using an engineered organism and they
00:23:06.11 include short-chained hydrocarbons that you'd find
00:23:09.06 in gasolines, long-chained hydrocarbons that you'd find
00:23:12.20 in diesels and jet fuels, and even aromatics that you find
00:23:15.10 in all fuels. So we went about engineering that yeast
00:23:19.21 and E. coli that we had first engineered for producing
00:23:22.03 artemisinin, to now produce hydrocarbons that would have
00:23:25.03 similar properties to the hydrocarbons that we get from
00:23:27.18 petroleum. And that engineering involved replacing the
00:23:31.21 genes that were specific for producing artemisinin
00:23:34.10 with genes that were specific for producing the hydrocarbons
00:23:37.02 that we'd use as fuels. Now, one great thing about these
00:23:42.05 fuels, which is very similar to the story I told you
00:23:44.29 about artemisinin, is that they are pumped
00:23:48.07 or leak out of the cells. And shown here is
00:23:52.14 a tube that had E. coli growing in it and you can see
00:23:57.04 at the top, there is an oil film at the top of that tube,
00:24:00.15 and then shown in this picture are E. coli swimming
00:24:03.15 in their broth. And you can see the blebs of fuel
00:24:07.02 that are being excreted and are collecting. Now this
00:24:10.14 makes the purification process for these fuels
00:24:13.22 really simple. You grow the cells up on the sugar,
00:24:17.04 they produce the fuel, excrete it, and it floats to the top.
00:24:20.16 It's a lot like making a vinaigrette, the oil floats to the top,
00:24:24.26 you skim it off the top, and if you've engineered the biology
00:24:28.00 right, you can put that into your tank and drive away.
00:24:30.12 Now I've talked to you about fuels and drugs, I want
00:24:34.23 to just mention the possibility of engineering
00:24:38.27 microbes for producing all of the other chemicals
00:24:41.09 that we now get from petroleum. Things like mouthwash,
00:24:44.06 perfume, carpet, and paint. As I mentioned earlier,
00:24:47.15 about 15% of a barrel of oil goes into producing
00:24:51.09 all of those products. And yet that 15% of the barrel of the oil
00:24:55.16 generates as much profit for companies as the other
00:24:59.07 85%, because of the higher value of those commodity
00:25:03.08 and specialty chemicals. But again, we have to produce
00:25:06.25 products that behave exactly like the products that we get
00:25:11.02 from petroleum, not things that are kinda like it.
00:25:14.00 But, shown here are just a variety of the products
00:25:17.08 that we currently get from petroleum. And this is
00:25:20.07 a pared down list. The list is vast and extensive.
00:25:24.17 But we can use a lot of the same synthetic biology
00:25:27.12 that I talked about earlier. Engineering the microbe
00:25:30.05 to produce identical precursors to the ones you get
00:25:34.04 from petroleum that could then be incorporated into all
00:25:38.05 of these products. And there we have a renewable
00:25:41.07 economy. Now, one of the challenges that we have in
00:25:46.16 engineering biology to produce all the products we want
00:25:49.10 to is that it's expensive and time-consuming. That
00:25:52.03 project I described about engineering yeast to produce
00:25:55.17 artemisinic acid, that took $25 million just for the research
00:25:58.26 and development alone. And about 150 person years
00:26:02.20 of work to get it accomplished. We need to be able to do
00:26:05.23 these projects with a tenth or a hundredth of that if it's
00:26:09.03 ever to compete with petroleum processes.
00:26:11.26 Why did it take so long? Well, just to engineer
00:26:16.06 that yeast to produce artemisinic acid, we needed
00:26:18.16 over 40 components, the genes, all the regulatory
00:26:21.06 switches. We had no standards for those connections,
00:26:24.26 no accurate models, and no details could be ignored.
00:26:28.24 And what's more, there was no BioShack to go out to,
00:26:31.12 to buy any of those components. Biological engineering
00:26:35.17 is slow because we design something, we then go
00:26:39.00 into the laboratory, we build it, and then we test it.
00:26:41.19 And hopefully, we learn from our mistakes.
00:26:43.28 But one turn of this cycle can take weeks, if not
00:26:47.26 months. And many years ago, it took a year
00:26:51.04 or more to go around this cycle. We need to turn
00:26:53.29 the crank faster and in higher throughput, so that
00:26:57.09 eventually, we can design once and have it be correct
00:27:00.19 after the first design. What's more, companies don't share
00:27:05.07 when they've gotten good at this process. Amyris became
00:27:08.27 good at engineering artemisinin, Dupont became very good
00:27:13.17 at engineering 1,3-Propanediol, but when they become
00:27:16.04 good at that, that's a trade secret. And they keep it
00:27:18.08 in house and don't share it. So the timelines don't
00:27:21.00 get shorter and shorter as we hope they could.
00:27:23.28 We really need the biotech industry, the next
00:27:27.08 generation of biotech industry, to look more like
00:27:29.06 the microelectronics industry, so that we can have
00:27:31.04 really powerful things like we now have computers
00:27:34.06 and cell phones that won't cost as much to develop
00:27:37.05 as they currently do in biology. And so, I'm advocating
00:27:41.17 for a next generation of biological foundry, just like we
00:27:44.23 have silicon foundries, we need to have biological
00:27:47.12 foundries that will allow any user to come in, design
00:27:51.12 on the computer what they want, send it off
00:27:53.25 to a foundry to be build, and come back to them
00:27:56.07 and work as planned. This is my goal and something
00:28:00.21 that I think the future holds. I think there's a lot of
00:28:04.03 promise in this. And it's something we have a head start
00:28:08.21 on. I'd like to acknowledge all the funding and the work.
00:28:13.05 There was a lot of work from people from my laboratory,
00:28:15.27 Amyris, and Sanofi for the development of the semisynthetic
00:28:20.18 artemisinin. The work at jbei and synberc for the fuels
00:28:23.20 and chemicals. And I'd like to thank all the funding from
00:28:26.28 the Bill & Melinda Gates Foundation, the Department of
00:28:29.14 Energy, and the National Science Foundation for all the
00:28:31.27 work this enabled. And I'd like to thank you for listening
00:28:34.18 to this video.

This Talk

Speaker: Jay Keasling

Audience:

General Public
Educators of H. School / Intro Undergrad
Student
Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: August 2015

Talk Overview

Dr. Jay Keasling discusses the promise of engineering microbes to create carbon-neutral products for a range of applications, including fuels, chemicals, and drugs. Keasling discusses the application of engineering microbes to the development of a microbial platform for the synthesis of the anti-malarial drug, artemisinic acid. This platform has helped stabilize the global supply. He also discusses additional applications of engineering microbes to fuel production, as well as some of the current challenges and possible solutions facing the metabolic engineering community.

Speaker Bio

Jay Keasling

Jay Keasling is a Professor of Chemical engineering and Bioengineering at the University of California, Berkeley, an Associate Laboratory Director for Biosciences at Lawrence Berkeley National Laboratory, and also chief executive officer of the Joint BioEnergy Institute (JBIE). He received his PhD in Bernhard Palsson’s lab at University of Michigan and was a post-doc with… Continue Reading