TBP01x Best Practice 2b Semi synthetic artemisinin production
Welcome to the second talk on this unit on semi synthetic Artesimic acid production
You remember the full title of the presentation is developing the most important anti
malarial drug using synthetic biology and fermentation technology.
Also you remember where we left of in the last talk. Should we work on E.coli or yeast or
develop a whole new chemistry. So let me tell you what we did.
We had a six month competition with two groups. We set up an E.coli group and a yeast
group. And the two groups competed against each other. The question was which team can
make the highest concentration of artemisinic acid in their fermentor. And this was to be a
schedulable industrial process. In other words the fermentation process we developed
producing artemisinic acid in either E.coli or yeast had to be capable of being scaled up to
ten of thousands of liters to hundred of thousands of liters scale in the real world. The
competition was to be judged by a panel of the world leading scientist in industrial
biotechnology after 6 months. So coming to the end of six months.
Prelude to decision day anticipation as running high. Fight night Amyris biotechnologies as it
was then called. E will grow on anything coli versus saccharomyces the evolved cerevisiae.
Free event, drink and appetizer provided. So what happened?
Well yeast won the competition. The E.coli could not make artemisinic acid at 30 degrees C.
It could make artemisinic acid at 20 degrees C, but isn t industrial scalable, you can t run
industrial fermenter at 20 degrees C, cooling costs even if it were feasible would be
prohibited. Yeast on the other hand produced 2.5g/l artemisinic acid at 30 degrees C.
And that is shown on the graph here. We were able to produce just over 2.5 g/l artemisinic
acid. With a minimal amount of Amorphadiene. But let me remind you of something.
Our target was 25 g/l of artemisinic acid, and we were only producing 2.5 g/l. So yeast won,
but the production still needs to increase by 10 fold.
So we developed a strategy to produce 25 g/l of Artemisinic acid in yeast. The first focus was
on amorphadiene production. We had two tracks of development. Strain engineering using
synthetics biology to engineer the yeast. And fermentation development to get the best
production out of the engineered strains. Once the Amorphadiene production target had
been reached, the 25g/l of Amorphadiene, we were going focus on increasing Artemisinic
acid production to reach the target of 25 g/l.
So let me introduce you to our strain of yeast. We build rebuild the entire terpenoid
metabolic production pathway in a new yeast strain, the original yeast strain was called
S288C much beloved by academic scientist. Most of the scientific literature on yeast
concerns this S288C. But as S288C did not have a known track record of used in an industrial
fermentation. On the other hand there was an other strain of yeast called CEN.PK2. Which
had some very desirable properties to been demonstrated. As an aside the work
demonstrating the superiority of CEN.PK2 in industrial fermentation was carried out by the
Technical University of Delft. Who are running this course that you are watching. The
connection between my company in California and Technical University of Delft of why I can
give you this talk is CEN.PK2. So moving forward with CEN.PK2 we called the strain yeast 2.0.
Let me show something about yeast 2.0 Amorphadiene production. We had a production
target of 25 g/l with the succession of fermentation and engineering improvements to the
strain we were able to produce over 40 g/l of Amorphadiene. And that is shown on this slide
here. This is in fermentation. So our Amorphadiene production target was attained.
So what we need to concentrate now was on conversion of the 40 g/l Amorphadiene in to
25 g/l Artemisinic acid. And how we going to do this.
So let me show something about Artemisinic acid production in yeast. This graph shows you
production in shake flasks experiments. The yeast 1.0, the original S288C produced just over
250 mg/l Amorphadiene, and the yeast 2.0 produced great improved, about a five fold
improvement of Amorphadiene production, on the other hand Artemisinic acid production
was low in yeast 1.0, but only increased slightly in yeast 2.0. It just about doubled. So in
someway the Amorphadiene production increases quite dramatically, but the Artemisinic
acid production barley changes. So we knew that we had to concentrate on the biological
oxidation.
Which is summarized here. This is what occurs in the plant and this is what we are having a
problem with in our production yeast.
So something I haven t told you is that in fact there two chemical intermediates between
amorphadiene and artemisinic acid. Intermediate 1 is an alcohol, intermediate 2 is an
aldehyde, and one conversion enzyme p450 did al three conversion. And we obviously
needed to work on this conversion.
What we did is we went back to the plant, Artemisia Annua, which produces Artemisinin and
investigated the conversion of Amorphadiene to Artemisinic acid in the plant. As a result of
our work on the plant we were able to ascertain several things.
And this shows what we did to the yeast to improve production. We altered the production
of the conversion enzyme in the yeast. But also we discovered there were three new plant
enzymes that we did not have before. We added a new enzyme to the first conversion. And
we replaced the second conversion with two completely new enzymes. So three completely
new enzymes from the plant were expressed in the yeast.
The new enzymes dramatically improved production in yeast. This shows the results of
shake flask experiments. And you can see in the dark blue the increased artemisinic acid
production The original strain produced a certain amount of Artemisinic acid our first change
in expression actually decreased it slightly but the yeast was much healthier. And then you
can see as we successfully added the other enzymes production increased and increased
until eventually all three enzymes and the changed expression we were producing almost
exclusively Artemisinic acid.
So how did this do in fermenters? Well it made a huge difference. As our 25 g/l target, and
you can see with a succession of genetic changes and engineering changes, our industrial
production target at 25 g/l of Artemisinic acid was obtained.
So the situation in 2014, as I am taping this. Sanofi is manufacturing with Amyris s yeast
strain. 2013 they made 20 million cures. In 2014 they made 120 million ACT cures. This is
about one third of the world supply of ACT s. This shows you some of the material that
came out when Sanofi first started production. We published a scientific paper in Nature
detailing all the work I just described to you. There were various press publication: Sonofi
launches malaria drug; Malaria drug makers see the light. Referred to Sonofi photochemical
process that I haven t talked about for the conversion of Artemisinic acid to Artemisinin.
The most important point is that semi synthetic Artemisinin is saving lives. Sonofia are now
distribution ACT s containing semi synthetic Artemisinin to multiple countries in Africa,
saving lives, mainly kids.
If you like to learn more about Amyris, the company I work for. You can go to our website:
www.amyris.com. And this shows three of my colleagues in the lab.
Over the course of this work, probably a 100 people have worked on this. In various partners
of Amyris, PATH, Sonofi and UC Berkley. This just shows many of my colleagues at Amyris.
Most of who contributed to this work. Thank you.