J. Craig Venter on Synthetic Biology at NASA Ames

Uploaded by nasaames on 12.01.2011

>> Dr. Pete Worden: Now he is one of the most frequently cited scientist.
Heís the author of over 250 research articles.
Also the recipient of numerous honorary degrees, public honors and
scientific awards including the 2008 United States National Medal of Science,
The 2002 Gairdner Foundation, International Award in the 2001 Paul Ehrlich and Ludwig Darmstaedter Prize.
Heís a member of numerous prestigious scientific organizations including the National Academy of Sciences,
The American Academy of Arts and Sciences, the American Science Society for Microbiology.
Itís my very great pleasure to welcome you, Craig to
NASA Ames and to all these cool people that are thinking about great stuff.
So, we look forward to hearing what you have to say. Thank you.
>> Dr. J. Craig Venter: Well thank you very much Pete
for the kind introduction and the invitation to come here.
This is not too many things excite my imagination as the implications of
trying to design organisms, even people for long term space flight and perhaps colonization of other worlds
as we try and use some of these same tools to clean up our own environment.
I understand you all have a wide range of background so I thought I would get us all on a common wavelength
by putting things in the context of genomics in general
and just some of the kinds of things that Iíve had the privilege of doing.
Perhaps like NASA I get to ask big questions like defining life, trying to digitize it,
which is what weíve been doing for the last 15 years,
how extensive and diverse is it, can we get down to minimal components,
and then can we go the other way, can we start in the digital world and recreate life out of that digital world
not within the computer but outside the computer.
And thatís what we announced earlier this year with the creation of
a bacterial cell controlled by a chemically synthesized chromosome.
So we actually synthesized a million base pair genome starting with four bottles of chemicals.
It was actually assembled in the eukaryotic yeast then
we had to transplant it out of the yeast and to a recipient bacterial cell
where it converted that cell into an entirely new species.
We call that a synthetic cell because everything in the cell was derived from that synthetic chromosome
and all the traces of the original species completely disappeared.
So how did we get here, what sort of the various steps that happened?
And what weíve been doing since the sort of the mid 80ís is what I call digitizing biology.
As we read the genetic code, we go from this analog molecule into the ones and zeroes in the computer.
And so as we read different genomes including the human genome,
itís a key part of digitizing that biology.
Now we can go the other way, we can start with those ones and zeroes and go back.
So in the 1990ís, we developed very rapid ways for discovering genes,
the numbers of these grew quite substantially over time with the EST method,
just pulling out the expressed part of the genome.
Itís amazing these tens of millions of sequences,
most of them are from human trying to understand all the different splice variants(?) in our own genome.
In 1995, the big breakthrough was a new mathematical algorithm, not just sequencing tools
but a way to assemble all the sequences and so we had approached by breaking the DNA down into little pieces,
sequencing those pieces and then reassembling those computationally in the computer,
we were able to come up with the first sequenced genome of living organism in history.
That was only 15 years ago.
We couldnít get funding for that, government review process said this couldnít possibly work.
We had to use our own money to do it and after we showed it worked, we were given more
money then we knew what to do with all kinds of species
and we had good money from DOE, and NIH looking at diversity,
first in the microbial world and then expanding to plants and working our way up
to in 1998-1999 doing the fruit fly genome and then in 2000, the human genome.
So, the human genome has roughly, each of us have about 6 billion letters of genetic code because we have two sets of genomes,
one from each of our parents.
So you here just pair up numbers or do we have 3 billion letters or 6 billion letters.
And the first draft of this came out in ìScienceî just about 10 years ago.
So in February, weíll be celebrating the 10th anniversary of this publication in a special meeting in San Diego.
But things have progressed first slowly but now a lot faster since then.
So in 2007, we published the first complete diploid genome.
Actually I used my own genome because then I didnít have to go through getting complex permissionsÖ
and people said of course he used his own genome.
But when we started this project, people were totally afraid of genetics and genomics
and how could we possibly read somebodyís genome and put it on the internet.
Since I did that and published this, and put my genome in the internet, itís now become de rigueur to do that
and I think biology will proceed now in an open fashion instead of a closed fashion.
But looking at my two sets of chromosomes from my parents, they actually differed from each other by about .5%
which was much higher than all the announcements in 2000, how we differed out one letter out of genetic code from each other.
People just looked at homologous regions, looked at SNIPs,
the single base pair changes and so we only differ by 1 out of a thousand.
When you try to compare any two of us, weíre comparing four sets of chromosomes
and so thatís how we get up to 1 to 3% when we look at all the insertions and deletions
and all the changes other than just the single letter changes.
In fact thereís more rearrangements and changes in the genome, more base pairs involved, in structural changes,
insertions, deletions, than there are in the SNIP variation.
So, itís almost 10 times the variation I previously thought.
If you think about that, so 44% of my protein coding genes have
one or more heterozygous variants in the protein sequence.
So if we all have that same sort of percentage differences, to me itís more amazing that our biology
is closely similar than that things donít work on in every single one of us.
Most drugs work on about third of the population; they have little
or no effect on another third and have toxic effects on another third.
Thatís not surprising when you see these kinds of numbers.
So understanding that variation is going to be key.
When youíre looking at 44% variation with as many of,
you know, millions of changes, thereís an awful lot to look at.
Genomics has expanded very rapidly in the last few years.
Due to technological innovations, so what was a $5 billion government worldwide program,
that we forced to go a little bit faster,
now you can buy a machine about the size of this podium for a half a million
dollars and sequence a genome in one or two days and that cost has going down substantially.
So genomes are pouring into the databases from around the world,
so looking at that 1 to 3% difference, so hereís looking at a Han Chinese,
Gubi, one of the recent sequence people from Africa done by Vanessa Hayes at my institute
then comparing it as a Northern European Caucasian and you can see the degree of overlap.
What Vanessa did is looked at 3 different populations in Africa,
including Desmond Tutuís genome, and there was more variation
within Africa than between Gubi, myself and the Chinese individual.
We all evolved out of these populations in Africa so itís not surprising thereís more variation there.
But itís sort of turning peopleís thinking on its head. So, if you think about in your field,
some of the reading Iíve done and tried to follow over time, NASAís been doing genetic selection for a long time.
You just didnít call it that, because that seem to have a bad connotation
but people had to pass rigorous tests, they had to be certain sizes.
These are phenotypic selections. So why not get smart and actually really do it
and screen for the things that might be meaningful for allowing space flight?
The study is showing that some inner ear changes allow people to totally escape the effects of disorientation in space,
things associated with bone regeneration, DNA repair from radiation and on and on.
That, probably this list could be thousands of traits long. All biology works on selection, NASAís worked on selection,
measure a few more parameters and what youíve been measuring and you probably can get a better result.
If weíre going to have people travelling for their whole lives and even multiple generations,
we might want to think about engineering these and other traits to enable those purposes.
But weíre not alone even in our own bodies, we actually have more microbes than human cells.
So we have roughly a hundred trillion human cells,
each of you have about 200 trillion of bacteria associated with you right now.
Nothing personal.
But they can get very personal.
So weíre actually born without these microorganisms and we acquire them quite quickly
and the gene population exceeds our own gene population by orders and magnitude.
So think about right now, maybe the person next to you,
especially if theyíre coughing have about a thousand different bacteria in their mouths right now.
If you look at, weíre talking about maybe 10 million genes in the microbes associated with each one of us.
We donít really know what most of these do.
Thereís new studies now just coming out of it,
these were discovered using the tools we developed for sequence in the human genome, the shotgun sequencing.
So we can just take samples from different body cavities and sequence at once all the microbes that are there.
This should have been being done by NASA for years now.
Each new person that goes up in the space station is bringing
perhaps 10 million new genes, organisms, pathogens with them on that trip.
Weíve been doing environmental sequencing as Iíll show you in a minute
and some environments such as submarines and others,
and certainly Iím sure the space station create a very unique microbial habitats.
So to understand our biology, we have to understand our own genetic code,
we have to understand the genetic code and the extent of these microbes associated with us.
We have to understand the interactions with our immune system and then with the external environment.
So itís getting more complicated by the minute but the exciting thing for me is now we know what the parameters are,
at least we think we do.
So I think for the first time we actually have a chance of making some progress.
Instead of being ignorant that all these things exist,
we can know about them and even manipulate them and understand them.
Hereís the change in different cancers since 1975.
Esophageal cancer is the fastest growing one.
And if we look at the microbiome in these individuals with the esophageal cancer,
they have a whole unique microbiota associated with them.
Now we donít know yet is that causal or is that the result of the cancer?
It becomes important to determine, but obviously esophageal cancer people think is
clearly environmentally determined and these microbes are key part of that environment.
So what else does this microbiome do?
When we look at physiology, our biochemistry only allows certain things to be made,
our microbes provide a lot of that additional physiology so what is that microbial potential?
So if we have 20,000 somewhat genes, maybe you can get a hundred thousand different transcripts,
maybe 100,000 to 300,000 different proteins depending on splice variants etcetera,
but the best quantitation of our chemistry is
roughly 2,400 different chemical compounds that we can make enzymatically from our gene set.
So what happens with those?
So if weíre to measure your blood stream after a meal,
we find around 500 different chemicals circulating in your blood stream.
Only 60% of those are from human metabolism.
30% or so are derived from all those different species you eat during your meal.
But 10% or on the order of 50 chemicals perhaps circulating in your blood stream right now are bacterial metabolites.
We have no idea what role they play in human physiology; do they make you feel better?
Do they protect you from disease?
Do they cause disease?
Do they make you depressed?
All the above, any of those things, nobody has any idea.
We just know theyíre there now, and they can be readily measured so we need to measure the genetic code,
we need to measure our microbes,
and we need to know all these different chemicals circulating in our blood in different environments.
This would be great studies to do just on an existing space trips now, let alone try to understand for longer ones.
So for using our imaginations and I was asked to do that, why not come up with a synthetic microbiome?
If we had a way with antibiotics or a way to sterilize an astronaut
before going into space and providing them with a synthetically compiled community.
We could eliminate completely disease organisms; maybe have no dental decay for example.
From reading ìPacking for Marsî, I understand methanogens and sulphur producers are a major problem in space flight,
except for those who want to propel themselves around the cabin.
But I understand the physics of that doesnít really work.
Body odor is primarily caused by microbes.
The French have tried to cover up things with perfume but the best way
to eliminate smell from your armpits or other areas is to kill the microbes.
So if you use something like the 70% alcohol, you can totally eliminate some of those odors.
So, if we come up with the right set of microbes,
we can eliminate some of these perhaps olfactory and even health problems.
Why not add bac-cells that make specific nutrients or vitamins instead of having to get those from the diet.
Unique metabolisms, so for example if we were making an algae based food,
we could metabolize every part of that algae with
cellulose degradation to sugars perhaps retaining calcium better, etcetera.
So Iím sure many of you come up with far better advantages than I can with the controlled environment.
So if we look at environments, based Carl Woese, who started measuring 16 sRNA,
our view of the world expanded dramatically.
But it turns out even the 16 sRNA was missing things by over an order of magnitude
and just from simple experiments in my spaceship,
itís a 95 foot spaceship that weíve sailed around the world taking samples every 200 miles in diverse environments.
Weíve come up with a very different view.
We just finished sampling in the Baltic and the Mediterranean and Black Seas this last summer.
We just simply filter sea water, collect different levels of microorganisms,
then sequence everything thatís on the filters.
We donít even see the organisms; we donít know what they look like.
We know what their genomes look like from compiling all their genetic data.
So itís a simple apparatus, we have a different one for sampling an air.
Weíve been doing the air genome and looking in all kinds of diverse environments.
These are complex plots but basically what youíd expect if there was limited diversity
when we first sampled in the Sargasso Sea in early 2001 through 2003, we were told weíd only find a few microbes.
So what we can do is put a genome sequence across the top of this and then compare things,
each one of these little bars is roughly 600 base pairs of DNA sequence.
So, if there was a simple set of organisms and not much diversity everything would be up around 100% range.
Instead what we found was this incredible diversity,
where all these organisms have basically the same 16 sRNA sequence
so we thought there was not this kind of diversity underneath.
But instead of being a single organisms or eleven maybe 20,000 different related organisms.
In fact if we look at this broadly of whatís falling out what were thought to be basically
single organisms now are of the major taxa that weíre finding in the ocean.
Some of these were unknown before seeing these clouds of organisms.
So if people talk to you about a single organism from the environment,
itís basically a meaningless concept whether that environment is your gut, the ocean or the atmosphere.
Weíve also looked at deep sea microbes.
So this is the high temperature vent in the Pacific Ocean were Holger Jannasch isolated this organism.
This is the first archaea genome that we sequenced in 1996 and this is a complete anecdote,
it doesnít need any organic compounds.
It makes everything it needs from life from carbon dioxide and hydrogen as an energy source.
So, this is one of the many CO2 utilizers out there in the sense
inspired us to go in directions of capturing CO2 for energy production.
But part of a program we have would be peer looking for microbes deep on the earth to do unique metabolism of hydrocarbons,
we came up with the same level of diversity deep on the earth that we find in the oceans.
But something is very different about them; perhaps because a mile deep theyíre shielded from radiation,
they certainly donít get any UV radiation there.
Instead of seeing these huge clouds going down to 50% diversity,
you can see things are clustered much more like people expected initially to find in the oceans.
Itís a much smaller set.
We have the diversity in terms of different types of organisms but the depth from all the mutations
you might see are caused by UV radiation in the oceans doesnít occur deep in the Earth.
So we have a whole range of different types of organisms that could be captured.
Now, if you want to discover a more mammalian genes we have to hope
and that thereís going to be strange mammals on different planets
because thereís no point in sequencing more mammals on Earth to discover new genes thatís basically saturated.
But if weíre looking at viruses, bacteria, archaea, weíre still in the linear phase of discovery
even though weíve exceeded 50 million genes now in our databases. Itís still growing exponentially.
We can take a sample anywhere in the world out of an aqueous system and majority of genes will be new in that sample.
So minimal life is something that came out of our early studies.
In fact, inspired by some NASA scientist.
When we sequenced this genome, itís the second one we did in 1995.
This has the smallest genome of a self replicating organism with only about 482 protein-coding genes and 43 RNA genes.
Now this was roughly around the time when some NASA scientist claim they found nanobacteria in some Martian meteors.
So there was a lot of discussion then about minimal life what it could be.
Turns out the volume of those so called nanobacteria were so small,
you couldnít even get a tiny piece of DNA or RNA in them so I think everybody pretty much concluded those were artifacts.
These are much larger cells.
Weíre trying to understand minimal life.
So we had two genomes, and we asked how many of these genes are essential for life,
whatís the smallest number of genes required for cellular life,
and ultimately could we design and construct such a minimal genome?
We took a variety of approaches, comparative genomics in the computer.
Trying to knock out genes to see which ones were essential and
we realized we could only get there by making a synthetic chromosome.
When we looked at the first two genomes that we sequenced and this was a study out of NIH from Kooninís lab,
you can see a pretty small overlap between the first two genomes.
They actually concluded that gene diversity in our planet must be really small to have this extent of overlap.
If they waited 6 more months till we had the methanococcus genome they would have come to totally different conclusions.
So comparative genomics could only take us so far.
Clyde Hutchison at the institute developed this technique called whole genome transposon mutagenesis.
So transposons are these small pieces of DNA that jump around in the genetic code.
Over half of our human genome is composed of these transposons.
Theyíre constantly jumping around and if they jump in to the middle of a key gene,
we can get a disease or next generations wonít exist with these.
But because we had the sequence of the genome,
we could put in these transposons which go in randomly and then
you can sequence off those and know exactly where they went in.
So we were able to over the years develop this map of the Mycoplasma genitalium genome.
Every place you see one of these little triangles thatís where a transposon randomly inserted in the cell.
So if the cell could live with one of these transposons in it,
we define that gene that it was in as a non-essential gene.
You know, see thereís some little bars on here with no transposons in them;
basically we define those as essential genes.
But the trouble is the term essential and non-essential is totally context specific.
It turns out this cell will grow nicely on both glucose and fructose and thereís a gene for the transporter for each one.
If you have both sugars in the media and you knock out the glucose transporter,
the cell keeps living and you say, well that glucose transporter must be a non-essential gene.
But if youíre only growing the cells on glucose and you knock off the glucose transporter gene, the cell dies.
So we can only define the genetics, including our own genetics, in the context of the environment that itís in.
And I think thatís an important concept when you think of these confined or limited environments.
When we look at the metabolic map of the cell and then look at all the genes that could be knocked out one at a time,
we decided this would probably not lead to a viable cell.
So it turns out thereís genes that cover the duplicate function,
you know, they are back up systems like NASA likes to have and so if you knock out one gene,
it doesnít really tell you whether thatís an essential function.
So after spending years doing this,
we decided the only approach was to make a synthetic chromosome
where we could control completely the genetic content.
So then we had new technical questions,
would the chemistry even permit us to be able to make these large pieces of DNA and if we could,
would we just have a large inert chemical or could we boot it up?
So jumping ahead to 2003, we tried a number of approaches.
We underwent some very critical ethical review that
Iíll get back to and we developed some techniques for error correction.
So the DNA synthesizers are not great machines.
They create errors in the DNA as they make one.
Itís an N minus one situation, the longer the piece of DNA you make the more errors.
And because of that, we either need error correction methods like we published here.
And what we did here, we started with the viral sequence in the computer.
We made these small pieces of DNA that we assembled together to make the whole genome.
And the exciting phase came, we inserted this piece of inert chemical in to E. coli
and the E. coli genome system started reading this piece of DNA. Started making all the proteins.
The protein self assembled to make the virus and the virus showed its gratitude by killing the cells
that made it which is how we detected with this clear plaques on a plate.
So we call this a situation where the software is actually building its own hardware.
All we did was put in a chemical piece of software and that
led to making this a physical structure that has biological activity.
But we didnít want to make just a small virus.
We wanted to make an entire bacterial chromosome and there were two aspects as I said,
one, could you boot up the DNA,
it was easy with the viral DNA in E. coli but we didnít think
it would be so easy with trying to boot up an entire bacterial genome.
So this study in 2007 that we published on booting up and transplanting the genome.
I think thatís actually one of the most important ones our teamís ever
published because we actually by changing the genome in a cell,
we completely converted one species into another.
And it seems like alchemy or something to many people until you really understand
the importance of DNA and the importance of genetics and how life actually works.
Because this is so important, I thought Iíd walk you through it.
So we isolated the DNA from M. mycoides.
These are a very simple cells with just plasma membranes.
We needed to know, for example were proteins required to do transplantation because if weíre just making chemical DNA,
we need to know if proteins were involved so we treated it harshly with proteinases and removed all the proteins.
We added a few gene cassettes so we could select for the chromosome,
and they would turn themselves bright blue if it got activated.
And we worked out ways to insert that genome into a related cell, M. capricolum.
And we thought about this for a long time.
We thought we would have to eliminate the chromosome in the recipient cell before we put in the new one.
And we worked on a lot of ways to do that with radiation damage,
chemical damage and finally after trying a number of things,
we decided maybe we donít have to do that,
and we could use the enzymatic systems in the cell themselves to do this for us.
So we have this very sophisticated movie to show you what we think happened.
So we inserted the new chromosome in the cell and for a brief period
of time now we have a capricolum cell with two different chromosomes in it.
As with the viral piece of DNA, the cell system started reading the new chromosome and started making proteins.
Some of the early proteins that are made are restriction enzymes.
The restriction enzymes that were made, recognized the capricolum chromosome as foreign DNA and chewed it up.
So now we have a capricolum cell with the information system the chromosome from M. mycoides.
In a very short period of time, we had these bright blue cells and when we looked at these cells,
all the characteristics of the capricolum species were gone.
All the proteins that existed in the cell were those coded for by the M. mycoides chromosome.
So simply by changing the software, all the characteristics of
one species went away and we had an entirely, a new one coded for.
So we knew now we could do transplants.
So jumping ahead to 2008, we had teams working diligently on the chemistry to make these larger pieces of DNA.
We knew we could make viral pieces accurately so we thought if we made a series of this viral size pieces,
we could perhaps assemble these with homologous recombination.
And thatís where a study of biology certainly helped us with different systems.
So we made 101 of these cassettes that were 5,000 to 7,000 letters each.
And then we went through this assembly process of assembling these pieces together on the lab bench,
first at the 6kb range then going up to 24kb.
And in each stage we cloned these pieces in a coli and sequenced them trying to make sure it was really a valid process,
but and we kept going until we got up over 100000 base pairs.
An E. coli would only take 2 of the 4 pieces,
so we started looking around for a new system as Mike Montague said earlier and we settled on yeast,
because not only did it happily clone these larger pieces,
the homologous recombination system and yeast assembled those.
Now, weíd spent years studying DInococcus radiodurans.
So, this is one of the earliest genomes we sequenced with the DOE.
This cell has 4 different DNA elements, 3 chromosomes and a plasmid
and it can take up to 3 million rounds or radiation and not be killed.
What happens is you get a couple of hundred double stranded brakes that chromosomes literally get blown apart,
but if itís in an aqueous environment, 12 to 24 hours later,
it reassembles its chromosome and the cell starts replicating again.
It could be pretty nice for space traveller if humans could do that,
but itís a much more complex equation where 6 billion letters of genetic code other than a few million.
But we thought we could use these processes and we expended several postdoc year lives on this,
but never got it to work outside the cell and so weíre delighted that we could jump ahead
and use simple brewerís yeast with itís powerful homologous recombination system to do this.
So, we were able to put just the 4 quarter molecules with proper overlaps and this simple vector.
And what this vector has in it, is an artificial yeast centromere.
And so just adding a centromere, a eukaryotic centromere to this bacterial clones,
yeast assembled all these immediately into the entire bacterial chromosome and thatís what we reported on 2008.
And that was the almost 600,000 base pair genome sequence
assembled from 4 bottles of chemicals and using this assembly process.
The trouble is, as Michael said, these cells grow extremely slowly.
We have still not been able to boot up this chromosome because we think in the
6 weeks that it takes to do that the selection processes arenít adequate.
We also discovered a few other things.
Some of these cells have nucleases on the cell surface and they just chew up the DNA as fast as you expose it.
For jumping ahead, we had to solve a number of problems.
Michael mentioned how we could throw in smaller pieces in good assembly and so Dan Gibson
who did all this work wanted to see if we could just throw in tiny DNA fragments in to yeast and get assembly.
So we could just put all simple nucleotides in to yeast and it will assemble those nicely into larger pieces.
But the real breakthrough that Dan came up with after studying these reactions is
this very simple single pot chemical reaction that actually allows us now to automate all these processes.
Itís just three enzymes, one that chewes back the DNA,
another one that ligates it together and then fills in with this fusion polymerase.
So itís a one step reaction at 50 degrees centigrade.
All you do is put in the synthetic pieces of small DNA and it assembles them into larger pieces.
We can go from the digital world to making an entire analogue of molecules potentially even booting them up.
Danny Hillis and I are talking about trying to build a robot,
that's a self-learning robot system that could do these experiments
and learn biology a thousand times faster than any scientist can.
We have a lot of biology to learn,
we don't know what most of these 50 million genes weíve already discovered is.
But if we can automate these processes going from the digital world into creating new life forms,
we have a chance to learn a whole lot faster.
So our problem was, we were assembling the bacterial chromosome inside a eukaryote.
To do the transplants, we have to find a way to isolate the DNA from the eukaryote and get it back into a bacteria.
And Gwen Binders at the institute cloned entire chromosomes in yeast, adding the simple yeast centromere.
So that gave us the ability to these trial experiments.
But we ran into a problem.
It didn't work.
We couldn't take the chromosome, that native chromosome out of yeast and transplant it.
So it took our team of roughly 25 scientists two years to solve this problem.
It turns out that the DNA, when we isolated it from the bacterial cell was methylated.
And that methylation protected it from the restriction enzymes that the Capricolum cell had.
The genome was using its own restriction enzymes to destroy that enzyme but it was getting destroyed first.
So if we purified the specific methylases and methylated the DNA,
we could then readily do the transplants out of yeast into the bacteria.
So we actually have this circle that allows us to make very rapid changes now genetically.
And for those of you who work with microbes, some of the biggest limitations
working with microbes is that they don't have genetic systems.
So simply isolating the chromosome from the microbe, putting it into yeast,
we can now modify that chromosome using a whole repertoire of yeast eukaryotic genetics.
We can then isolate it, methylate it if necessary, and transplant it into a recipient cell forming a new species.
And we can go around this circle very rapidly.
So we made the decision early on because of the problem with the slow growth of the microplasmic genitalium,
to make a leap, knowing that we could transplant the mycoides genome,
to resynthesize that genome, even though it was a much larger project.
And initially we thought DNA synthesis was going to be the limitation of the biology.
So this is what we reported this spring and so the process was again starting with these all of the nucleotides
but now using this new single-part assembly method.
So we could start with 1-kb pieces.
In fact John Mulligan, who is here, made all those for us at Blue Heron.
To speed up the process we then took ten 1-kb pieces; put them together and made 10-kb pieces.
We then took ten of the 10-kb pieces together and made 100-kb pieces.
And then there were eleven 1100-kb pieces that we put together in yeast
to assemble this entire million base pair chromosome.
At this stage I was totally certain that it was now just a matter of simply doing the experiment.
And I boldly predicted that we would have the first synthetic species by Christmas last year.
Obviously I was wrong.
For some reason we could never get a living cell out of it.
So just like software engineers have proof-reading software, we had to developed DNA proof-reading software.
And what we did was actually made naturally occurring 100-kb pieces, so we could substitute those.
And so we could get ten synthetic pieces and one natural one, and we could boot that up.
So we knew there was a problem in this one piece.
And part of the problem is that the new sequencing technology is not as accurate as the old Sanger sequencing.
So even though we'd sequenced that, it couldnít find this one base pair deletion.
So one error out of a million base pairs and we got no life.
So we re-sequenced it with Sanger sequencing, and found the single base pair deletion in the central gene.
Then remade the piece, and booted it up, and hereís the complete map
and hereís the cells that resulted from the transplant.
Let me go back a minute ëcoz we did some things that when you think about the problems you could have with serum,
you could fool yourself and fool others.
Our biggest concern was a single molecule contamination of the native genome.
With living cells we could think that we actually had made a synthetic genome activated.
That would have been a contaminant.
So we started this concept of watermarking the DNA.
And the first genome we made we just signed our names in it and people thought that was very unimaginative.
So we got a little bit more imaginative with this second genome.
And Mike Montague and Harold Smith and Clyde Hutchison developed a new code within the code within the code.
So in the first watermark is the code for actually translating DNA into English
or English into DNA with complete punctuation.
A large number of scientists have now solved this code. And there's an email address built into the genome.
So they solved the code and sent an email to the web address proving that they had adequately decoded it.
But once you decode that it tells you how to read the rest of it.
We have 46 names of all the different scientists that have been involved in this project.
And we tried to get creative and add a few quotations from the literature.
So we had one from James Joyce, one from Oppenheimer's biography and one from Richard Feynman.
And just to show you can never get away for free with anything,
after this was published we got a phone call from James Joyce Estate
saying that we hadn't sought his permission to use this quotation.
And you know I know weíre, we have powerful techniques but I didn't know quite how to do that 'coz he was dead.
But... So all these is built into the genetic code.
I think the chances of this occurring naturally is pretty close to zero.
So the difference is we can insert all of this, what would appear to nonsense DNA.
In fact, part of our code is to put frequent stop codons into it
so we don't introduce new biology in to the cell by making new fragments.
On the other hand, if we have one error in the central gene, you get no life.
So where it is in the genome and what it is is obviously very critical.
This is the map of the whole genome.
So unlike our genetic code which is only about three percent of our genome codes for protein coding genes,
and this, itís well over 90 percent.
You can see, there's not a lot of gaps between the genes.
So it's a much more efficient system.
Again, when we checked, there were no capricolum proteins left.
It was just the proteins made from this modified genome.
So this is the size range thatís happened over this period of time, now being over a million base pairs.
These techniques are so robust Dan Gibson reassembled the genome for each experiment
instead of trying to use a clone variety of it.
So they're truly robust and now theyíre able to be automated.
I think we're gonna enter into a new era.
So I like to think of all these genes weíve discovered today as design components.
In the electronics industry, people in the 40s and 50s had far fewer design components to work with.
By the time we finished characterizing life on this planet,
this number could be two or three hundred million unique genes or genes that are part of complex gene families.
We actually have software synthetic genomics for designing software of life
to create new organisms where we can modularly build in the type of metabolism.
Is it going to be metabolizing sugars?
Is it gonna take to CO2 to methane etcetera as building a backbone to try and design future organisms?
Because there is so much gene diversity and so few scientists on this planet,
we have to come up with new combinatorial approaches to make some rapid progress here.
So just think if you have a metabolic pathway with only ten genes in it,
and if you have ten versions of each of those ten genes.
That's ten to the tenth combinations.
It would take forever to get there.
So we're trying to build this robot that could make a million chromosomes a day just for one or two scientists to work with.
And if it really is self-learning, those scientist are probably just gonna watch the robot work like it happens with robotics.
And then it gets down to what with all biology has been selection.
Can you set up the right assays for selecting what you want out of it?
So, some of my different audience, hopefully not this one ask why do this?
But there's a lot of different reasons.
Obviously burning all these fossil fuels were weíve exceeded the equilibrium of CO2 capture on our planet.
The oceans are the largest sink.
This number keeps changing I think it's up to 3.8 million tons of new CO2 in the atmosphere.
We're making more people faster than we can provide the means to feed them and provide medicine and housing and clean water.
We're at 6.8 billion now.
Within 35 to 40 years we'll be over 9 billion.
Like any number, I like to put it in context.
So I was born in 1946.
There's now three people alive on the planet for everybody that existed the year that I was born.
Soon there'll be four.
All in almost a single generation do we have these huge changes.
We can't provide all means for feeding the ones now.
So weíd need new approaches for the next generations.
When we look at plants, plants are really not very productive systems.
They're pretty limited.
So we, like others, have looked around and even looking at kind of modest numbers ëcoz we have to greatly exceed
ten thousand gallons an acre with microalgae to make it truly cost effective for making fuels.
It's orders of magnitude better than any of the plants systems.
We've been working on oil palm and jatropha and others.
Look at the bottom of list - corn.
We have an economy in this country based on trying to make ethanol from corn.
It's only because of thereís a corn lobby, not coz it's a smart thing to do.
So we have to try and move in a different direction.
So we've been trying to work on designing what we call 4th generation of fuels and cells,
where sunlight is the energy source and CO2 is the carbon source.
We can almost go in any direction from CO2.
You can make materials.
You can make food.
You can make fuel.
You can make unique chemicals.
You can make proteins.
And so we've been working on this and our team led by Paul Rustler had a really nice breakthrough in the lab.
By changing some genes and some enzyme systems,
instead of treating algae growth like farming of growing up a lot and trying to squeeze the oil out,
we got the cells to pump the oil out of the cells on a continuous basis.
Here's a cell that makes pure C8 and C10.
By changing anything along the pathway, we can make any size lipid.
This is one of the main reasons why Exxon put 600 million dollars on the line to work with us
to try and scale up the production of carbon compounds from CO2.
I don't say fuel because the goal is just to create a biocrude
from the algae to go into the existing refineries to make gasoline,
diesel and jet A fuel, totally consistent with the existing infrastructure
instead of trying to adapt to burning a different type of lipid.
So we're making some good progress along these lines.
But to get to the billions of gallon scale, these are not short-term projects.
So we think the soonest there would be anything on a substantial economic scale will be ten years.
But itís progressing.
Itís where we need synthetic genomics, synthetic biology, cell engineering to take over.
Because it wouldn't make sense for an algae to evolve to produce as much hydrocarbon as we need from CO2.
So we have to change evolution.
We have to take over.
Weíve looked at thousands and thousands of algae strains.
And there is nothing within an order of magnitude naturally to get where one needs to be.
People have obviously looked at algae for producing food for a space flight and other processes.
It's pretty inefficient in terms of what was done before.
So using natural algae like this, I think it would take
a pretty large volume just to produce enough to feed a single astronaut.
Going up exponentially in a production scale and engineering these cells to produce different substances.
I think itís totally within the realm of the next few years.
Ken Nielsen at the Institute has a great electrobiology group thatís working on microbial fuel cells.
That, microbes just naturally select the anode or cathode and we can take complex mixtures including raw sewage,
generate electricity and convert that into close to drinking water.
Nobody wants to do that final experiment yet, but it takes it a very long way and itís in part due to understanding this.
The teams discovered that bacteria actually make these nanowires that can live off of metal surfaces,
pulling in electrons out of the metal and using those for metabolism.
We just announced formation of a new vaccine company, to use these synthetic approaches to very rapidly make vaccines.
This is based on 15 years of work weíve had with Novartis on the new meningitis vaccine,
which is the first genomic-based vaccine and just finished phase-3 clinical trials in Europe.
Meningitus B is one of those diseases that by the time you diagnose it in young people, it is too late.
They are dead shortly thereafter.
So vaccine is the only preventative approach.
And now we are applying this to making very rapidly new influenza vaccines.
So NIH has funded my Institute to make synthetic fragments of every influenza virus that we and others have ever sequenced.
So, we are going to have all these fragments just on the shelf.
And if there is a new pandemic, we are actually going to be making vaccines with the new emerging ones for next year,
within less than 24 hours we can make new vaccine candidates that can go right into their new cell production system,
and get very rapid production of vaccines that we think are going to be far more effective
than our century-old technology we are using today of growing things in chicken eggs.
So, we are trying to apply these tools in a wide variety of areas.
As I said, we asked ethical questions before we started the first experiments.
There is a series of reviews that have been published along the way,
including one that the Sloan foundation has funded my institute along with MIT to look at security concerns.
On our announcement this spring, President Obama has asked the new bioethics committee
to deal with this as their first priority and their report out is due very soon.
This is a report from the Royal Academy of Engineering last year,
saying these synthetic biology,
synthetic genomic tools are likely to be the number one wealth generator for the next century.
For countries, for companies, for individuals,
because they have a chance to completely change how we make everything from food to fuel.
So, we are just at the early stages of this.
The first stage took us 15 years.
We didnít think it would take that long when we started out.
But we developed our own funding to go along with our belief that we will get there.
If we relied on government grants, it would have probably been withdrawn a long time ago.
These are the early stages.
What took us years to do, you can now do in a day.
Hopefully, what we can now do in a day, within a short while we will be able to do millions of times a day.
And just think how that accelerates biology and our understanding.
So, just compiling some things for long term.
Space flight, obviously, we have to start looking at the genetic code of people to understand their range of biology,
how to prevent diseases, to understand and predict what's gonna happen.
Identifying traits compatible for long-term space flight.
With the microbiome, understanding how microbes contribute to health and disease and trying to get positive traits,
placing pre-existing ones, and then everything from food to chemicals to materials.
I think this list could be extended indefinitely.
These microbes can be self-correcting with as Deinococcus does, with radiation.
Perhaps the only way to do that with humans is to send up
a set of a lead-covered container for stem cells to do replacement.
But we think we can use synthetic tools to even improve on stem cells to do exactly what we want them to do.
Then ultimately, if it's gonna really be generational space flight,
we might want to go beyond selection ultimately to engineering.
Thank you very much.
>> Moderator: Thank you very much, Craig.
That was great!
If people would like to line up on the microphones over on this side, if they have questions.
If you could identify who you are and where are you from.
If there are any press in the audience that will have questions,
then please also line up here and move to the front of the line.
>> Dr. J. Craig Venter: Pathogens would be amongst the most difficult things to try and deliberately create.
And youíd have to be starting with something pathogenic public to do that,
but biology is still there is a stage of surprising people all the time.
I think doing these experiments;
weíre designing everything where it could not survive outside the laboratory or facility where it was produced.
We can do this with suicide genes, chemical dependencies.
We have almost twenty or so years with this, probably been tens of millions of experiments done in
molecular biology with E. coli, that can't grow outside the lab because of such a chemical dependency.
So we know how to control these things.
I think the ability to rapidly make vaccines,
I think new emerging infections are orders of magnitude greater risks than
humans making anything infecting deliberately or otherwise.
>> Audience 1 (female): If you were to choose chassis environment, chassis microorganism for space environment,
what do you think it would be?
So it's a containment organism with minimum genome that you would start modifying and rebooting, for example in space.
And the second question, I'm sorry for not letting you answer it right away.
And the second question is, what type of space stressors would cause
a colony of multiple organisms to take on the whole chromosome, you think?
Maybe it's radiation?
And how easy it is?
What is the probability?
Thank you very much.
>> Dr. J. Craig Venter: I don't think there is any one type of microbe
right now that we know about that would be any better than others.
In fact diversity is gonna be important if we're gonna try and recreate a microbiome, generate energy, create food, etcetera.
But they need to be robust.
We obviously like to build in the kinds of things that Deinococcus,
where the endurance has where they can self-repair very robustly from continuous radiation.
So DNA repair is important for us as it is for these microbes.
And there is a wide range of DNA repair systems in microbes going all the way to something like Deinococcus.
In terms of exchange, you and I were talking about this earlier.
I think in the environment on our planet we see not just from our deliberate transplants but when we look back at the environment,
we see all kinds of organisms that have multiple chromosomes in the microbial world.
The first one we saw was with cholera.
One of the arguments based on 16S RNA, there was no point in sequencing the cholera genome because it's very close to E. coli.
But when we sequenced the genome it actually had two chromosomes,
one that was very similar to E. coli and one that was very different.
And so we think we see this all the time.
So Deinococcus has four chromosomal elements, some quite different from the core ones.
So I think many species have acquired entirely new traits.
In a heartbeat, literally, you could add a thousand new traits in evolution to a cell.
We don't understand these mechanisms very well.
We don't know if there's cell fusion.
Some cells take up DNA quite nicely.
So obviously, continued evolution is something we have to understand with the microbes
that we are putting together in stressed environments.
>> Orlando Santos: Hi, Craig.
I'm Orlando Santos from NASA Ames and I wanna ask you about technology development timelines.
Because if you take out your crystal ball, given what you heard all day today,
what do you think would be the first application of synthetic biology to
NASA's mission and how long do you think it will take to get there?
>> Dr. J. Craig Venter: How much money you got?
[Crowd laughs, Venter chuckles]
>> Orlando Santos: Good answer.
>> Dr. J. Craig Venter: I mean, you know, without knowing the level of effort going into it that's impossible to answer.
But I think it could change the shape of everything NASA does if you make the commitment to do it.
You know, itís, I found it really hard to predict the future ever when
I'm a month or two away from something that I'm sure will happen.
I've been wrong many times.
But, yeah, we're a small group of scientists that have done this work independently over a fifteen-year period.
The fact that you or anyone else now could take these techniques that we've published and these kits that would
be available and do what we did in three years and do it in an afternoon
or so means you could move a whole lot faster and learn a whole lot faster.
So, without the financial commitment and the intellectual commitment, yes, it's not just dollars.
It's having the right people doing the right experiments for the right reasons.
I can't think of an organization that has more potential to use synthetic genomics and synthetic biology than NASA.
I mean, Exxonís doing pretty well but itís trying to stay on this planet, right now?
>> Andrew: Hi. My name is Andrew Bingham.
You mentioned openness a couple of times in your presentation in posting
the human genome online and how that would spur research in the area of biology.
My understanding is that currently up to 20% of the genes in my body have actually
been patented by different companies involved in the area of biology.
>> Dr. J. Craig Venter: You must have some cool genes.
[Crowd laughs, Venter chuckles]
>> Andrew: And Iím wondering,
how did you strike the balance between the need to give private companies incentive to
do this type of research versus the idea of patenting things that are then around
in nature for hundreds of thousands of years.
>> Dr. J. Craig Venter: Well I would urge you to try and understand the patent process and why we have one in this country,
ëcoz itís designed to encourage and force openness.
The Coca-Cola formula was never patented and was kept secret for a very long time ëcoz it was a trade secret.
So part of their reasons for having patents is to force people to completely disclose all the information about their invention
so that somebody else could build on it The trade-off is that our
government makes with inventors is you get a period of exclusivity,
not to own the data, not to block every people from using it, but to commercially develop it.
Itís the basis of our economy.
I doubt that your 20% of your genes have any value rather than to you.
But some of them do, you know.
Thereís millions of diabetics that are...
>> Andrew: Wait. Theyíre not inventing it. I mean look at...
>> Dr. J. Craig Venter: Let me finish, please.
There are millions of diabetics that are very pleased UC San Francisco patented the gene for insulin and Genentech produced it.
You know, it doesnít matter in this country.
The law is, it treats invention and discovery exactly the same.
If you have an issue with that take it up with a lawyer or a congressman.
But itís the law of the land and thatís how science and industry move forward.
>> Doug Messier from Parabolic Arc.
This is space manufacturing some of us, looking at that, can you see organisms being built to assist with mining,
manufacturing life support systems, those types of things?
>> Dr. J. Craig Venter: I am (inaudible name), to what extent are you talking about mining?
There are several groups, companies on this planet trying to see if microbes can enhance mining of copper, mining of gold.
Obviously the oil companies weíre working with BP to see if we can use these deep-earth
microbes to enhance oil recovery, change viscosity of oil etc.
So, how and if these things apply to work on Mars is beyond my expertise.
>> Doug Messier: OK. Thank you.
>> Wayne White. You said that some microbes have the ability to self-repair DNA.
It occurs to me that if you could introduce that trait into the human genome,
then one of the consequences of your work would be life extension.
>> Dr. J. Craig Venter: Well, all humans have the ability to repair DNA.
In fact it was a discovery early on that we made with Burt Vogelstein that mutations
and some of these DNA repair enzymes are associated with colon cancer.
So if you canít repair the damage that weíre constantly being subject to, others can increase incidents of cancer.
Basically, all organisms can repair their DNA.
Not all of them can do what Deinococcus Radiodurans can.
It can be blown apart and reassemble their chromosomes in the same way.
But it turns out there is probably a very large number of organisms on this planet that can do it.
People think it wasnít necessarily evolution to deal with radiation
as much as it was drought resistance, maybe very similar mechanisms.
Being able to completely reassemble our chromosomes like Deinococcus can, would be an interesting phenomenon, right?
Iím not sure with the biological consequences of that would be.
But itís an intriguing idea.
>> Patrick Fu: Hi. Iím Patrick Fu.
You have shown that chemically synthetic genome is doable and it has been created.
Do you have any plan to design for the synthesis of the membrane proteins
so that those chemically synthetic genome can live inside,
it does not need to just borrow other bacteriaís body to accommodate it.
>> Dr. J. Craig Venter: Weíre not specifically doing that.
It would be, weíre trying to see if we can design sort of a universal
recipient cell that we could put a variety of chromosomes into.
And there would be enough diversity of reading that DNA so it could start making almost
any protein system until it can get its own system going.
It needs to get early life going from DNA, we have to be able to read that DNA.
So we need tRNAs and a few other components.
Weíre trying to see if we can make sort of a basic cell that has diversity of
those that could deal with a wide range of codon usage and other things.
So, I think thatís gonna be essential in terms of standardizing things for the field, but that may not work.
I think what you are talking about in terms of unique membrane protein production;
weíre not doing that per se.
But I think thereís ten million things that we can all think of doing that we alone arenít gonna do.
So hopefully you will.
[Venter chuckles]
>> Audience 2: There are some experiments which should not be done on earth.
For example recent work in high energy physics can make micro black holes
which is they considered safe because Hawking says theyíll evaporate.
But if they donít, the Earth is?
The question is, are there experiments that we would like to do that are so potentially hazardous,
it would be a good idea to do them in some remote laboratory of earth?
And is that a good reason or not a good reason to go into space?
[Venter laughs]
>> Dr. J. Craig Venter: It depends what the legal system in space is, I guess.
Itís a, I was offered an island off of Belize but I read a book about that one too.
So, I donít think that is such a good idea.
I donít think we should have a different ethical system necessarily in space.
But I think human engineering is one of those things that we sort of
all agree on you canít do ëcoz you canít do human experimentation.
You know, that leap from selection to engineering is gonna be a very complex one for society, I mean if it ever does.
I donít think doing that in space makes it any better, easier.
>> Silvano Columbano: Silvano Columbano from NASA Ames.
To what extent are you able to predict functionality from your designed genome?
And if itís merely trial and error in a computer,
could you do the trial and error in a computer, as a form of simulation?
>> Dr. J. Craig Venter: Yeah.
I wish we could do everything by computer modelling and simulation right now.
But we canít even completely model this very simple 500-gene cell.
It's basic metabolism can be.
But even with that cell 100 of the essential genes for biology.
In other words, if we remove that gene or disrupt it the cell dies out of unknown function.
So, biology is fundamentally in a discovery, not by a first principle method right now.
I mean thatís what these combinatorial methods,
if we can learn rapidly from them we can get to the first principles that there is not a single genome
where a scientific community understands the function of every gene in it.
We are not even close to that.
So, empirical science is a big part of this field for a long time to come.
Thatís why we need good ways to measure, good ways to speed it up.
>> Silvano Columbano: Thank you.
>> Dr. J. Craig Venter: Weíre done!