So I was thinking about the lecture yesterday, I believe, on plant, you know, plants, natural

products. And so thinking along those same lines, I was wondering if, in your case, you have

operons but there is also the possibility for specialization, right, so that you could have orphan genes outside of the operons, but I was wondering if maybe more generally the question is how you choose, since you have so many candidate operons to work from, how

do you choose which ones you want to wake up, you know, is this based on novelty or because you know that there is not less of a possibility for this kind of specialization. So priority, prioritizing these gene classes, which one do you want to wake first?

So it depends on what you want. For us it could be something that's completely novel, if you want a diverse chemical structure then we can deduce from the gene sequences the kind of class of natural products it will be. So we can go for something that's completely different, but also sometimes if you want

different but similar activity, let's say, okay, we've got penicillin, but we want a different variety of penicillin, modified penicillin, then you can choose those that are similar as well too. And in most cases in the microbes, it is clustered, these classes are usually clustered

unless if you've got like the one enzyme step or something like this for natural products. The modifying genes can be single enzymes somewhere else sometimes, but it's very rare. So most of the time in microbes it's all clustered. This is the difference from plants, a big difference from plants.

But the priority, yes, well for us it depends, like I said, on what you'd like to do in the end. Does that answer your question? Yeah. So I have a question for you, Errico, which is how do you deal with best known cases,

that is, cases where you know more or less what the actors are in the metabolic pathway leading to various compounds such as antibiotics, can you apply some metabolic computational

analysis, has it been done successfully, is it interesting from the point of view of setting up the modifications, alterations to the pathway, what's the practice? So you're talking about the whole cell metabolic pathway, metabolic modelling, or is that

leading to more? Yes, but that would focus on the pathways of interest for the production of various compounds such as antibiotics. So you're just thinking of the biosynthesis pathways for the antibody, is there metabolic

engineering going on there, or modelling, rather, modelling? Yes, because there are some difficulties and stuff, so it's not directly easy. Yeah, well, I don't know if it's as, it's not as easy as a normal primary metabolism modelling, I think it's just as difficult in fact, but not a lot of people have done

this. But having said that, the companies who actually produce compounds as a product, they know a lot about these enzyme pathways, they know which ones are limiting, which ones are not limiting, and they go off and change them. So in a way they are modelling these pathways as well to understand which enzymes they need

to go up, down, for example, or cofactors as well is a big thing, and also the starting material, substrates. If you don't have enough substrates, that's of course a problem as well. So it's not published as modelling these pathways per se, but industry has lots of

background knowledge, I think, for modelling. But I think it's also interesting to do, too. Any other questions?

We've seen that refactoring of this metabolic operon can be very challenging, I'm sure you're aware of the work of Chris Voigt, for instance, so what's your take on that? Do you think that there could be also alternative approaches to refactoring, like

for instance adopting the sigma factors from the host where it comes from or different things like that? Sure, absolutely. Any comment on this? So I think it was really, Chris did a great work, I don't have anything against Chris's work at all, it's just that he chose a pathway that was a primary metabolism pathway,

really, associated with nitrogen fixation. So if he had chosen a different pathway, a little bit more simplistic pathway, perhaps he would have gotten a better result. So I'm alluding to the fact that Chris's group did a lot of refactoring, then in the end their production of the fix of the nitrogen was the same as wild-type after they

refactored it. So perhaps if he used a different pathway maybe it would have been better, perhaps. I think, yeah, your idea is great, why not? You can do lots of other things, put different things into refactor pathways. So I think, yes, yes, I hope I'm answering your question.

So you can do lots of things. You can, like I said with the refactoring, promoters are important, ozone-binding sites are important, both strengths are important, terminators are important, directions of the genes are important, all of these things you have to think about and you just have to dissimulate it and try it.

So this is for a general question, not only for you but maybe for the whole audience. So I just wonder about the long-term stability of the constructs that we make in the laboratory, right? So when you want a population to obey your orders, you have basically two strategies.

One of them is to enslave the population and then to punish all those that do wrong or you can try to make your employees happy and then they work happily and then they may be working for you for much longer. So I think that the only strategy that I've been using from the beginning is to punish bacteria

and they do wrong to your orders. I just wonder whether we can think of some type of increasing bacterial happiness by the time that we engineer them to do something. You know, microbiologists are paranoid about bacterial stress,

but I'm interested in bacterial happiness. So I don't know whether people have figured out ways of making bacteria happy at the same time that they are in our structure. Because if we do that, the long-term behavior will be more stable and more predictable. So what's happiness for the bacteria? Happiness for the bacteria is easy, it's growing.

Is it? That's the only thing that they understand. You have this phenomenon called caloric restriction. I mean, if you grow fast and get fat, it's not a good idea. So you don't have to find the right thing. The human species is proving you wrong, right? I just wanted to follow up on Victor's suggestion

and fairy tale about microbial happiness and so on. It depends whether your bioprocess is going to be static. You know, you do fed batch and then you use it for making something.

Or you are going to have proliferation along the way, which is a very different thing. So in Darwinian terms, leaving more progeny is happiness. No matter what it takes, you know, resisting radiation, storing fat at some point, or phosphorus for using it.

There are many, many algorithmic ways of representing it, but leaving more progeny is okay. And I think at this current stage of understanding life, we are absolutely, it's beyond our understanding to do evolutionary design. Not only could we construct something that is a true achievement,

to have something like you describe, then having the proper, maximizing the formation of some secondary metabolites by fine-tuning, but this will be for just, let's say, a few generations.

Now if you grow a culture like that, the first thing it will do is get rid of this viola scene of our liking. You see what I mean? So evolutionary design, we have not even started to think, even to pose, to make the statements, the specifications about it.

There are only empirical things and I think the main concern about this is how to prevent dissemination. So there is some efforts ongoing in several labs because that's the least that we can promise to the public, not having our Frankenstein very sane, making bugs, spreading in Amazonia or wherever.

But this is only the beginning. And I think it is, you know, since we gather regularly now, I think that this evolutionary design part should become a systematic session, systematically unforced session, together or aside by breaks and things, which are far less important for the future.

That's a suggestion. Can I just respond? Yes, please. You first, though. Sorry. I just wanted to ask, you said that to proliferate is the happiness for the bug, right? Not me. Charles Darwin, say.

But then they also die. So does that mean you want long-lived bacteria or you just want numbers? What do you think? It depends. You know, I know species of jellyfish, you know, the males just burst out in sperm cells, but they are happy to do this because next generation, they will burst even better, but the number of jellyfish increases.

So it depends exactly. You could say the figure of merit that I want to improve is this, and then algorithmically, you could try to figure out how to increase that particular trait, but we must keep in mind that it is the very capability of generating more protein that the species of the creature that we are evolving or engineering are after.

So they can die. They can die if you have lots of babies. It's a mortality rate. They must not all die, of course, but you must understand that. One comment or maybe two comments to the happiness of bacteria. I think at the end, bacteria always find strategies somehow to maintain the species.

That means somehow to survive. And this means sometimes faster growth or sometimes better stress adaptation depends on the situation. At one point, what we observe very often is that it forms under certain conditions

many, many different sub-populations. And when I look for all these modeling studies, I mean how we ever want to integrate these sub-populations is another challenge, which is, of course, almost underestimated, and I mean you can just knock out one gene and suddenly find heterogeneous behavior, or you can overproduce and find homogeneous behavior.

So you can narrow the difference. Any comment on your side? I'll take a different stance.

I think you can predict to some extent, because you can analyze the different components that you're engineering into things like bacteria and work out what the relative fitness decrease of running that component will be for the cells in the conditions you want to grow up. You can analyze DNA sequence and look at the predictability from the sequence level

of that sequence being deleted or changed due to things such as repeat sequences. Maybe, for example, the way you've cloned your plasmid leaves a similar sequence multiple times, then it's going to have a more likely chance of being ejected out later in design.

So although I don't think we can easily work out ways to keep the growth rate perfect to avoid our engineered bugs from being out-computed by ones that have not been engineered or that have got some mutation, we can do things about the design that can improve our chances that our DNA will be kept for longer periods of time

because certain components cost the cell less effort, so RNA-based regulation, probably less cost to the cell than transcription factor-based regulation, and designs that avoid certain sequences where transposons are easily going to come in and cause damage to your DNA and also be avoided.

So you can do something about it. Ultimately, I don't think we really want to get to the point where they grow even better than wild-type E. coli because, for example, if you're doing this work in E. coli, that would be breaking what was set out at Asilomar in that our lab work in E. coli would not create strains that grew faster and were fitter than the wild-type natural ones.

If I may interject my own view of that, I think it's a matter of domestication and styles of how we domesticate bacteria and yeast. Think of cows and maybe some wheat and corn.

From that point of view, actually, there is probably an incentive, if you want your bug to run during months in a 1,000 cubic meter fermenter and not change properties too much, to be perhaps a bit on the nice side of domestication that the cells are somewhat happy with what you added to their constituents

so that they stay robust. And Philippe does not agree. I know the complaint of yogurt makers. So they are making cubic meters of yogurt every day that God makes.

And sometimes it tastes like cat piss. Because some viruses were there that resisted. So the only way they have sometimes is for instance millions of euros, actually. When a ferment of yogurt like that is lost. So they have a real incentive in understanding and steering this evolutionary thing.

They have absolutely no model, no way, no other thing, like no more than so many variations and so on. So that's what I said earlier. I'm sorry to repeat myself. This thinking has not even started. So let's just follow up with what Philippe says. You said we have to think on evolutionary engineering.

Design. Part of the design should be... So what I want is to have things that don't evolve at all. Absolutely. And what I want is an existence that can be realized. Definitely. And then that takes me to the other side of the coin. I use this issue about growth. So I think that one of the big problems also in engineering and so on

is that bacteria grow and that's a component of the whole system that makes the whole thing very different from what an engineer has. So an airbrush does not grow or a radio does not grow, right? So I just wonder whether we can also... We're also thinking of bacteria in terms of something that grows from the time. So what I would like to have would be to something that you would call an adult bacteria.

So we grow at the beginning. We have a long adult life. We do our jobs and everything and then we peacefully die. So can we think on adult bacteria that didn't grow however they do what they ask them to do and everything? So I think that it's been touched laterally this issue of uncoupling growth from activity

and to this day I think that this is still a big issue. I mean how can we use synthetic bacteria to uncouple all together? Growth from synthetic activity, not stressing the cells because this is the standard way to do it and still having a happy adult catalyst. I don't know. Do you have any ideas about that?

If I had them I would most probably not communicate them here. Rather silently pursue them. I think the problem description is very accurate. Also the solution would be great. I think nevertheless it's remaining a very tricky endeavor

because these regulatory processes that define whether the cell is happy to whatever definition can not really be pinned down to a limited number of things.

We need to turn this up. We need to turn this down. So to me they seem to be distributed over a large number of aspects and to change them will a. require better understanding and b. lots of interventions at different spots.

So I think what you describe is great but it will take some time. I can add a comment perhaps. Antibiotics or secondary metabolites as you know are not produced in primary metabolites. They wait until they stop or reduce growth and that's the only point when they start producing.

So maybe they're happy producing these compounds because they're not thinking of growth. So one of the ideas of course of pharmaceutical companies and we did before as well was to engineer so that they only produce when they stop growth. You engineer the expression so that they come on only when you have a fixed mass

and then you can switch it on or off whenever you want. I had a question for Pablo. So one of your final results was very nice that you showed that biofilm or organisms in biofilm had higher degradation of the chlorobutane.

I was wondering if you had maybe at this early stage but any speculation at least on mechanism for this? Is it greater chemical resistance? Well one of the things that for sure is different in a biofilm is the way the cells can withstand the substrate.

It's extremely toxic so you can only add it up to 0.5 mm more. So cells in the biofilm are much more resistant to the substrate itself which is toxic than in the planktonic state. That's why we think we recover most of the allogeneic activity in the biofilms.