Do you have a question you'd like answered on a future episode of Translating Proteomics? Send it to translatingproteomics@nautilus.bio!

Proteins adopt a wide variety of functions depending upon factors like their location in the cell, their modifications, and the biomolecules they interact with. While many of us may have been taught that single genes produce single proteins that have single functions, protein function is far more dynamic than that. In this episode of Translating Proteomics, Nautilus Co-Founder and Chief Scientist Parag Mallick sits down with University of Cambridge Professor and proteomics expert Kathryn Lilley to discuss our evolving understanding of protein function. They cover:

• How they came to realize protein function is more complex than one gene, one enzyme, one function
• Factors that give rise to the dynamic complexity of protein function including proteoforms, protein localization, and moonlighting
• Steps we can take to better understand and teach others about the complexities of protein function

Research diving into the complexities of protein function

• Research from the Beltrao Lab using bioinformatics techniques to identify functional phosphosites (Ochoa et al. 2020)
• Work from the Lilley Lab integrating techniques to investigate ome-wide localization of both RNA and protein (Villanueva et al. 2024)
• Lilley Lab preprint investigated protein localization changes in a cancer cell line as a result of ionizing radiation treatment (Christopher et al. 2024).
• Collaborative work with the Lundberg Lab mapping subcellular proteomics (Thul et al. 2017).

Additional protein function resources

• MoonProt - A database for moonlight proteins from Professor Constance Jeffrey's Lab
• Translating Proteomics Episode 5 - Why the Biology Surrounding Biology's Central Dogma is Wrong

In this episode of translating proteomics, host Parag Malik chats with University of Cambridge professor Katherine Lilly about the intriguing complexities of protein function. Their conversation covers how parag and Catherine came to realize protein function is much more complex than one gene, one enzyme, one function.

Factors that contribute to the complexity of protein function, steps we can take to better understand and teach others about the complexity of protein function. To introduce Catherine and this fascinating topic, here are your hosts, Parag Malik and Andreas Humer of nautilus biotechnology.

On this episode of translating proteomics, we welcome special guest professor Katherine Lilly of Cambridge University. We discuss a topic that seems simple, but becomes very complex very quickly. Protein function.

Specifically, we cover how proteiforms may impact our understanding of protein function, why an evolved conception of protein function must account for the context in which proteins are found, and what we gain when we think about protein function in context. With that, let's watch my conversation with catherine.

As you just heard today, I have the privilege of speaking with Catherine Lilly. Catherine is a professor at the University of Cambridge and director of the Cambridge center for Proteomics.

She has deep experience in quantitative proteomics techniques and has developed a variety of means to study proteome dynamics.

In particular, her pioneering work has assessed proteomic differences between subcellular compartments and shows just how important protein localization is to biological activity. And her research is helping us develop a new understanding of what the term protein function means. Catherine, welcome to translating proteomics.

Thank you very much. Pleasure to be here.

To get started on the podcast, we've discussed some of the ways that classically we think about protein function. But for many years, people followed the paradigm of one gene, one enzyme, one reaction. I'd love to start back in time a little bit.

Your earliest days as a biologist, do you remember being taught about protein function and beta lintatum? Is it wrong? If so, why?

One gene, one protein, one function. I think I'm so old that I wasn't even really taught that. I remember being taught about proteins.

I remember being really excited about the prospect of amino acids in different order. I was fixated with the sequence of proteins.

In fact, my first job was as a professional protein sequencer using Edmond degradation in terms of linking gene to protein to function.

It's not something which I really took seriously until I moved to Cambridge, because at that point I was running a proteomics facility and people were giving me their samples. I was trying to look for differences in protein expression, and to my mind, then the fun started.

So you'd have a list of proteins that were changing in abundance. Sometimes this list seemed not to make much sense, that the proteins within it didn't seem to be particularly related.

But then you had to try and work out, well, what does this mean for the cell? If you've got more glycolytic enzymes being expressed, if you've got less of something else, then what does it mean for the cell?

And I think years ago, you read proteomics papers where the first part of the prot, the paper was talking about how they got the datasets. But then there was an awful lot of conjecture at the end of what it might mean.

And I think at this point, I got excited, really, about the size of the proteome and what the proteome is capable of.

Size of the size of the proteome. To most people, the genome has about 20,000 genes, therefore there are about 20,000 proteins.

I wish there were. It'd be great if there were only 20,000 proteins. That would make life so much easier.

So what we know for the human protein is it's got a fairly small genome. Given what we have to do as human beings, we know that there are tens of thousands of splice variants.

We know that there are millions of potential single nucleotide polymorphisms in the coding region. We know that once rna is made, there's adenosine to inadine modifications.

And all of this together will give many different chemical entities, which then expand still further by the amount of post translational processing.

So this could be truncating a protein, making a smaller version, a shorter version of it, or it could be through post translational modifications, such as phosphorylation, acetylation, glycosylation, we know that there are huge numbers of potential sites for post translational modification.

And if we just imagined, and there's some lovely papers in the literature, many spear headed by Rudy Abersold and his collaborators, who have done thought or had a thought processes. So how many different proteins could exist?

So even if you imagine a cell expressed, expressed as 10,000 proteins, and each one of those, you could have 100 different modifications, then that would mean a million different chemical entities. So we call them protein forms. So these are separate chemical entities in a cell.

And we haven't really got a head around the combinatorial nature of these post translational modifications. Some of them may be constitutive, some of them may not be, some of them may be regulatory.

They're all going to be potentially in different orders and different combinations, and at worst, you could imagine that every single protein in a cell is different, so every copy is unique, taken together. I think what I'm saying is that there are huge numbers of proteforms and we don't actually know what that number is, which is pretty scary.

I think that's actually terrifying because we have a hard enough time dealing with the complexity of transcripts. And so now you're talking about orders of magnitude more complexity.

And I feel like you even simplified it, that beyond just in your own research, beyond the chemical entities themselves, where they are, how their dynamics change over time, are extra layers of regulation on top of what the chemical entities are?

Absolutely. And something else to think about is that proteins are extraordinarily gregarious. They don't go off by themselves.

They are generally part of multiprotein complexes. The components of those complexes are highly dynamic as well.

Maybe the that they're controlled by post translational modification, maybe they're controlled by location, that proteins will form complexes with other proteins when they're in a particular part of the cell. And a final layer of complexity is that proteins are shapeshifters.

They can adopt multiple different structures, and the structures themselves will dictate what those proteins are capable of, what their function is.

So now, the very thorny issue, which we haven't quite got to, is, if there's all these different protein forms, do they all do the same thing as each other?

So if you've got, I don't know, say a thousand different versions of enolase, does the enolase glycolytic enzyme, does it function the same if it has different modifications, if it's in different parts of the cell, if it's got different interacting partners?

And I'll ask you for your opinion on that. First, do you think all of these different variants are impactful?

I don't think we know.

That's a very.

The reason we don't know is because we can't measure them, or we haven't been able to measure them at the depth that we need to measure them to be able to answer your question. So the problem is that the way that we identify proteins, that we quantify proteins generally, is mass spectrometric. This is the way we do it.

And what we do is to take our proteins, we digest them to peptides, and then we measure the peptides, and we use those as a surrogate from the protein from which they were derived. So we never measure the whole lot. So 5% of proteins, only 5% of proteins in most proteomics experiments, will have more than 50% coverage.

And what I mean by that is we've identified more than 50% of the, the peptides which have been derived from them.

And the other thing that we don't do because, or we can't do because of the way that we collect these data is we're looking at individual peptides and we can't then piece them back together again.

So even if we can work out which ones are phosphorylated or which ones are acetylated, we can't stitch them back together again to know actually in the cell which of those modifications were part of the same polypeptide chain.

So this is a problem that we don't have the depth of coverage and therefore we can't really say what the proteins, or if a particular proteform has responded in a different way to a stimulus which might give us some insights of what that protein forms. Function.

Yeah that's a really, that's a great point, is that if you can't measure it, it's really hard to learn about it. You're making lots of inferences under uncertainty.

And we also, we don't know. It's great to be able to speculate for a protein how many different protein forms may exist, but we can't measure them.

So some of them may never occur and some of them may also not be of any functional importance.

There's been some lovely work done from Pedro Beltrao's lab where he's looked at phosphorylation, which are regulatory post translational modifications and which aren't, which are just are there, but don't make any, have no bearing on the function of the protein.

I think this will ultimately be a really important question because on some level we're asking about how many stars are there in the universe. And that's a layer of question. But some of them we can't see in the night sky, some of them blink, some of them go away over time.

And we really don't even know what the map of the universe is when we're talking about this proteiform landscape at all.

Absolutely. Yeah.

For me, I remember there was this distinct point where my view of protein function changed. I was originally trained very much as a structural biologist and so I viewed the world.

If we could figure out the structure of a protein, then that would immediately tell us its function and everything we would need to know about it and wed be done. Then the concept of proteins and their friends and their interactions started to come into my awareness.

And recognizing the layers of regulation that existed above the individual protein molecule in its structure, how it played with its friends and neighbors, and even more things like transcription factors that have no enzymatic function.

They are structural, they move from place to place, they help with things, they're clearly important, but they don't have an enzymatic function in the way that we think about canonical enzymes.

It was really me changing my frame of reference from the function of a protein is what it does as an enzyme, to the function of a protein is how it contributes to the physiology of a cell and an organism. And so that frame of getting a little larger was a bit of an aha moment for me. I'm curious for you.

Where was that moment for you where you said, oh golly, there's more to this than I thought.

I suppose for me, it was looking at the data that we generated more than ten years ago.

Now, in the lab, using our method for subcellular fractionation, what we do is that we take cells, we lyse them gently so the organelles are intact, and then we carry out the subcellular fractionation.

And then each fraction, we look at the protein content and its quantity, and from that, using patterns of how proteins are distributed through these fractions, we can create a map of the cell, and on a cell wide scale. So we've got clusters of proteins that are all doing the same thing, that are within our fractions, which are mitochondrial.

We can determine which proteins are in the Golgi apparatus, which are in the endoplasmic reticulum.

And we teamed up early on with Emma Lumberg and her colleagues when she was still in Stockholm, so that we could marry our datasets with data sets that were being generated from her lab using immune fluorescence. And so using completely separate methods, we were trying to give a protein an address in the cell.

And surprisingly, it turns out that both methods had a high degree of agreement, which was great. But then we started to dive into our own data.

And what we did, because we had a lot of it, was to have a lab hackathon where we all went into a college room somewhere else in Cambridge, and we each took a piece of biology and just mapped it. And I took glycolysis because I thought glycolysis was quite straight. I've been taught university where glycolysis takes place.

It's a cytosolic thing. And so I expected my glycolytic enzymes ought to cluster in the cytosol, but they didn't, and three of them were doing something a bit different.

I thought, this is strange.

And I'd forgotten all the niceties of glycolysis at that point, so I had to go back to textbook and lo and behold, the majority steady state location of the glycolytic enzymes was cytosolic, apart from three. And the three that had mixed location, that were in more than one place were the enzymes, which are the points of control.

So the kinases, so pyruvate kinase, phosphoryptokinase and hexokinase. So that then got me intrigued.

Moreover, the proteins themselves, about 50% of them, fell into that sort of category of the glycolytic kinases, that they were in multiple locations. You could map to see that they were in different places. So then I thought, okay, what's another set of enzymes that I know about?

So I thought, oh, I know the amino acid tRNA synthetases, I know all about those. At least I thought I did. And I imagined that, well, I'm not sure what I imagined.

They, I thought they would be in and around the translation machinery. That's where they needed. And it turns out that many of those have got mixed localization as well.

And when I looked into this, it turns out that as through evolution, those enzymes seem to have picked up extra sequence, and so they moonlight, they can do different jobs. So this then got me down the track of trying to understand why, although I've just. I'm going to go back on what I've just said.

The proteome is massive. It is. But why? For 20,000 genes, we've got so many different protein functions.

So therefore, the idea of moonlighting and the fact that a protein may be coded for by a particular gene, but through different processes, once it's been translated, can adopt completely different function. So I think that's what got me into thinking of proteins, is what you've just said. You know, they're sort of their presence in a cell.

Which then brings us onto the thorny issue of when you try and drug some a protein, or when you try and knock it down, if your protein is doing multiple different jobs, you don't realize what those jobs necessarily are.

You don't understand the full repertoire of what your protein is involved in, then you're going to get all sorts of effects that you're not expecting.

That's a really interesting question. I'd like to just double click on that a little bit on the.

What does it mean to drug a protein that does multiple things in multiple places at multiple times?

Well, you have to know those pieces of information. Maybe we don't I mean, if we have. For many proteins, we think they know what they do. For an awful lot of proteins, we don't.

I mean, there's a lack of functional annotation for quite a large proportion of the proteome.

The ones that, the proteins that we think are the ones which are going to, if we can stop them carrying out their function, we're going to cure a disease.

We have to know what the functions are and where the proteins are and what they interact with and how this changes in different situations, in different cell types, for example. So yes, it's relatively. It's scary when you think about it, if we don't know what we don't know about the proteome.

But on the other hand, it explains life is pretty complicated. Humans are fairly complicated. So if it were really as simple as a clock, then that would be somewhat disappointing.

So you had a paper last year, Villanueva et al, where you were looking specifically at both rna and protein and subcellular localizations, how they changed over time. And so I'd love to hear just a little bit about how much movement are we seeing of proteins in general from place to place.

How does that relate to simple rna expression, and what does that mean for how cells work?

Well, I would go back further than that paper to papers from my group and many others that have captured two pieces of information during a dynamic process. So about the proteome. So those pieces of information are how much is the protein abundance changing, and is the protein changing location?

So you end up with two pieces of information for as many proteins as that you can measure. And you sort of try and marry these together.

And in the systems that we've looked at, and, okay, they're limited coverage in terms of the biology that we're trying to capture, but the two seem to be quite decoupled. So you can have changes in location with no net change in abundance.

So depending on what the system is, that could be a result of two different processes.

It could be that you have a pool of proteins in one place, and in response to stress or whatever the perturbation is, you move them somewhere else because they're needed somewhere else, and off they go. And that may be through post translation modification. They're making new interactions and they're retrafficked somewhere else.

It could be that you degrade, or the cell degrades a pool of protein in one place and makes new protein, and that is then trafficked to a new location. And in most of the studies that have looked at location and abundance, this isn't captured either.

You know, the protein is now somewhere else, but you don't know which of those eventualities have caused it to move. Has it been degraded and resynthesized, or has it just been moved lock, stock and barrel?

So what this throws into sharp relief, I think, is many of the proteomics workflows that people have adopted to date, where the readout for whatever's happening in the cell is to do simply with abundance.

And I'm not saying that that's a bad thing to do, but these studies could be missing a trick, because it may be that the major cellular process, which is controlling whatever it is that they're looking at, is to do with relocalization rather than the change in abundance.

For projects that we've had in my lab, generally you don't see massive changes in location, but the changes in location are important and they make sense.

A paper from a few years ago now, Claire Mulready, a former very talented postdoc in my lab, was working with a monocytic cell line, which she then treated with lipopolysaccharide, and looked to see how this monocytic cell line was evolving. So it, it goes off down a macrophage like trajectory. And we did this all blind again, hypothesis generating. We weren't trying to test anything.

And she noticed a whole complex worth of rho gtpases that were moving to the plasma membrane. And we thought, well, this must make sense, because as the cells are differentiating, then they become more motile, their morphology changes.

So this made perfect sense to us, but we weren't looking for that. It just popped out of the data set that we had. And importantly, none of these proteins were changing in net abundance.

So again, localization was key here, whether they were moving or whether they were being resynthesized and the newly synthesized set of proteins were moving to the plasma membrane, I don't know, but it showed us the importance, really, of looking for localization changes as well as abundance changes, because we would have never picked it up if we simply looked at protein abundance.

The paper that you were referring to is we went one step further, where not only did we capture protein movement in protein location, but also the transcriptome as well, to try and map where that was and how that changed.

And what we were looking at, particularly in this study, although I don't suppose we necessarily set out to study it, is the formation of granules and the fact that there seem to be very large scale changes in rna localization, particularly rna that was messenger rna that was associated with the endoplasmic reticulum, where there was some change in protein location as proteins were relocated to deal with the formation of these granules. But actually, the transcriptome in this study was the major mover, not the proteome.

That's really interesting.

I think one of the things that's coming out to me, and one of the things that we think about in my lab quite a bit, is where is the point of regulation?

And that point of regulation could be at the level of the transcript and what gets transcribed or what gets degraded, how much the steady state level of the transcript is. It could be at the level of how much protein is there. It could be further at where is that protein?

It could be what is the modification state of that protein, whether it's degraded or phosphorylated?

And it may be that any given protein is regulated by one or many of these mechanisms, alongside that next question of who its friends and neighbors are. So changing the. I can change the function of protein a by bringing it around protein B.

But I think we also have to bring in other molecules as well. So rna binding is.

There are other molecules other than proteins.

Other molecules than proteins. So the rna binding proteome is the largest functional class of proteins. I mean, there are a lot of proteins which bind rna.

Some of them, it's quite clear why they do it and how they do it for a lot. It isn't because they don't have canonical rna binding domains. And then you sort of wonder, why are they binding rna? Is it, whose benefit is it for?

Is it for the benefit of the RNA because the RNA is being protected, trafficked, whatever? Or is it the benefit of the protein that the rna is actually manipulating? Protein activity, protein function.

This sort of question, just to go a slightly different direction. One of the things that we saw in my lab were with proteins like agr two and trop two.

These were proteins that were canonically membrane proteins and had one function.

But then when they were cleaved, they might have a completely different function, change the cell behavior, drive metastasis in a completely different way when they translated to the nucleus, again, completely different function. And so do you think, from your experiences and observations, does every protein do this?

Is this a rarity for proteins to have three different activities in different places? What is the scale we should be thinking about this as.

I honestly don't know. I don't know. And I think it needs a community wide effort to really look at this.

So I, with my interest in moonlighting proteins, have followed closely the work of Constance Jeffery and her group, and they have a website called Moonprot. It's now Moonprot 3.0. And you can go onto this website and type in your protein of interest.

And if they've been able to determine or go into the literature and look to find empirical evidence of proteins having multiple functions, there, there it is. So their recent version, which is 3.0, they've added 500 proteins from 2.0. So this is a real labor of love.

You know, I think they're doing an amazing job, but to trawl through that amount of literature to find evidence of proteins having multiple different functions is, it's a gargantuan task. So how do we, how do we, how can we capture proteins in, in different geysers?

And I suppose a lot of information is out there already, but it's got to be, it's got to be curated, and there are many different consortia who are doing this.

So, uniprot next prop from the Swiss Institute of bioinformatics are using natural language processing, deep learning methods to really trawl through not only the literature, but trawl through sequences, just to try and get an idea, even from linear sequences, what proteins could be capable of. Are they actually falling into different families? And therefore, is there more to.

More to than immediately you see when you look at a protein sequence, of what a protein can be capable of? So I can't answer your question. I have no idea.

In some cases, it's like the rna binding protein, it's going to be obvious if they're chopped up that, you know, they're going to perhaps be fold differently, make different interactions and be trafficked to a different part of the cell and they'll have a different function. But for a lot of proteins, it's not going to be obvious that they can shape shift and moonlight.

So, Catherine, one of the things that comes about when we think about this complexity, about what protein function means, is we go back to what is a gene, and we have so many genetic variants and markers that we've discovered over the years. How do we think about that? How do we think about these things that are seemingly quite predictive, but don't account for all of this complexity?

Well, they may be predictive, but they don't give us the mechanism necessarily.

So maybe if you're blessed with a certain set of, or a certain genetic makeup, it may predispose you to some disease, but we don't necessarily know how the genetic makeup translates into the disease.

So something about sort of. It's an association, but it's not unveiling the process.

Yes.

Yeah, yeah. I think one of the places where that comes up is thinking about things like, oh, this disease is caused by a mutation in p 53. Okay?

And so there's a mutation in p 53. Is it changing its DNA damage sensing mechanism? Is it changing its DNA repair mechanism?

Is it changing how it interacts with five other downstream proteins? So I think there's associations, and the associations are valid. They've been shown a thousand ways, but they're incomplete.

Yeah, I agree. I agree.

So I think there's a broader question here about how do we teach others about protein function when it is so complicated?

There's an initial reaction to be like, oh, that's too complicated, I can't handle it, I'm out, I'm going to go study something simpler, or, I just don't understand. How do we teach people about protein function?

Well, I think we teach it at different levels. We have to teach what proteins are capable of. And some of that is very obvious. You know, catalytic activities, have they got a Rossman fold?

Have they got a kinase domain? Have they got membrane spanning regions?

But then I think we have to teach the fact that they are highly dynamic, that they can be modified, that one gene doesn't equal one protein, that there are huge numbers or potentially huge numbers of different protein entities.

And then I think we have to teach technology possibly better than we do about what technologies are capable of and where the gaps in technology are, because there's going to be some really, really bright people coming up.

Well, there are really bright people coming up in, in the generations below, as I have more generations below me than you do, who are, you know, are going to be the tech, are going to be the next technologists that are going to really develop. Some of it will be empirical, a lot of it won't be. A lot of it will be AI.

And I think, embrace new technologies and not be scared of new technologies as well. Also, something which I think is not always very well received is the idea of science, apart from hypothesis testing.

Because if you test a hypothesis, you're testing the same thing over and over again.

And if you test the same thing over and over, the same sets of proteins over and over again, you may learn about every different shape, modification, location, and everything to do with p 53 at the expense of a great proportion of the proteome.

There's an analogy that I give in class to explain systems biology to people, which is I show a picture of a student of mine, Kim, it's not her real name, but I show her when she is just in class and looking like an average student.

And then I show her when she's interacting with a separate entity, her boyfriend, Brody, and she's all dressed up and she's wearing a fancy dress and he's in a tuxedo. It's hard to. These entities behave very differently.

And then you put her in a triathlon outfit, and then she's fierce and terrifying and yet a different entity, so still the same person, but the context completely changes what you would measure about. About her, what you would observe, the behavior. And I think that's likely true of proteins as well, is that that context just matters so much.

I agree, I agree.

So I think this is more of a call to funding agencies that hypothesis generating experiments are extraordinarily worthwhile and needed infrastructure and support to give the sorts of data at scale that we require as a community, really, to start delving into the issues that we've been discussing.

Well, I think that's a very profound statement, and something we really need to focus on in the community is both doing the right experiments to collect an amazing trove of data, analyzing it in the most elegant and sophisticated way we can, and then communicating it so that it doesn't stay locked in a little bubble, and that the entire biological community can benefit from those learnings. So, Catherine, I think one of the things that this conversation has highlighted for me is just how much complexity there is.

And the further we dig, the more complexity we find.

And so I'm curious why, after digging and digging and digging and finding more and more and more things that we don't know, what keeps you motivated to keep on digging, as opposed to throwing your hands up in the air and saying, oh, my gosh, it's too complicated, I'm out.

I think I want to understand how cells work, and by continuing to add different layers, we will start to understand, I think, what the limits of this complexity are. I mean, the proteome is very complex, but some of that complexity is non functional complexity. So trying to work out what's important and what isn't.

So, an easy example would be phosphorylation, which. Which amino acids, when phosphorylated, has some effect. Maybe a lot. A lot of phosphorylation events is an occupational hazard of being serine.

It's your own fault. You shouldn't have been serine. You're going to get it. It's coming to you. Lysine is even worse. Lysine is a very, very reactive amino acid.

If you're lysine, you've got it coming to you.

So many different ways in which it could be modified, but to actually pick apart what's important and what isn't because the complexity is not going to go on forever. So to perhaps work out what we should be focusing on and what we can ignore a bit keeps me going.

And the more that we can, the more systems wide data that we get, I think in response to different perturbations, the more perhaps we'll be able to work out what the scope is, what we need to be focusing on.

Thank you. That's very incredibly well said. I agree. I think for me, I like solving puzzles and I like understanding how things work.

And when I look at all this complexity, I feel like there must exist some simpler fundamental rules under the hood. Katherine, thank you so much for joining me today. This has been so much fun.

And I feel slightly more terrified of the proteome now than I was when we started this conversation. But also excited.

I don't think one should be terrified of the proteome. I think one should respect the proteome.

Perfect. Thank you again for joining us.

You're welcome.

All right, what do you think? Is protein function more than just about a protein's isolated structure?

Absolutely. But let me mention first, this was an excellent lecture, I think for everybody who wants to learn about proteins and proteome.

This is the 101 lecture everybody should listen to. I think what Catherine builds as a picture of a proteome and for proteins in particular, is that very dynamic environment.

And so proteins have to really adapt to the environment. And what I was, you know, it's not only about protein, the structure in pdms, but also about where they actually.

Are in the cell and how much that changes, just how dynamic it is that they're constantly moving from place to place to place.

Yes.

And I was aware that, you know, kinases, for example, can be promiscuous, but I think the better characterization that she brought up is that they actually have a second job. They're moonlighting.

And so as they're in a different part of the cell, to have a different job, like me being at work, have a different functional role than if I go home, then I have a different role there. So, kind of fascinating.

Yeah.

I think what I really appreciated about her perspective is she studied the trafficking of proteins so much and that they can have completely different behavior depending upon who they're interacting with, where that interaction takes place, and then on top of that other biomolecules, so interacting with rna, for instance, or interacting with a membrane, can completely change the function of a protein.

Yeah. Another part of the conversation that was fascinating is how big is that space?

And I think you can do many types of speculations and calculations, what it might be, what is the number of proteforms?

But I think what was fascinating that she approached that whole topic of how many may they be, how many proteforms may be in the cell, is by purely trying to match her observations to the functional information that she had and that sort of restricted the space of the protein forms that she studied.

Well, ultimately, in my mind, it comes down to what alterations in the protein, whether it's temporal dynamics, chemical structure, in terms of modifications or location, actually have a consequence on cell behavior. So as we go up length scales, which ones actually change how cells grow, which ones change how they respond to their environment?

And many of those variations may not have any consequence at all.

And that's the real question, is amongst the trillions of possible variations, which ones are actually at a particular time, at a particular moment in the growth of the cell, causing any functional consequence.

So one of the takeaways in this lecture was that the environment determines what type of proteins are going to be present and what type of function they have, but vice versa, the change in protein will also change their structure and then create a new environment. Bi directional communication.

Right. The dynamism that interplay that we have these feedback loops.

And so what happens in one place at one time, 10 hours, 20 hours, three years later, may have an impact.

Yeah.

Thank you again for joining us on translating proteomics.

I think what we heard today was so much discussion from both Catherine and Andreas myself about how protein function is much more than maybe we anticipated. And so I'd love to hear from you. What do you think protein function is?

What are the things we should look at and consider when thinking about protein function? Hit us back in the comments and let us know.

Also, when we just think about proteins and states and time and change, how do we wrestle with so much complexity? How do we get to the heart of it?

We hope you enjoyed the translating proteomics podcast brought to you by Nautilus biotechnology. To contact us or for further information, please email. Translatingproteomics at nautilus bio.