Hi, my name is Rob Woods, and I'm going to talk to you today about modelling carbohydrates,

which is some of the work our lab does. And first of all, I thought maybe it's a good idea to explain where carbohydrates are found. Typically, what we're looking at are the carbohydrates that are on the surface of our cells and on the surface of pathogens like bacteria and viruses. For example, in this slide here, we can see on the surface of this endothelial cell,

this fur, which is a combination of glycoproteins and glycans, which are the names for proteins and sugars in biology, you can see it's a very complicated surface. It's also the first thing that a pathogen like a virus or a bacterium sees when you

get an infection. So the first interaction is not with a protein necessarily, but very likely with a carbohydrate. So if we're trying to develop therapeutics to either block this infection or vaccinate against bacteria or viruses, we need to understand the 3D shape of sugars. The 3D shape of sugars is a little different than studying the 3D shape of other biomolecules

like proteins. What I mean is proteins are relatively rigid structures compared to carbohydrates. And this concept of rigidity means that you can model a protein based on similarity to other known 3D structures. For example, if you have a protein that's 75% similar in sequence to another protein,

the likelihood is it will have the same 3D shape. In contrast though with carbohydrates, a tiny change in the sequence can profoundly affect the 3D shape. And this means that we have to have some other way to model 3D shape. And we do this using a computer where we simulate the actual motion of the sugar.

These simulations can run anywhere from a week to two months, depending on how much time you want to simulate of the motion. So for example, if we look here, these are two sequences of sugars from the surface of two different bacteria. The first one is group B streptococcus, and the second one is pneumococcus.

And both of these are infectious bacteria. You can see that the sequences are almost the same, even if you don't understand anything about carbohydrate chemistry. The only difference is that this one has this particular sugar here, and this one doesn't. So in terms of percent of similarity, they're very similar.

But in terms of their 3D shape, they happen to have very different properties. And we can see this biologically in the sense that if you immunize with one or the other of these, you get different immune responses. And the antibodies that recognize them do not cross-react. They don't recognize each other. So they have different 3D shapes. And one of the best ways to study this is with simulations, which is what we see here.

On the left is the group B strept, polysaccharide. On the right is the pneumococcal polysaccharide. And right away, you can see that they have very different 3D shapes. For example, here we can see the structure is basically a loose helix. It's certainly not rigid, but it's not random either, whereas this one spreads over

a number of different shapes, and in fact tends to collapse together quite a bit. And the implication for this is antibodies that recognize this would not recognize this. And when we talk about recognition, it raises concepts that we often don't really think about in terms of flexibility.

After all, how do you recognize a flexible shape? Because your immune system has to do the same thing. And the concept might be, well, perhaps the immune system or the infection recognizes the average shape of the sugar. But if we think of that, if we think of this as being like a snake, if the snake wiggles equally right and left, its average shape is a straight line, which never really

looks like the snake. So we have to be very careful and precise when we talk about shape. And what we really mean is we're talking about multiple shapes, and what's the most likely shape, rather than what's the average shape. And this has implications for the design of drugs to block such infections, as well

as for vaccines. When a flexible molecule binds to a protein such as an antibody, given that the antibody must recognize a particular shape, you can imagine that it becomes more rigid, or that there's a shape something like this. And this is, again, a computer simulation.

This is the same group B polysaccharide now bound to an antibody that recognizes it. And instead of flopping around like a loose helix, what the antibody has done is evolved to recognize the most common shape, which is this helical shape here. Once you know that the structure is relatively rigid, it allows you then to approach the

design of a therapeutic agent or a drug that you might use to intervene in something like this where maybe you don't want an interaction. So for example, the proteins on a bacterium that bind to a sugar, maybe you want to block that in the first place. And so for example, if we look at this, here's the glycocalyx again, and here's

the model of the surface. When you have an infection, a pathogen like a virus or bacteria, the first thing it may do is grab on to the sugar. It grabs on and then infects the cell. And you can think of this as the bacterium having like a hand to open the door handle into the cell.

And one way to block this is to provide some small molecule that looks like the carbohydrate, but that actually binds to the pathogen and prevents it from entering the cell. So you could imagine this is something like blocking the hand so it can't grab the door handle. So once we have an understanding of the 3D shape of a carbohydrate protein complex,

we can start to design molecules to act as drugs to inhibit this. For example, in the case here, we can see a carbohydrate bound to a protein. It's a natural sugar. And by a small change like this, we've made it into a potential drug. Now we don't know how good it is yet until we take it into the lab and test it.

And when we test it, we can go back and we can say, well, how does it compare to the original? Is it better or worse? And if it's better, is it good enough? Or do we have to do a bit more design and then take it back into the lab? So this is one of the most exciting parts of carbohydrate modeling is the ability to

go back and forth from experiment to theory and try to improve and make better drugs. And this is exactly what our group is working on now. So thank you for watching this video. And if you have any questions, please contact us through the website.