People of the Fediverse!
-
People of the Fediverse! I have accomplished science. Allow me to elaborate...
"Discovery of a sulfotyrosine-motif in the human TrkB extracellular domain
required for agonist activation"https://www.biorxiv.org/content/10.64898/2026.05.19.725324v1
#Science #Neurobiology #StructuralBiology #Biochemistry #Pharmacology #SeeMyScience #OpenScience
-
People of the Fediverse! I have accomplished science. Allow me to elaborate...
"Discovery of a sulfotyrosine-motif in the human TrkB extracellular domain
required for agonist activation"https://www.biorxiv.org/content/10.64898/2026.05.19.725324v1
#Science #Neurobiology #StructuralBiology #Biochemistry #Pharmacology #SeeMyScience #OpenScience
Why did we do this?
TrkB is a receptor in your brain. It transmits signals from outside your neurons to inside your neurons and tells them to stay alive, and help reinforce synapse formation.Having neurons is important. Not having neurons leads to neurodegenerative conditions.
Having synapses is important - learning & cognition need synapses and synaptic plasticity.
We know that if TrkB signalling doesn't happen, bad things happen to your brain, and this can be a hallmark of and contributing factor to a host of neurological conditions, from Parkinson's to BPD and beyond.
Normally, a small protein called BDNF activates TrkB, so in an ideal world we'd just make BDNF and give it to people, and that would be problem solved...
However, BDNF is a bastard. Difficult to make. Incredible sticky (hydrophobic) but also highly negatively charged. Insoluble, unless the pH is below ~5. It has probably evolved these characteristics as part of it's regulation, but that means that as a therapeutic intervention, clinical trials have been disappointing.
-
Why did we do this?
TrkB is a receptor in your brain. It transmits signals from outside your neurons to inside your neurons and tells them to stay alive, and help reinforce synapse formation.Having neurons is important. Not having neurons leads to neurodegenerative conditions.
Having synapses is important - learning & cognition need synapses and synaptic plasticity.
We know that if TrkB signalling doesn't happen, bad things happen to your brain, and this can be a hallmark of and contributing factor to a host of neurological conditions, from Parkinson's to BPD and beyond.
Normally, a small protein called BDNF activates TrkB, so in an ideal world we'd just make BDNF and give it to people, and that would be problem solved...
However, BDNF is a bastard. Difficult to make. Incredible sticky (hydrophobic) but also highly negatively charged. Insoluble, unless the pH is below ~5. It has probably evolved these characteristics as part of it's regulation, but that means that as a therapeutic intervention, clinical trials have been disappointing.
So what do we do when we want to mimic a thing without using the thing? We build a mimetic. This was done by our collaborators several years ago, and they came up with a antibody-like biologic that very faithfully mimics BDNF activation of TrkB - we call it ZEB85.
You can read about how it was developed here:
https://www.pnas.org/doi/full/10.1073/pnas.1806660115tl;dr we used a ribosome/phage display-like technique coupled to a functional readout of TrkB to pan 10^11 different sequences to pick a winner.
In our recent work, we figure out how ZEB85 works, compare it to BDNF, and discover some interesting things out about TrkB along the way.
-
So what do we do when we want to mimic a thing without using the thing? We build a mimetic. This was done by our collaborators several years ago, and they came up with a antibody-like biologic that very faithfully mimics BDNF activation of TrkB - we call it ZEB85.
You can read about how it was developed here:
https://www.pnas.org/doi/full/10.1073/pnas.1806660115tl;dr we used a ribosome/phage display-like technique coupled to a functional readout of TrkB to pan 10^11 different sequences to pick a winner.
In our recent work, we figure out how ZEB85 works, compare it to BDNF, and discover some interesting things out about TrkB along the way.
First off, we wanted to understand where both BDNF and ZEB85 bound to TrkB. To do this, we did some biochemistry and determined some x-ray crystal structures.
BDNF binds where it was predicted to. This was no surprise, because there is already a structure of the evolutionarily related TrkA/NGF signalling complex. But we confirmed so ideas about BDNF binding, and were able to show how it differs from other systems. Importantly, getting a good look at the binding surface allowed us to design some mutants in TrkB that render it insensitive to BDNF.
ZEB85 threw us a curve ball. It bound the bit of TrkB inbetween the BDNF binding site and the cell membrane - what we call the extracellular juxtamembrane (eJM) region. The eJM is predicted to be unstructured ( spoiler: it is) but was interesting to us because the TrkB eJM is very different from the TrkA and TrkC eJMs, which helped explain why ZEB85 is specific for TrkB.
Then we wanted to do some biophysics and study some interaction kinetics - but when we made a synthetic TrkB eJM it barely bound (micromolar KDs) to the ZEB85 at all, whereas when we used recombinant TrkB extracellular domain we got sub-nanomolar KDs (very tight binding)
-
First off, we wanted to understand where both BDNF and ZEB85 bound to TrkB. To do this, we did some biochemistry and determined some x-ray crystal structures.
BDNF binds where it was predicted to. This was no surprise, because there is already a structure of the evolutionarily related TrkA/NGF signalling complex. But we confirmed so ideas about BDNF binding, and were able to show how it differs from other systems. Importantly, getting a good look at the binding surface allowed us to design some mutants in TrkB that render it insensitive to BDNF.
ZEB85 threw us a curve ball. It bound the bit of TrkB inbetween the BDNF binding site and the cell membrane - what we call the extracellular juxtamembrane (eJM) region. The eJM is predicted to be unstructured ( spoiler: it is) but was interesting to us because the TrkB eJM is very different from the TrkA and TrkC eJMs, which helped explain why ZEB85 is specific for TrkB.
Then we wanted to do some biophysics and study some interaction kinetics - but when we made a synthetic TrkB eJM it barely bound (micromolar KDs) to the ZEB85 at all, whereas when we used recombinant TrkB extracellular domain we got sub-nanomolar KDs (very tight binding)
So what's going on? We looked at the eJM sequence and figured out that there were a couple of possibilities for a post-translational modification (PTM).
PTMs are cool. When mammalian extracellular proteins are made, they pass through the golgi network (remember the weird blobby bit inside your cells from high school biology?). The golgi contain all manner of enzymes that modify (post translation from mRNA) your proteins. These modifications can alter the form and function of the protein.
The eJM sequence we were interested in had a threonine in it, so our initial thought was an O-linked sugar, but there were also some tyrosines there... inside the cell tyrosines frequently get phosphorylated by tyrosine kinases to propagate signals - TrkB is itself a receptor tyrosine kinase, althouigh TrkB's kinase domain is inside the cell and the BDNF and ZEB85 binding sites are outside the cell.We sent some protein off for mass spec analysis - and lo and behold that part of the protein was 80Da (80x the mass of 1 hydrogen atom) too big, that that extra mass was on tyrosine 400. Extracellular kinases do exist (VLK), but they hit quite specific sequence (GYP) and our sequence was YEDYG.
-
So what's going on? We looked at the eJM sequence and figured out that there were a couple of possibilities for a post-translational modification (PTM).
PTMs are cool. When mammalian extracellular proteins are made, they pass through the golgi network (remember the weird blobby bit inside your cells from high school biology?). The golgi contain all manner of enzymes that modify (post translation from mRNA) your proteins. These modifications can alter the form and function of the protein.
The eJM sequence we were interested in had a threonine in it, so our initial thought was an O-linked sugar, but there were also some tyrosines there... inside the cell tyrosines frequently get phosphorylated by tyrosine kinases to propagate signals - TrkB is itself a receptor tyrosine kinase, althouigh TrkB's kinase domain is inside the cell and the BDNF and ZEB85 binding sites are outside the cell.We sent some protein off for mass spec analysis - and lo and behold that part of the protein was 80Da (80x the mass of 1 hydrogen atom) too big, that that extra mass was on tyrosine 400. Extracellular kinases do exist (VLK), but they hit quite specific sequence (GYP) and our sequence was YEDYG.
Some extracellular proteins get sulfated on tyrosines (sY) - sulphate and phosphate have very similar masses (almost indistinguishable) - this is a rare and difficult-to-detect modification - most protocols that might detect sY often start with a hash acid step, and the sY is acid labile - so you destroy it before you can find it. We had an inkling though
So, we made some more synthetic peptides, this time incorporating the sY modification and we were able to recapitulate the sub-nanomolar KD for ZEB85.
No sulfotyrosine - no binding - no activation.
We were able to mimic this in recombinant protein by making a Y400F mutant that lost all binding to ZEB85. We also did some other agonist engineering stuff and showed that the spacing of the sY binding sites is of critical importance as well.
What it interesting is that in characterising an agonist, we learnt something fundamental about TrkB biology, and this might in turn lead to better agonist design and hopefully TrkB-activating therapeutics for neurological and neurodegenerative conditions. It also might lead to new TrkB biology, although that, dear reader, is a story for another time...
-
Some extracellular proteins get sulfated on tyrosines (sY) - sulphate and phosphate have very similar masses (almost indistinguishable) - this is a rare and difficult-to-detect modification - most protocols that might detect sY often start with a hash acid step, and the sY is acid labile - so you destroy it before you can find it. We had an inkling though
So, we made some more synthetic peptides, this time incorporating the sY modification and we were able to recapitulate the sub-nanomolar KD for ZEB85.
No sulfotyrosine - no binding - no activation.
We were able to mimic this in recombinant protein by making a Y400F mutant that lost all binding to ZEB85. We also did some other agonist engineering stuff and showed that the spacing of the sY binding sites is of critical importance as well.
What it interesting is that in characterising an agonist, we learnt something fundamental about TrkB biology, and this might in turn lead to better agonist design and hopefully TrkB-activating therapeutics for neurological and neurodegenerative conditions. It also might lead to new TrkB biology, although that, dear reader, is a story for another time...
@xtaldave when I say my job is to build instruments to allow people to science, it’s exactly this kind of thing I mean
I don’t know what most of these words mean though
but I’m happy to keep it that way as it means more mental capacity for the things I do need to know -
F folfdk@helvede.net shared this topic
