Subscribe
Share
Share
Embed
Proteomics in Proximity discusses the intersection of proteomics with genomics for drug target discovery, the application of proteomics to reveal disease biomarkers, and current trends in using proteomics to unlock biological mechanisms. Co-hosted by Olink's Dale Yuzuki, Cindy Lawley and Sarantis Chlamydas.
Welcome to the
Proteomics in Proximity podcast,
where your co-host
Cindy Lawley and Sarantis Chlamydas
from Olink Proteomics.
Talk about the intersection of proteomics
with genomics for drug target
discovery, the application of proteomics
to reveal disease biomarkers,
and current trends in using proteomics
to unlock biological mechanisms.
Here we have your host,
Cindy, and Sarantis
What if your chronological age
only told part of your story?
What if your brain,
your lungs, your ovaries?
What if they were all aging
at different rates in your body,
and that you could know that you could
find that out from a simple blood test?
Today's guest,
Tony Wyss-Coray from Stanford University.
We'll talk about just that.
He's one of the pioneers who's building
organ aging clocks out of simple Olink.
Protein measurements.
Tony and his team at Teal
Rise are revealing remarkable insights
into biological age,
organ health, and even future
disease risk years before symptoms
at all.
Of course, we're delighted to have
our customers get more out of the data,
so we're partnering with a team
at Teal Rise to ensure
customers do just that.
In this episode,
we talk about young blood vampires,
immune cells, organ aging clocks
and so much more.
I hope you enjoy it.
Hello everyone.
Welcome back to Proteomics in Proximity.
We're very excited to have, guest.
We haven't had before here to talk about
aging and some amazing innovations
happening in aging.
Sarantis,
do you want to introduce our our guest?
Thank you.
Thank you Cindy. Welcome everybody.
We are excited and honored to have with us
a leader in the field of aging research
professional, Tony Wyss-Coray.
Tony is a professor of neurology
at the Stanford University School
of Medicine, director of the and Benny
Knight Initiative for Brain Resilience,
where he leads biomedical research
on aging and age related diseases.
His work focuses on blood proteomics
and circulating factors
and how they influence healthspan
and lifespan, with a goal, of course,
of developing interventions
to help us live longer, healthier.
Tony is a pleasure to have you with us.
To start, we would love to hear
a little bit more about your journey.
What led you to aging research
and what continues to inspire your work
every day in the lab?
Thank you.
And it's hot, hot, hot right now.
This aging stuff is on fire.
It's so exciting.
Yeah. Thank you.
Thank you so much for having me.
Yeah, I was when I
started out in research,
I started actually immunology.
I really had no interest
in aging research, to be honest.
But, I got into neurodegenerative research
and studied
how the immune system affects the brain,
and particularly in Alzheimer's disease.
And, I realized
that age is really the key risk factor
for this disease.
And we, did more than 20 years ago,
some very simple experiment,
where we asked patients,
for blood samples, and,
could we see changes in immune, factors
in the blood
of these patients with the disease
compared to those without?
We saw some differences,
but we realized that in the healthy
controls,
the changes are actually often bigger.
And so there were a lot of changes in the
composition of proteins that we measured.
And at the time, it was actually a filter
area where we measured 120 proteins.
And these changes in healthy
people are often
bigger than between disease and control.
And that really got me
into aging research.
Meaning that over time,
the longitudinal aspect
showed changes despite them
not having any apparent disease,
but that there were changes
that happened longitudinally that were
because so much of the big population
health projects are one sample, right?
The baseline sample, like in the UK
Biobank data, which went into the paper
that we certainly want to talk
about the organ aging paper today,
but those are baseline samples.
And so this was really about longitudinal
sampling way ahead of its time
I'll say 120 proteins. It's a lot.
At the time
it actually wasn't longitudinal.
It was still cross-sectional.
But what we saw
is that people who were younger
had different concentrations of proteins
than those who were older.
So this was cross-sectional, but
it just showed that, you know, age
seems to be associated with changes
in concentrations of proteins.
And, you know, now this makes total sense.
And we know, of course,
you know, most molecules
change in concentrations with age
and in organisms
and certainly function changes with age.
We look different as we get older.
So the fact that we can capture this
in the blood, provided us
with an opportunity to start asking,
are these changes cause or effect?
And we can talk a little bit about what
that led to?
Yeah, definitely
want to talk about causality for sure.
And I'm guessing that this is what drives
you later on to parabiosis experiments.
Right.
Because it was a big actually discussions.
They make a lot of, impact
in the community
also of these parabiosis experiments
parabiosis
Yeah. Let's explain. So exactly.
So Sarantis is actually absolutely
right.
So this was really the
motivation for
us to get, into, into this model.
And so basically what this is,
is it's a model where you can,
pair animals,
of, you know, different composition
or different age or different geneticbackgrounds.
And in this case,
and it's called parabiosis
So it means living next to each other.
But essentially
when you use this model, the,
the circulation of two mice in
this case is shared.
And so we paired young with old mice.
And we did this actually in collaboration
with Tom rando, who,
you know, in a seminal paper
with the Con Boys,
showed that he can change
the age of muscle stem cells.
So use an old mouse, which has very low
stem cell activity in a muscle.
He paired it with a young mouse
and could show that
this regenerated
and rejuvenated the old muscle.
And so he had this model
and he actually recruited me to Stanford.
So we asked
what is the effect on the brain?
Because we saw these changes
in blood composition with age.
And again, this allowed us to ask,
is this cause or effect?
Are the change a reflection of the aging
organism, or do they even influence it?
Turns out they do both.
And this is really the basis of now
the diagnostic potential of this,
and the
implications for biology in general.
But what we did is we basically used,
the same model
that Tom had used and asked,
is it the brain of a of an old mouse?
Affected by exposure
to a young circulation?
So young blood factors.
And that was indeed the case.
So we could rejuvenate the old brain.
These mice, had less inflammation,
more stem cell activity.
But then what we also showed
for the first time, actually,
if this was Saul Villeda, who has now
his own lab at UCSF, he showed that he can
simply take blood from young animals
and repeatedly infuse it into old.
So he did this
every three days for three weeks.
He infused small amounts
of just the liquid part
of the blood, actually, not the cells,
just the plasma, and could show
that this reproduced pretty much
the effects of this para bias model.
And, most importantly
it improved function of these mice.
So when mice get old,
they get cognitively impaired,
they have difficulty navigating a maze.
And when they're exposed to young plasma,
they can navigate it better.
And their function improves.
And so that proved to us
that there are factors in the blood
that can not just that,
not just reflect the aging
of the organism, that they can actually
change the age of an organism.
But this is amazing.
I mean,
from the way that I understand this,
like these factors,
they target specific
cells in brain, right?
Because at the end, the brain seems that
this is the the final,
actually hit on these on these factors.
Do you know what type of cells
are you targeting with these factors?
Are we aware
I'm guessing more of them than
most of them are stem cells now, right.
Yeah, that's a great question.
So initially you can imagine
when when we first reported these results
and also when Tom reported these results,
people thought that's too good to be true.
How could young blood
just have these effects?
That's folklore.
Right?
People thought,
you know, give me Young Blood and Dracula.
Exactly. Little Dracula stuff.
But then with the advance of single cell,
genomics,
where we can take cells
basically of every,
organ tissue in the body,
of a mouse,
we can study how
every cell is affected by these, factors.
And so because we saw these effects,
I was really interested
in, in harnessing the,
the transcriptomic approaches
and worked with Steve Craig,
who was one of the pioneers
in developing these, technologies.
First genetic diagnosis, first genetic
diagnosis made by Stephen Quake on himself.
Yeah, yeah, yeah.
That's it.
Yeah.
So we, we built this atlas,
we called Tabula Muris,
where we profiled
the cells of all major organs.
And then we used this approach
to ask in a very unbiased way,
how is, how are the cells
in the parabiosis model affected,
which cells, as you asked Sarantis,
which cells are most affected?
Globally, throughout the organism,
we find that, stem
cells are a major target,
especially hematopoietic stem cells,
but also, hepatocytes, for example.
And as you know, hepatocytes produce,
the majority of proteins,
at least in quantity,
in the blood, like albumin and coagulation
factors, component factors and so forth.
But what was really also exciting
is that almost every cell responds
to, this hetero chronicity.
So whether you give an old animal
young blood or a young animal old blood,
which accelerates aging, most cells show
transcriptional changes.
And you can then ask
what are key pathways.
So one of the key
pathways are mitochondria.
For example, there's a general, reduction
in inflammation,
from young plasma to to old animals.
And I think with respect to the brain,
we don't know
exactly how these factors
get into the brain,
but there is sort of, reduction
in inflammation in the vasculature.
So for that, factors
don't even have to go into the brain.
And that may be one prominent way
to get benefits.
But we also know now from other studies
that Andrew Yang, in my lab,
you know, is at the class on institutes,
these simply labeled, proteins,
in the plasma and injected them into mice
and sees that they're actually broadly
taken up into brain tissue,
you know, unexpectedly, you know, people
always think there's this barrier.
Nothing goes and this is totally not true.
So we're
now following up on different cell types.
Neurons
take up proteins from the circulation,
but we pursue
a very specific type of microglia
that are specialized
in taking up proteins from circulation.
So I think there's a whole biology
of how this might actually work.
But, yeah.
But I think the key message
was really that factors in the blood
can modulate aging.
And so let's define hepatocytes. Right.
These are the functional cells
in the liver.
And as you said they they are incredibly
important in producing proteins
that show up in the blood.
So I think that that leads us down
a path of
proteomics that you've explored
since then.
Can you tell us about how you got to
where we are today?
And in particular, I'm, I'm pretty excited
to talk about the origin organ
aging work that, that your team
published in Nature Medicine.
So this is plasma proteomics
links brain and immune system.
Your, your,
your foundation in immunology, which is,
makes complete sense that
that the immune system is key to this,
and healthspan and longevity.
So maybe it's useful to define healthspan
as opposed to lifespan
and then talk a little bit about the path
that led you to, work
with the UK Biobank data and build
the organ aging clocks that we,
we're excited to see leverage today.
Right? Yeah.
So maybe just the definition of health and
and lifespan.
So, healthspan is, is described as the,
the time of your life
where you're generally healthy
and you don't have any major diseases.
So now I think the field sort of describes
that we live, roughly 80 years of age.
You know, in, in sort of,
economically developed countries
with good health care systems.
And of those 80 years, roughly 50,
55 years where were totally healthy.
The average population.
Right?
So there's of course, always people
who get sick at different ages of,
or different times of the, of their life.
But, the 55 years
would be the health span.
And then we're getting more
and more diseases as we get older.
And some people call this sick span.
But the goal, really, of,
efforts in this field of, aging research
is to extend the health span
so that you lis healthy until you die.
Ideally. The sled ride to the bottom.
Yeah. Like, just I just want it to.
Yeah.
You fall asleep and or you say, I'm tired.
I think I've seen it.
You know, you're you're. 80
and you're done 90.
Years old and and that would be it.
I mean, that would be a nice, way
to, to die
I think, rather than suffering from
multiple degenerative diseases and,
I think the worst for me,
but for many other people,
would be to lose
your cognitive abilities
and don't even know,
you know, who your children are, who your
your loved ones are, things like that.
Yeah.
Yeah, those big ones.
Cardiovascular disease,
cancer, Alzheimer's disease
or neurodegenerative disease.
Exactly.
Those are the biggest risk for getting
those living long enough to to get those.
Yeah. That's true. Yeah, yeah.
But maybe getting back to
how did we get here.
So having having shown that,
the blood is really sort of
able to affect the aging
or even the physiological state
of cells in the body.
Of course, we wanted to know
where do these factors come from?
That was one of the motivations
to do the single cell transcriptomics
and ask, well, which cells are potentially
producing factors,
but also which cells
are targets of these factors.
And so the question was
then more generally,
if we look at the changes in the blood,
can we ask where these factors come from?
And that's that.
Tell us something about the state
or the physiology
of a given cell or organ.
And that's really a very trivial question
if you think about it.
And we pursued this for decades.
You know, in clinical chemistry,
if you go to the doctor,
they take a blood sample and they measure
a number of different analytes.
They measure lipids.
They measure also a few proteins,
very few, actually,
but one, you know, very prominent,
to come back to the liver protein
that is measured in our transaminases, liver transaminases
They tell you
if the liver is damaged or not.
So if they go up, that means
your liver is not really functioning well.
But we've used this mostly in medicine
to record abnormality,
to record pathology.
So people
go to the doctor when they feel sick.
And then the doctor says, yeah,
I can see your liver.
Your liver values are too high
or you know, there's something wrong
with your heart
based again, on on a couple of proteins
that are measured from the heart.
We measure,
of course, some lipids, cholesterol,
that's a prognostic factor.
But we have very few factors
that really look in the future.
So we asked, you know,
can we get information from,
the brain in the circulation.
So are there proteins in the blood
that might be derived from the brain?
And what was really sort of a revelation
for us if we once we started to use
these multiplex platforms like the one,
you are, developing
or have been developing a link,
what we see is that there's
lots of proteins in the blood
that are not made as secreted proteins.
Remember I said we started out
by looking at immune factors in the blood
because I thought, well, the immune system
sort of communicates across the body.
So we may find some cytokines
and chemokines
that tell us
about the state of the organism,
but it turns out you can literally measure
almost every protein
that the body produces in the blood.
If your assay is sensitive enough
and one of the proofs
for that statement
is that we can now measure a-beta
or tau with the key pathological hallmarks
of Alzheimer's disease in the blood.
If you told anybody 20, 30 years ago,
they would have said, that's not possible,
that won't be too noisy.
We will not be able to pick this up
because there is no relationship.
But it will also not be, measurable.
And now we, you know,
we can measure transcription factors,
we can measure kinases.
Any protein seems to make it
somehow into the blood that,
could then potentially tell us something
about where does that protein come from?
And, if we know that,
does it tell us something about the cell,
where it comes from or the organ,
the organ that it comes from?
Yeah. So so plasma.
I've thought about this a lot,
and I'm curious because I don't
I don't necessarily think of plasma as
or even the organ aging clock which,
which we'll talk about in this paper
as, reflection of the tissue state
as much as a reflection of,
you know, secretion, leakage,
turnover of cells, signaling
between cells, maybe between organs.
But but
maybe because of all of those activities
and the way they change with age,
then maybe
plasma does give you
tissue state. Exactly.
Yeah. I think,
you know, intuitively,
we think about the proteins in the blood,
maybe like you just stated, right,
that they're a reflection of some
either secretion or damage,
but it seems not to be the case
because in a young, healthy organism,
you find any type of proteins,
again, cytosolic proteins
that were never meant
to be secreted, synaptic proteins.
And I think they're a reflection
of cellular,
activity.
And many of these proteins
are maybe the majority, certainly
of intracellular proteins are released
through some vasicular mechanism.
Exosomes.
Maybe there's others that we haven't
really characterized that described well.
But exosomes is,
you know, a key part of this.
And that's part of, a communication system
that people start to discover
and describe.
So, you know, immune cells,
in particular, macrophage
type cells,
they release tons of these exosomes.
And maybe it's a part of communication
with, with the rest
of the body cancer cells, of course also.
But it turns out every cell can release
exosomes
and people that now describe markers
in these exosomes that tell them
the exosome comes from a neuron
or the exosome comes from a macrophage,
or it comes from a liver or a tumor.
So or a cancer.
When we
when we then apply the plasma proteomics
and measure the concentration
of proteins with your tool.
You know,
we, we look at all the proteins
that are present in the blood,
whether they're in a vescicle or not.
And so we really have all this information
wherever it comes from.
Some proteins are secreted.
They're meant to be secreted as a,
you know,
an endocrine, communication.
But and others are shared
from the surface.
And, if a protein is shed,
it may also have a physiological function.
And yet others come through these,
you know,
access
exosomal mechanisms and other parts.
And of course,
some come from damaged cells.
But I think it's it's really a combination
of, anything that happens in the body.
And I really,
state
now, in all my talks, the blood,
is an endophenotype
of organ physiology.
That's great.
That's amazing.
Yeah, that's a good point, Tony.
I mean, stay on this.
And I want to follow up on that question.
Now, having technology
that Olink democratized proteomics
and we can identify more, much
more proteomics and low abundant proteins,
now we are able to predict right
to have biomarkers with predictive power.
You mentioned p tau right now.
It's already for late stages.
I mean, already for your sick.
I think now what you bring in the field
and how you, you are the pioneer in
the field is like having predictive
biomarkers at the level of organ.
And that's really super powerful
for precision medicine.
Yeah.
How how would you see this moving forward?
Yeah, I think I think it
has the potential really to,
change the way we,
we do medicine and apply
sort of this concept of,
you know, blood pressure and cholesterol,
which are predictive markers
to, to many more organs by simply
looking at molecules
that are derived from these organs.
And if they change with age
and physiology.
And I really see the state of an organ
reflected in the age related changes.
So in other words,
if the function of an organ changes it,
it is because the molecular composition
of the organ changes
and we can measure that in the blood.
So in other words, we can measure how,
your liver changes with age,
how your heart changes the age,
how your brain changes with age.
And if it shows an accelerated aging,
it is more likely to develop what we call
disease.
Now, disease is a human construct, right.
And if Vadim Gladyshev, you know his analysis
of,
the same data we analyze, his
conclusion
is that all these diseases that we give
specific names are a reflection
of accelerated organ age.
A very bold statement, but I think
you might be right for many of them.
Of course, if you have to pass,
a pathogen that causes a disease,
then that would be exempt.
But these chronic diseases
of dysfunction of an organ
like Alzheimer,
they're a reflection maybe of,
the aging of that organ
and the failure of that function.
We give a name, we call it a disease.
Now, we know, of course, there's
many different diseases in the brain
and there or the brain age,
for example, is not a good predictor
of Parkinson's disease.
But I think that's just a question
of resolution.
As we get more and more proteins
that are specific to different brain
regions, for example,
we may pick up a signal
from the substantia nigra
or from specific neurons
that degenerate in Parkinson's disease.
And that will give us information
about Parkinson's disease.
And in fact,
we just showed this recently with measures
that, estimate the age of
cell types, not just organs,
based on changes with age
and the age of muscle, of muscle cells
is a strong
predictor of future risk for ALS,
again,
based on 50,000 people in the UK Biobank.
So we have a blood sample from people,
when they enter the study healthy
between age 40 to 60,
we measure their blood proteins
with the Olink platform
and we derive signatures of different cell
types.
We can derive about 40 different
cell type signatures
based on proteins
that are cell type specific.
And we can measure in the blood.
And so we can estimate
how old your skeletal myocytes are.
Your muscle cells.
And if they show an accelerated
aging phenotype,
you're three times more likely to develop
ALS in the next 15 years.
So this brings us to the paper
where you leverage the UK
Biobank population, 54,000 individuals,
and developed organ aging estimates
relative to known chronological age
and demonstrated
what you've demonstrated before that aging
is heterogeneous, meaning it changes.
It's different across different organs,
maybe more heterogeneous in some people,
over others, depending on on age,
probably as well,
but that some of these nodes are,
are dominating outcomes of longevity.
Healthspan likelihood of death.
How would you how would you explain that?
Yeah.
So what what you're bringing up
is a very interesting point
that when we. So again,
we built these models,
we measured proteins
that are derived from specific organs.
We have about 11 organs
where we have good enough signals.
And we can measure proteins that are
produced in the liver, but not in
other organs or mostly in the liver.
And then if we add 5
or 10 proteins together
that are all enriched in the liver,
and we can measure in the blood and
they change with age, we can build a model
that estimates the age of the liver.
We can
then apply this model to every person
and ask how?
How well does our model predict your age?
And the deviation from that age,
which we call an age
gap, is a reflection of faster
or slower aging of your liver.
So if somebody has an accelerated
aging of the liver,
they're more likely
to have cirrhosis later on and so forth.
So we get these organ specific diseases.
But then when we look at these
50,000 people and we ask,
how many
people have accelerated organ aging
and then how many people have one
accelerated organ age or so,
one organ that is aging, two or 3 or 4?
We find most of the people
have only one organ,
showing accelerated
aging consistent with this concept
that Mike Snyder came up,
which he called ageotypes.
So it seems like you have one organ
that seems to get sort of out of control,
and that may be the disease
you develop later on.
But then we do
find people who have multiple organs age,
organs older,
and we can predict mortality.
So we can
then ask which organ is the strongest
predictor of, of future death.
And, it turns out that an old brain
and an old immune system
are the strongest risk factors.
And this is because we've
we collected those samples
almost 20 years ago, and we know who died.
We know who got what diseases we've got.
Those data in those outcome data
from the UK Biobank.
And so the strongest predictors of death
are brain and the immune system.
Yeah. In this UK Biobank population.
And what was even more surprising
is if you have young brains
and young immune systems,
you live actually longer.
It's not a traumatic, increase
in longevity, but it's clearly significant
And it's young
relative to your logical age.
And then you're looking at this across,
you know, a normally distributed
population, presumably, and looking at,
you know, the mean of that.
Yeah.
So the way you do
this is basically you can imagine like
like anything
height is always a good example, right?
If you look at the population,
you most people who are similar size.
But then you have some people
are very short and some are very tall.
So similarly,
if we predict the age of your of brains
and 50,000 people, most people's
brains will match their actual age.
So if you're 50, your brain
is predicted to be 50 years old.
But then some people have a little bit
younger, a little bit older.
It's not normally distributed
all the time.
So you often get a skew towards older
because
and that makes maybe sense
because our organs get older all the time.
But we
really look at the outliers
at the very tail.
So, you know, those with the very youngest
and those with the very oldest.
And that's usually only 1
to 2% of the population.
You can make this cut arbitrarily
anywhere you want.
But we have decided to really
look at this top 1 to 2% .
And that's where you find the strongest,
effects or risk predictions, right?
Like we know if you smoke, your risk
to get lung cancer increases
by, you know, two fold.
Here we find if you have an older brain
you risk to develop Alzheimer's increases
three fold.
And if you compare those with the youngest
brains and those with the oldest,
there's a 12 fold difference
in risk to develop Alzheimer's. Wow.
And and what about protective.
What what about the other side
of the curve?
The those folks that are outliers.
On the younger side.
Yeah. Yeah.
So again
what we find is those with young brains,
they live longer and have very reduced
risk to develop Alzheimer's disease.
And and in a study that will be published,
next month, I think
in Nature Medicine,
we made similar, predictions
at the cellular level
as I, as I indicated earlier.
So we can
build models that estimate the age
of a cell in your brain
or a cell in your muscle.
So you get more resolution,
basically not just a.
Cell at the cell level.
You get resolution of the cell level,
actually.
Yeah. Yeah. That's amazing.
That's something.
And so there we find a cell type
called astrocytes.
So this is a very important cell
type in the brain
that is involved in glucose transport
from the blood to neurons.
It's involved in,
in modulating synapses,
and, making them work
and actually building them.
And we can estimate the cell of the cell
type in every person based on protein
measurements in the blood.
And what we saw
is even a more dramatic risk
association with Alzheimer's disease.
Particularly it's interesting.
In individuals who also have
the genetic risk factors called APOE4.
So that genetic factor is, a predictor
of future Alzheimer's disease,
similar to having an old brain.
So it's about a three fold increase.
And if you have two
genetic copies of the E4,
it's a 10 to 12 fold increased
risk for Alzheimer's disease.
But what we find if you're also unlucky
and you have very old astrocytes,
that increases by a further three fold.
Wow, wow.
So we find actually in the UK Biobank
that almost 40% of the individuals
who have the genetic risk factors
and old astrocytes,
almost 40% will have Alzheimer's
disease over the next 15 or 17 years.
Which of them do
you think is more predictive?
The genome genetics, proteomics,
combinationcombination
It goes I would like combination.
I would like to highlight here
the value of having proteomics data,
because so far we're really focusing
on genetics data. Right.
How would you see this moving forward.
And like a follow up question is
to the clinical relevance.
Would you see the clocks to be like,
a nice tool
to have in patient stratification
for clinical trials,
for clinical diagnostic in the future?
Yes, absolutely.
Is moving forward the field.
Before we change from the astrocyte story
and we can we can switch over to that.
I just want to talk about
the oligodendrocytes that you the pathways
that you saw
that were important in brain aging,
which may be to some extent,
a function of which proteins
you were able to measure
because it is a targeted assay.
Yeah.
But this these were proteins
that weren't neurodegenerative proteins
that I typically think of,
you know, that show up and say,
Parkinson's disease, for example,
when you've got these neurons dying.
They were myelin structure.
They were the sheath on these.
Can you just comment a bit about that.
And then I absolutely want to talk about
the genetics, the proteomics, you know,
bringing these modalities together
and that importance.
Yeah it's super interesting.
So when we make these models,
the algorithm,
first we make the assignment
out of the thousands of proteins
that are on your platform,
we make the assignment.
Where do they potentially come from
based on gene expression data.
So if a gene is only expressed
in an in the brain,
then we say, okay,
this is a brain protein in the blood.
And when we when we then ask it,
when we don't develop a model
that estimates the age of the brain,
we know, of course,
which proteins are making that signature,
and we can ask, what is the biological,
role of these proteins?
What what is known about them?
And as, as you just mentioned,
the proteins that make up the brain age
model are largely involved in synaptic,
a synaptic structure
that wraps around synapses and is,
produced in part by oligodendrocytes,
maybe also some other, cell types, but,
it, it
it reinforces this notion
that maybe synapses are the first part
that that declines as we get older
and that get this function dysfunctional.
So was point.
And that's what we miss communication.
And that's what we may be picking up
long before
you get amyloid plaques
and tangles in the brain.
Amazing.
This is also consistent
with parallel studies that we did
in cerebrospinal fluid, where we looked at
which proteins are strongest
associated with cognitive function
independent of amyloid pathology and tau.
And there two we found plasma.
We found synaptic proteins
are the strongest predictors.
And specifically this NP two.
And why w h h three
that we found are very strong predictors
15 years into the future
with a hazard ratio of 15 or so.
So a 15 fold increased risk if you add,
very high levels of that protein.
Amazing. Yeah.
So to come back for what
that all mean and,
and how do we integrate this
with current knowledge?
Genetics is incredibly powerful. Right.
But what we've also learned is that,
there's very relatively few genes
that are really predicting your risk
to get a specific disease.
April E4 is a strong exception.
That link to Alzheimer's disease is almost
unprecedented in any other disease.
So we need something on top of genes.
And that's where proteins are,
because proteins capture where we live.
They integrate many different genes.
And there are that's why I call this
the end of phenotype.
They capture, our life experiences and,
our daily, challenges.
Now, some of these are very noisy,
but the beauty of it is,
if you measure a lot of proteins,
you can pick the stable ones
that provide information on disease risk.
15 years later.
And that's really what
we can call the proof in the pudding.
Right.
Because we can say what your risk is in
an unknown sample from a person.
We've never seen before.
We know nothing about this person
except the concentration.
No lifestyle factors, no clinical.
We don't need. Any other knowledge.
And that answers
your question, Sarantis.
Right.
That means you can potentially translate
that to the clinic
and provide information
to a patient about their future risk.
And, you know, while that might not
be helpful for some diseases
at this point,
I think you could change lifestyle.
And for some there are drugs.
And we you know, we started a company
to really, take advantage of this
and hopefully bring it to the consumer
in the coming years.
Vero Bio Sciences, that is trying to
to harness this information
where, the consumer can then,
you know, have their blood measured.
We give an indication of which organs
show accelerated aging.
And then with advice from a clinician,
the clinician helps interpret the results
and what you can do about it.
Because of course, you don't want to
just know that you have an old heart.
You would
like to know, what should I do about it?
You have to regenerate. Yeah.
And then the beauty is
that you can follow up
three months later and say,
okay, did my intervention do something?
Exactly.
And this is really
what we're working on now to prove that,
these,
these measures are functionally tied
to, to, to
to outcomes and, and specific organ.
And so what would be needed
to leverage organ aging clocks
as a surrogate endpoint
and a surrogate endpoint would be,
you know, shifting your biological age
of an organ or overall biological age
would be an endpoint in a clinical trial
to demonstrate effectiveness
of an intervention, either a therapy or,
you know, certainly we do
clinical trials on some, some companies,
support clinical trials on their vitamins.
Exactly.
Things like this.
Yeah. So what what will be required is to,
you know,
keep validating these tools
across the community.
And it seems a lot of studies,
you know, have very similar findings,
and I've shown the value, but there's,
of course, more work to be done.
I think the,
the platforms have to be stable,
what with whatever,
you know, from a technical perspective.
And I think they
they show that they can get
the same results with repeated testing.
And then what we also need to show is that
if you measure these,
if you make these estimates in a,
in a person,
you know, once a month, over a year,
that they're relatively stable, right?
That it doesn't change
based on the breakfast you had.
So again,
we need to figure out what are the,
what are the signatures that, that really
have this predictive value on are stable.
And and we're getting there.
We have such a data set
and it seems they're remarkably stable.
But then they need to be tested in
the, in, in the traditional,
placebo controlled clinical studies,
where you show that they track function
and then you can start using them as a,
as a secondary endpoint, maybe,
and then as a primary end point.
So it's a, a stepwise validation
in the scientific community,
then into, clinical trials.
And they're also stepwise.
But I think we can get there.
In the biomarker world.
Then not only will you envision
a model based on proteomic signatures,
because now I have the impression
that we are moving
from the epigenetic laws
that initially Steve Horvath has launched
with, like the methylation profile
to something more functional
to the to the biology, the proteomics,
then the future will be bringing
these proteomic signatures
in the in a model that you can predict.
This is how you see this moving forward.
And we definitely want to talk about
Teal Rise, another company that,
that is leading the way and
their expertise for building such models.
Yeah.
So the idea is that
that you would be able to make models
that, that predict function
and that, track
with efficacy of drugs and allow you
then to use these models as surrogates.
And this is what, another company
that is really trying
to provide these services that we started
Teal,
Teal Rise, is
is partnering with,
with Olink to provide that service.
Not everyone has
maybe the tools to,
to develop these models.
And teal has assembled a large, database
that it uses to train the models
and can
then provide all these, services, to,
to really across
any type of disease and prediction,
that, that you can imagine and see how,
how valid they are.
And I see
this is you know, hastening us to that,
that state Barrett, by which we do
see these as surrogate endpoints.
We do see these as ways to confirm
and corroborate interventions
or even discount interventions.
You talked in the paper, Hamilton,
you know, first author,
and your team talked a bit
about ibuprofen,
about vitamins,
you know, looking at associations.
I just want to clarify,
we talked about causality earlier.
Those associations are are helpful.
They're one step toward understanding,
causality.
But causality is about finding
what are the levers,
what are the pathways?
What are the cells.
What are the, triggers that are actually
moving us toward disease?
And those are the magic that,
that pharma can leverage if they're,
if they meet the right criteria
to build therapies
to nudge someone out of a disease state,
back to a healthy state.
Right.
So the vision here is precision medicine.
Being able to treat the individual
based upon
what their needs are,
to bring them safely back
into a healthy state
and extend their health span.
Absolutely. Yeah.
So as you said,
we have been able in the UK
Biobank to look at,
people's lifestyles, people's,
use of medications.
But this is all retrospective
and some of it is self-reported.
So it's noisy.
It looks very promising.
It looks interesting.
Like I often
mentioned glucosamine
this in our first study
with a different platform
on 5000 individuals.
We had a signal on glucosamine
use associated with younger organs.
We felt
this was not good enough to report yet
because this was a relatively small
number of individuals, 5000.
So the exact same signal in 50,000 people
with the Olink platform.
And we reported it then,
but it's still it's retrospective.
People report I use glucosamine that is
associated with some younger organs.
So now the next step, as you
said, is to do a clinical trial
where you have people
placebo controlled,
they get the supplement or not.
And you do
these measurements of the clocks
and you say, okay, does it.
Actually track with the functional
benefits, does it track with drug use,
and so forth.
And we're starting to do that.
We have a collaboration with Heike A. Bischoff-Ferrari
who ran the DO-HEALTH trial.
This was a 3000 individual trial
with vitamin D,
exercise and omega-3 fatty acids.
Was a four arm trial
to see what control they had, some,
functional outcomes. And,
Steve Horvath showed that
epigenetically some people got younger.
And so we're now, running,
with, Vero and Teal,
we are running the proteomics
to, to say, okay,
is there a signal in the protein clocks?
And indeed we,
we see signals suggesting that, you know,
you can pick up, functional outcomes
and, you get also this,
agreement between epigenetic and proteomic
signatures.
Amazing.
So, so can you tell our audience
where might they track or
keep an eye on some of the interventions
that you might be involved in
if there are clinical
trial calls for participants?
I know Mike Snyder has has this, often,
these QR codes to invite participants
that are of a particular demographic.
Are there opportunities
for people to get involved and support
the work that your team is doing?
Yeah, certainly.
They can, sign up on Vero
for, for updates.
And, we're still going to launch an IRB
approved, beta trial,
so people can sign up for that.
To analyze data sets.
They can, of course, reach out to you
and, work,
with Olink and, and, Teal Rise
and otherwise.
Yeah.
Stay tuned with the publications
that we're working on.
Exciting, exciting.
Well, so happy to have you.
Thank you so much.
We had a little bit of a hiatus
in our, in our, our, podcast publishing.
We had some months off.
This is an exciting episode
to be our first one back.
So thank you so much for agreeing.
Much on it.
Yeah. Thank you.
If we can stay all day asking you.
Exactly.
Exactly how, you know.
I, I, we have to stop at the point
because they have so many questions
popping up all the time.
You know, it's definitely
we will catch up again soon.
Definitely.
Because there’s so many things
we need to know from you.
Thanks so much. Thank you so much.
Well, that wraps up
this episode of Proteomics in Proximity.
Huge thanks to our guests and authors
of such impactful publications.
I also want to thank you for tuning in.
Really appreciate you being here.
If you enjoy the content of this episode,
please think about sharing it
with friends or colleagues
you think might be interested
in the content.
In addition, if you'd be willing
to head over to Apple or Spotify
or wherever you digest your podcasts
and give us a rating and review,
this will help others find the podcast
when they're searching for proteomics
or precision medicine podcasts.
And mostly I want to say
we would love to hear from you.
So we have a dedicated email address.
Pip@olink.com. Please reach out.
Let us know what you're interested
in hearing, about what you care about,
and any feedback on the episodes
that we have already done so far.
This is all about you,
and so we're really keen
to make sure that we're meeting
what you'd like to hear about.
Thank you so much, and we'll see you soon.
In.