Dr. Guido Kroemer on Autophagy, Caloric Restriction Mimetics, Fasting & Protein Acetylation
Hello, everyone. Today I’m
sitting here with Dr. Guido Kroemer,
who is a professor at the University of
Paris Descartes. He is a cell biologist
who has made major contributions to
understanding the role mitochondria play
in cell death. He has also published
numerous publications in the fields of
cell biology, cancer biology,
immunology, aging, and autophagy.
The latter of which, I suspect, we’re
probably going to talk a lot about today
because it’s such
an important topic.
So, autophagy is a
spectacular phenomenon in cell biology,
one that you can see with your eyes
because cells become vacuolated when this
process is induced. So you can see
it by light microscopy and better,
of course, by electron microscopy. It is
a process that consists in sequestering
portions of the cytoplasm of the cell
and then digesting them to recycle the
material and to degrade macromolecules
into micromolecules, metabolites,
and to allow rebuilding the
structures that have been destroyed.
So technically, it works in the sense
that the mouth that is involved in this
self-digestion process is the
autophagosome. So, it will sequester
portions of the cell in the cytoplasm.
It can be an entire organelles including
mitochondria, it can be protein
aggregates, it can be bacteria or viruses
that invade the cell, and this
autophagosome, once it closes, will fuse
with the lysosomes, which is the sort of
stomach of the cell. And in the lysosome,
it has been fusing with the autophagosome,
which is then called Autophagolysosome,
the luminal content
will then be digested.
So you mentioned the digesting
of these multiple organelles or
mitochondria, but also protein
aggregates, and viruses, bacteria,
pieces of chromatin, and things all seem
to be sort of things that at least in the
sense, if you’re looking at the
aggregates, and you know, damage that
occurs in a cell, seems to be something
that’s associated with aging in general.
So it’s sort of like, kind of seems like
it’s getting rid of all these damaging,
potentially damaging, not just aging
but also obviously, an infection.
But this process is getting rid of these
damaging, potentially damaging
molecules and aggregates and
mitochondria which are defective.
What is the actual goal of autophagy is?
Like you said, is to get rid of these
defective things to
then provide energy?
Well, it’s actually
interesting to look at the history of
autophagy. The name comes from “autos
phagy” in Greek which means “self-eating.”
And it’s actually a linguistic invitation
to think about self-mutilation and
self-destruction and cell death. And
actually, the phenomenon is observed
mostly in the context of stress.
So cells, when they are stressed,
will often undergo an autophagic reaction
which occurs before the cells die.
And so, this chronology of the phenomenon
has been also an invitation to think that
autophagy is a mechanism that leads to
cell death until it has been understood
that inhibition of autophagy, which can
only be achieved in a specific way by
genetic tricks, will actually sensitize
cells to cell death induction.
And so, this means that autophagy is a
means of adaptation to stress and a
technique of the cell to avoid cell death.
So, the primary goal of autophagy is
adaptation to changing conditions
and adaptation to external stress,
and at the end, avoidance of the
unwarranted demise of the cell.
You bring up so many important
points that I’d love to touch on.
The first is the external signals
that are actually causing autophagy.
You mentioned it’s a response, it’s
a generalized response to stress.
Like, you know, mostly probably a hormetic
type of stress, but even…this is another
question I’ll have for you later, the
differentiating between the type of stress
that can cause autophagy versus
actually pushing it over to cell death.
But in terms of the actual external
signals that…the main ones that we know
about that cause autophagy, a lot of
them have to do with nutrient sensing.
perhaps the most physiological way,
to induce autophagy that is
phylogenetically conserved from yeast to
primitive animals to ourselves is
nutrient deprivation, starvation,
hunger. And so, the idea is that a
cell that is deprived from its energetic
supply, which can be the absence of
nutrients or the absence of growth factors
that are required for these nutrients to
be transported from the outside world into
the intercellular space or the absence
of oxygen, all these factors can induce
autophagy, and the cells actually will
destroy its bioenergetic reserves,
which are macromolecules,
and ribonucleic acids
to generate energy.
In terms of the energy part, I
was reading about like three really major
pathways that seem to lead to autophagy.
One, being the actual energetic charge of
the cell, like ATP status. When that
lowers, that activates the AMP Kinase
pathway. And then the amino acid sensing
pathway, which then when you don’t have
enough amino acids, that can basically
inhibit mTOR. And then there’s a third
one, the protein assimilation pathway,
which I’m not as familiar with essentially
how that activates autophagy.
So, it’s actually
extremely easy. When you think about basic
biochemistry, one of the central
metabolites is Acetyl-CoA. And so,
the cytosolic pool of acetyl-CoA
determines the level of protein
acetylation for the simple reason that
acetyltransferases, which you see acetyl,
moiety of Acetyl-CoA, to transfer
it on lysine residues in proteins.
Acetyltransferases are having a low
affinity for Acetyl-CoA as compared to
kinases which have a high affinity for
ATP. So if you vary the ATP concentration
in the cell, it has little impact
on phosphorylation reactions.
But if you vary acetyl concentrations,
it has a major impact on the acetylation
level of cellular proteins. So
that’s a major difference. And so,
since Acetyl-CoA is built in the
degradation of glucose, a glycolytic
pathway, from pure beta or in the
catabolism of branched amino acids,
as well as a final product of
beta-oxidation. All major nutrients are
actually supplying Acetyl-CoA as an end
product. And taking away glucose or amino
acids or fatty acids will cause a
reduction in the Acetyl-CoA pool which is
important to note, it is a cytosolic acid
Acetyl-Co-A pool that is accounting for
autophagy regulation. And this reduction
of Acetyl-CoA in the cytosol will cause
deacetylation at the end of cellular
proteins, hundreds of different proteins,
and hence, a sort of multi punked
induction of major subpathways of the
apoptotic process. Autophagy is
actually a very complicated process that
involves dozens, perhaps hundreds, of
different proteins and is regulated by
hundreds, perhaps thousands, of additional
proteins. And so, this common regulation
by acetylation is very efficient in
stimulating the autophagic pathway.
As a side effect of deacetylation
reactions, you usually also observe from
the inhibition of mTOR and the activation
of AMP Kinase. So, everything comes
together at the end. There’s no
exclusivity for one or the other pathway.
They are connected.
Really? So, changing the
acetylation status of proteins affects
mTOR and AMP Kinase?
we don’t know how this actually
works in molecular details.
Because that was sort of, my
next question was I know that if you take
cultured cells in a dish and you
just…you’re doing some, you know,
specific nutrient withdraw, you withdraw
amino acids when you withdraw glucose,
or you withdraw glutamine, you can induce
autophagy. But in the whole organism,
for example, in mice and in humans,
ultimately, you know, can you just limit
your protein intake for a week and induce
autophagy even in the presence of a
normally caloric diet where you’re
still getting enough energy.
That’s a good question.
We have never tested to selective
completion of one or the other nutrient,
I suppose that this would work,
because protein depletion may affect
a newer endocrine factors like
insulin growth factor that, at the
end, will, due to its depletion,
decrease the transport of glucose into the
cells. And hence, stimulate autophagy.
But it has not been tested thoroughly in
mice. What we usually do is we starve mice
and sometimes human volunteers completely
from any kind of caloric uptake,
and in this case, we do see, at the whole
body level, that in all major cell types,
perhaps with the exception of the brain
that is somehow buffered against this
effect, protein deacetylation
occurs mostly in the cytoplasm.
And that’s a biomarker
of autophagy, you would say?
Well, it’s too early to
say that it is a surrogate or proxy of
autophagy. So far, we have not been
able to dissociate the two phenomenon,
autophagy and protein deacetylation,
in the response to nutrients.
However, when you induce autophagy by
pharmacological tricks, such as cell
permeable peptides that
dissociates an inhibitory interaction
between a Golgi protein and Beclin 1. You
can induce autophagy without that protein,
deacetylation would occur before. And
similarly, when you give chemical
inhibitors of mTOR like rapamycin or
the rapalogs, there’s also no protein
deacetylation. So, you can induce
autophagy without protein deacetylation
which means that the proxy would be
imperfect. So we do have a system to
measure autophagy which is relatively easy
to be used in experimental systems which
is the study of the redistribution of LC3
and other members of the same family that
are usually diffusely distributed all
over the cell, mostly in the cytosol,
and then we’ll aggregate or redistribute
towards autophagosomes and autolysosomes.
So they acquire a punctate distribution
small dots in the cytoplasm and these dots
can be seen by fluorescence microscopy if
LC3 is labeled by immunofluorescence or
when it is fused with green fluorescent
protein or similar biosensors.
And so, in humans, the only accessible
cell type is the circulating white blood
cell, the leucocyte. So we can draw blood
and determine by immunofluorescence the
redistribution of LC3 from a diffuse
to a punctate from a pattern.
And this is then a sort of detection of
autophagy that can be applied to human
beings as well.
That seems like it’s kind
of complicated though for your standard
clinic to be able to use
immunofluorescence to look at some
leukocytes, circulating leukocytes, right?
I mean that’s more…
Well, you need some
technology. Especially, ordinary
cytofluorometry cannot be used for this
kind of approach, because the standard
cytofluorometer just measures an intensity
of fluorescence signal per cell not its
subcellular distribution. So there are
cytofluorometers that take pictures of the
cells that are flying in front of the
detector, and using these pictures and
analyzing them by image analysis
software, allows them to quantitate the
redistribution of LC3
So there is hope for a
non-invasive clinically relevant biomarker
for autophagy but there still, it seems
like there needs to be more work done
before that actually happen…before
I can go to my doctor and say,
“I did a four-day fast. I’d like to
see if I’ve activated autophagy.
Can you please draw some blood.” Right?
We’re not quite there yet.
Yeah, it would be
wonderful to have the reward of measuring
autophagy as a result of fasting
and to get an objective incentive
as a biomarker for doing that.
Right. So it kind of brings up
another question I had which related to
when you started talking about how you
can fast, and fasting in organisms like
rodents and also in some human
volunteers does induce autophagy.
And the question that I had for you
is, like, I’ve talked with Dr. Valter
Longo, he was on the podcast, and
he talked quite a bit about his research
on prolonged fasting in both rodents and
also in humans and how the prolonged
fast, at least, in rodents is 48 hours,
which in humans is around 4 days,
4 to 5 days. And that was able to very
robustly, not only activate autophagy, but
also cell death, and that was
followed by a regeneration period.
But, the question is do we know what the
minimum amount of fasting time is for
humans or rodents that can activate
autophagy? So for example,
when I’m not pregnant, I usually followed
a very time restricted eating schedule
where I like to eat all of my food within
at least 10 hours, and then I fast for 14
hours every night. Some people do even
more strict. They eat within 8 hours
and they fast for 16 hours. Does that
16-hour fast induce any autophagy in any
of our tissues? Is there
any evidence, do we know?
We don’t know. So, Craig
Thompson published a paper on circadian
variations in hepatic autophagy.
So you know that mice don’t eat during the
day and they eat during the night, and
so, the entire cycle is inversed.
And he observed that as a result
of not eating during the day,
there was more autophagy in the liver.
So this result is intriguing.
It has not been, to my knowledge,
extrapolated to other organs.
And it still certainly requires
more profound studies.
When you say during…
So what we did on
circulating leukocytes is that we needed
to wait for three or four days to
see a massive induction of autophagy.
There’s a fundamental difference
between rodents and humans, and so,
the two days that you have been alluding
to cause a 20% weight loss in mice,
that are, at this time point, at the verge
of death. Another day would potentially
kill them. And so, 20% is a lot,
so imagine this for yourself.
In two days.
In four days, a human
being only loses one to two percent of his
or her weight.
Is that because they have
a higher metabolism, rodents do, or…?
Yeah, it’s certainly
linked to the change in the surface volume
ratio that is classically associated
with an accelerated metabolism.
Yeah, okay. That sort of…
So we don’t really know to what extent
autophagy can be occurring in a shorter
intermittent fast. There’s some hope that
it does. I mean, I know, for example, you
mentioned IGF-1 and how IGF-1…lowering
IGF-1 is important for inducing autophagy
because in the whole mTOR pathway and so
on. But I do know that the half-life
of IGF-1 is, in a serum at least,
is around 12 hours. So, the question
becomes well, okay, if you start to lower
IGF-1, after 12 hours, is do you still
need more to occur, like more ATP
depletion, more…what…you know, what is
it that needs to happen to actually send a
signal to the cells to go, “Oh, I’m
stressed. I need to start eating my
whatever organelle or damaged proteins
or something.” So is that something that
people are currently investigating? Like,
the minimum amount of time that it would
sort of take to induce, for fasting
at least, to induce autophagy.
It’s an extremely
interesting question that is easy to be
answered in rodents, and difficult in
humans. Because it may be easy to find a
volunteer who fasts and allows
for regular blood drawing,
but it will be very difficult to find a
volunteer who fasts and allows for liver,
muscle, or skin biopsies.
Right, right. It kind of
reminds me there was a study I was
actually reading the other day that was
done in the Caloric Restricted Society.
You know, there’s a group of people
that are out there practicing caloric
restriction which typically is eating
around, what, 30% less food than you
normally would eat or something like that.
A lot more difficult for people to
maintain, I think, than intermittent
fasting is, but there was a study that was
published. And these individuals had been
doing caloric restriction for about six
years, plus or minus. I’m sure you’ve seen
the study. But they did muscle biopsies
on them and they measured LC3. They
measured some of the biomarkers of
autophagy, I think Beclin and some other
things. And then they measured heat shock
proteins, which are also a stress
response. And it was, you know,
like in some cases, the heat shock
protein, HSP70, was elevated by 12 fold
compared to age-matched, lean controls
that eat more of a Western-type diet.
But you know, the fact of the matter
is that they did do a muscle biopsy.
Autophagy was activated. You know,
the stress response pathway,
in general, was activated. But six years
of doing caloric restriction is not very
sustainable for the majority of
the population in the, you know,
at least in the United States
and in Western world.
Actually, in mice, you
can obtain exactly the same longevity
extension that you would obtain, the 30%
of caloric restriction by intermittent
fasting. So, it’s logistically much more
difficult. Imagine, you have to weigh, for
each mouse, the amount of food that
they would eat normally, subtract 30%,
put it in the cage, individual cages
because calorically restricted mice tend
to eat each other.
Yeah, they become
aggressive because they are hungry.
Yeah, they become
cannibals. So, it is logistically much
more simple to take out the food from the
cage completely, and to put back the food
on the next day. So it’s one day without
any food, and another day with normal
nutrition. And at the end, so, you have
an oscillation of the weight of the mice,
10% every day. These oscillations tend to
become smaller because the mice somehow
adapt to this sort of stress, but the
final result is that the intermittently
fasted mouse has the same weight as a
normally fat mouse. A difference with the
calorically restricted mouse, it weighs
also 20% to 30% less. And in spite of this
difference in the body weight,
intermittent fasting allows for a lifetime
expansion in the same way as does caloric
restriction. So, one can also consider
that this may be more amusing to have, if
I was a mouse, I would probably prefer the
intermittent regimen, because it means
satisfaction during one day and
dissatisfaction on the other day,
but not permanent dissatisfaction.
Most people prefer
doing intermittent fasting. I mean,
it’ll be very interesting to see more
studies come out on, you know,
the translation of this to humans and,
as you mentioned, you needed three days
of…was it a water fast they did? Was it
a complete fast or they had coffee or…?
Okay. No food.
no milk. And water.
Okay. So three days was enough
to, at least, show signs of autophagy in
circulating leukocytes. And
Valter’s work has shown, you know,
four to five days and he’s done. You know,
he’s got his fasting and then he’s got the
fasting-mimicking diet into… Some
also hints that that also is enough.
So, that’s sort of encouraging. It would
be more encouraging to have like a 24-hour
fast or 48-hour. I mean, that’s
so much easier to do in general.
But the other thing that induces
autophagy, you’re mentioning the stress
response and oxidative oxygen, and it
sort of reminded me of exercise and how
exercise also induces autophagy. I’ve
seen some studies where in humans,
they’ve looked at muscle, skeletal muscle,
and how aerobic exercise and eccentric and
concentric exercise all can activate
autophagy in skeletal muscle.
Do you know if it activates autophagy
in multiple tissues? Exercise?
that we have not studied. So,
it is known that endurance training is
particularly efficient in mice to induce
autophagy and that it mediates
anti-obesity and anti-diabetic effects
that are depending, in a way, on autophagy
induction, because genetic modifications
of the process that leads to
autophagy induction, it’s inhibition,
specifically by exercise, can prevent
these anti-diabetic effects.
Oh, I didn’t know that the
role of the exercise in preventing
diabetes was shown to be dependent
on autophagy to some degree.
That’s very interesting. So, do you
think that has to do with the liver,
and the like pancreas,
somewhere… I mean, is it known?
To know this in detail,
it would be necessary to inhibit
autophagy specifically in different
tissues and, to my knowledge,
this has not been done yet.
Okay. So do you think fasting
while you’re like exercising in a fasted
state… Now, that’s another thing
that’s…do you think that would be
important? Or do we… I mean, I’ve seen
some studies in mice where they claim it
is, but mice have a very high metabolism.
And so, there’s a synergy there.
But when you look in humans, it’s not so
important. Like, the exercise can still
induce autophagy in skeletal muscle in
humans even without being in a fasted
state, but the question is like
will you synergize more and…
speculate, but we don’t know.
I’m giving you a lot of ideas
here. So maybe we can kind of talk,
shift a little bit into the general role
that autophagy plays in some of these
age-related diseases, like
neurodegenerative disease, cardiovascular
disease, and cancer. Talk a little
bit about the microautophagy,
or is that what you call it, like,
when you’re talking about the specific
degradation of organelles like
mitochondria or protein aggregates?
So normally, when
we refer to autophagy, we talk about
macroautophagy, which is the phenomenon
that you can see easily by microscopy.
Because of the formation of the
autophagosomes that are big enough to be
seen by conventional microscopy, face
contrast, and especially of course when
you enhance a solution by
immunofluorescence or similar
technologies. So, there are other types
of autophagy that are less well studied.
Like, chaperone-mediated autophagy or
microautophagy where basically proteins or
portions of the cytosol are
introduced directly into lysosomes.
So you don’t need the mouth of the process
or autophagosome, you just need the
lysosome. And they are much less studied.
And then there’s a special case among
different kinds of macroautophagy. So
to be very simple in the dichotomy,
there is the case that autophagy is
dictated by general stress or general
absence of neutrons, which means
that it is dictated by demand.
So the cell needs to eat some portions
of itself to adapt to nutrient stress,
and the other kind of autophagy is
dictated by the offers. So a damaged
organelle will change the composition of
its surface in a way that it is decorated
by signals for stimulating its
engulfment by the autophagosome.
And so, it’s another kind of
autophagy that then can be specific.
Specific for organelles of different types
like mitochondria, and it is called
mitophagy, or for peroxisomes, and it is
called pexophagy, for the endoplasmic
reticulum, and it’s called reticulophagy,
specific for ribosomes,
ribophagy. Perfect, yes. And specific for
viruses, and it is called virophagy.
And the two processes may also interact
in a way. So when you stimulate general
autophagy by activating the nutrient
sensors, AMP kinase, inhibition of mTOR,
or by provoking deacetylation, then you
increase the demand, and the autophagy
machinery actually prefers in a way to
sequester and to destroy those organelles
that are already slightly marked for
destruction. The protein aggregates that
are not yet harmful enough to emit
a signal per se but they are there.
And so it’s a sort of preferential
cleaning of the slightly damaged and
slightly aging portions of the cell. And
this may actually explain why stimulation
of autophagy in cells, when they are
monocellular organisms or at the
organismal level at different organs,
can be a sort of device against aging.
Wow, that was very beautiful
explanation. It actually answered a
question I was going to ask you which
was, you know, the difference between the
signal, for example, nutrient generalized
autophagy, when you have the nutrient
sensing stress that even that can, to some
degree, selectively degrade mitochondria,
for example, but the actual signal that
really does activate mitophagy…when
you’re talking about mitophagy,
it’s a little different, right?
It’s the actual mitochondrial
damage, the membrane potential…
Yes.so when a
mitochondrion is suboptimal in its
function, it will decrease its
mitochondrial transmembrane potential.
And this is a signal to activate enzymes
on the surface of the mitochondria that
cause ubiquity relation, recruitment of
autophagy adapters, and leads at the end
to autophagy because of the
organelles or the organelle in a way
offers itself, it proclaims its
sacrifice by autophagy. And so,
of course, this is not an all-or-nothing
phenomenon. So mitochondria can be aging
in the cell, and as they age, they
gradually decrease the performance and the
mitochondrial transmembrane potential.
So, those mitochondria that are most
dysfunctional, they will be eaten first
if you increase the demand for autophagy.
That is very cool. And if you
are selectively degrading these damaged
mitochondria, which are you know, or aged
which are damaged, do they get replaced by
new mitochondria? Is that a signal
for mitochondrial biogenesis?
Yes. So in C.elegans,
this was a study that actually the whole
turnover of mitochondria is regulated.
So, there’s a sort of coupling between
That’s good to know.
So it’s very clever,
how the system has been designed.
It’s great. So it’s not like
you’re losing…you’re not losing the pool
of mitochondria. You’re effectively
losing the defective pool,
and you’re almost making younger
mitochondria. If you’re going to make a
new mitochondria, then it’s going to
be young and fresh and not damaged.
So it’s very elegant
way to sort of
replenish your mitochondrial
population, it seems.
So we have to make the
difference between homeostatic conditions
and, for instance, cellular
differentiation when cells change their
metabolic program. So the easiest example
is yeast that you suddenly place in the
glucose-containing medium to allow for the
fermentation of glucose in wine or beer
production. So these yeast cells don’t
eat much oxidative phosphorylation,
and they essentially rely during the
process on glycolysis. So they adapt to
this change by destroying most of
their mitochondria, by mitophagy.
And this makes actually a metabolic
adaptation of the yeast cell efficient.
Do you have similar examples in the
embryonic development of the retina for
retinal ganglion cells or the
differentiation of macrophages from
so-called M0 to M1 macrophages, in
which the cells change from oxidative
phosphorylation respiration to an
essentially glycolytic metabolism that is
coupled to mitophagy. And so, inhibition
of mitophagy actually avoids the
differentiation process in both
examples that I just gave to you.
That’s really interesting. So
obviously, these processes are not just as
a stress response, they’re
part of development as well.
They can be used
in multiple different instances.
Very interesting. And
in the case of mitophagy here,
it’s also it plays an important role
in the prevention of neurodegenerative
Yes. So, most known
neurodegenerative diseases are either
caused by the aggregation of poorly built
protein cell that somehow create protein
aggregates that are toxic for the cell.
Or they can also be caused by septal
deficiencies in the autophagic and
lysosomal machineries that lead to the
accumulation of unfolded proteins at
the end. And so, either the excessive
production of unfolded proteins or
their reduced removal causes to a slow
accumulation of these toxic
protein aggregates. Remember that
neurodegenerative diseases are slow
processes in most cases that manifests
with old age. And so, one strategy
to treat neurodegeneration,
at least theoretically, is to
increase autophagic turnover. And so,
one technique is actually then to
stimulate general autophagy by increasing
the demand, by starvation, or by
biochemical trickster that substitute for
starvation, and to reduce
the protein aggregates
that are the cause of the disease.
So these protein aggregates
like amyloid beta plaques in Alzheimer’s
disease or alpha-synuclein in
Parkinson’s disease. So, basically,
the clearing out of those protein
aggregates obviously would play an
important role not only in prevention
but presumably also, to some degree,
in help with the treatment. Of course,
that’s you know, an speculation,
but… And then the mitochondria, the one
I was thinking about with mitophagy,
was the role, at least some of the
proteins that are involved in that,
like, the PINK/Parkin, and how they seem
to be important for Parkinson’s disease.
Is that accurate?
Yeah. So the
PINK/Parkin pathway is one pathway
among others, that allows for marking
mitochondria that are damaged for
destruction. And so, inhibition of this
pathway leads to the accumulation of
malfunctioning mitochondria with major
consequences for the cell that harbours
cells’ mitochondria because all of a
sudden bioenergetic metabolism becomes
inefficient, reactive oxygen species
are produced, and as you know,
mitochondria are latent bombs in the sense
that they enclose potentially dangerous
proteins that once released will
activate the apoptotic machinery
and cause cellular suicide.
Yeah. So, that’s the question.
Do we know the threshold for the stress
threshold for, you know, activating
autophagy, and when that pushes the
mitochondria then to permeabilize
and cause cell death?
Like, where, for example, with Valter’s
work in mice, he had done 48-hour fasts
and there was both autophagy and
massive apoptosis occurring. So,
is it just the intensity of the
signal that can then say, “Okay,
autophagy is not going to work here.
We got to die.” Or do we know?
Well, autophagy in used
in most cell types, while apoptosis is
occurring in selected cell types.
So what Volter has been observing,
if I remember well, is destruction of
leukocytes, right, white blood cells,
which are very easily to be rebuilt. And
so, the loss of 50% or 75% of leukocytes
can be easily repaired in a few days. And
it is a way to adapt the repertoire of
immune cells to changing circumstances.
It is a way also to inhibit unwarranted
inflammatory reactions. So depending on
the context, induction of autophagy can be
actually a subtle way to avoid excessive
inflammation. One example is the so-called
sickness response. So, a cat or a dog or
a human being or a mouse that is sick,
that has a bacterial infection, will
hide away, avoid light and noise,
and will not eat. It’s a classical
phylogenetically conserved reaction
in most cases of bacterial infection. And
so this phenomenon leads to changes in the
metabolism. Ketone in the production of
ketone bodies, the reduction of glucose
levels, presumably also induction of
autophagy, and altogether these mechanisms
avoid excessive inflammation that may be
lethal. So Aslan Medzhitov published a
paper in cell last year showing that
force-feeding mice or just increasing the
glucose levels to a normal concentration
was sufficient to make bacterial infection
that otherwise would have been
able to cope with lethal.
Wow. So, I know in humans too.
And we have a bacterial infection,
for example, a stomach virus or
something that’s bacterial of origin,
you don’t eat as well. So, it sounds like
it’s sort of a protective mechanism.
That’s really interesting. I
didn’t know that. It’s very interesting.
I want to kind of move on to cancer, just
for time purposes. So cancer is another
sort of very, it’s been, in regards to
autophagy, something that I’ve always sort
of been unsure about, because it’s
very clear to me that preventing the
accumulation of damage, you know, pieces
of nucleic acid and pieces of chromatin
and all sorts of things that can cause
inflammation by having damaged proteins
around and things like that. Obviously,
clearing those out would be very important
for preventing cancer. But when it comes
to treating cancer, it’s not as clear.
There seems to be… I mean, for example,
you know, there’s a very classic drug out
there, chloroquine, right, that inhibits
autophagy that’s used to kill cancer
Well, it’s not exactly
[inaudible] or chloroquine is a
lysosomal inhibitor. It’s a molecule that,
due to its charge, will specifically
enrich in the membranes of lysosomes, and
then causes lysosomal membrane damage,
potentially also inhibition of autophagy.
But it fundamentally also liberates the
potentially toxic content of lysosomes
into the cytosolic space.and so,
there are a few reports around showing
that inhibition of autophagy is not the
sole mechanism by which chloroquine
can mediate such a toxic effects.
that’s good to know.
The other thing that is
important to notice is that chloroquine
and hydroxychloroquine, which are
antimalarial agents that have been used
for a long period, and also actually used
for the treatment of rheumatoid arthritis
because they have anti-inflammatory
properties. Only introduced into clinical
trials, most in combination with
chemotherapy or radiotherapy to treat
cancer. And those clinical trials,
so far, are not convincing.
Okay. So what about
the fact that some cancer cells
do activate autophagy?
So one relatively general
mechanism may be that early during
oncogenesis, the deletion of tumor
suppressor genes or the activation of
oncogenes leads to autophagy suppression.
So there are several examples for this.
And it is part of the process that
leads to cellular transformation,
because autophagy is a homeostatic
mechanism that, if inhibited,
favors genomic instability and
malignant transformation of the cells.
So there are examples on the literature
also that dived inhibition of autophagy is
sufficient to cause oncogenesis, in
particular, in the context of leukemia.
And so, later on, when the cells strive
and adapt to an ever more hostile
microenvironment, hostile because there’s
too little vascularization for the
expanding cancer cells, so initially there
are hypoxic areas, there’s no normal
tissue architecture, so the cells are
usually undernourished. The doctor may
apply some chemotherapeutic agent
which is an additional stress.
So there are internal and external stress
pathways that the cell has to cope with.
And it is an advantage for the cancer
cells to reactivate the autophagic
process. And so, it has been proposed that
inhibition of autophagy would be a way to
make the cancer cells more fragile and
vulnerable to therapeutic intervention by
chemotherapy, radiotherapy, targeted
therapies. The problem is that nothing is
simple in oncology and that cancer is
not just a cell-autonomous disease.
It is more. It is not just that one
cell has become wild term and has been
accumulating genetic and epigenetic
changes that make it selfish.
No, a cancer cell will only survive if
it escapes from amino surveillance.
So the immune system, fortunately for us,
is usually very efficient in eliminating
aberrant sells, premalignant cells,
and the initial cancer cells.
And actually, the inhibition of autophagy
that occurs during early oncogenesis maybe
also a way for the cancer cells
to hide from the immune system.
And so, it is complex. It’s immunology,
multiple different players come into
action. Autophagy, for instance, is
required for stressed cells to release ATP
into the microenvironment. You know,
of course, ATP is the most important,
energy-rich metabolite in the cell. It’s
like the equivalent of the dollar for
bioenergy clinics in the economy. And ATP,
when it appears all of a sudden outside of
the cell, is considered as
non-physiological. It is a dangerous
signal. It is perceived by so-called
purinergic receptors that are present,
among other cell types, on leukocytes,
and particularly myeloid cells.
And a cell that undergoes autophagy may,
especially, when this occurs before cell
death, release ATP to attract myeloid
cells into its proximity and to start an
immune response against tumor antigens
in the context of the initial oncogenic
events. And so, autophagy is required for
some steps of the immunosurveillance
process. And it is exactly this process
that makes cancer therapies efficient.
So, in contrast to the official dogma
that has been en vogue for several
decades, chemotherapy is not just killing
the cancer cells as if we used an
antibiotic that specifically paralyzes
the metabolism of bacteria.
No. It is true that chemotherapy induces
cancer cell death, but the important point
is that chemotherapy must provoke this
cell death in a way that it later leads to
an immune response. And so, if you have
a long-term effect of chemotherapy,
for years or decades, that continues
beyond removal of the drug,
it is due to an anti-cancer immune
response. And so, since this is so
important, the capacity of the
chemotherapeutic agent to induce autophagy
is actually required for the
long-term efficacy of the treatment.
I did not know that. I had no
idea that the induction of autophagy would
stimulate the immune system through
this extracellular ATP mechanism.
And how that is, I mean, obviously the
immune system is extremely important for
killing cancer cells. But that’s very
cool, and probably leads to the next topic
on some of your work with the fasting,
so-called fasting mimetics like
spermidine, hydroxycitrate that you’ve
done. Maybe can you kind of just briefly
explain… I’ll start with spermidine.
What is spermidine? What does it do?
Well, it was first
start to explain what are these
fasting mimetics is as you say and
caloric restriction mimetics as we say.
So the CRM’s, caloric restriction
mimetics, are actually inducing the same
biochemical changes in the cells
as would do starvation or fasting.
So, we have been discussing on the
importance of acetyl-CoA and protein
deacetylation resulting from the
inhibition of Acetyl-CoA in the context of
fasting. And caloric restriction mimetics,
similar induced deacetylation reactions to
stimulate autophagy. And this can be
actually achieved in three different ways.
First, you simply inhibit the generation
of Acetyl-CoA. The enzyme that generates
Acetyl-CoA in our cells, the most
important one for the cytosolic pool,
is ATP citrate lyase and hydroxycitrate
or pharmacological compounds that inhibit
this enzyme cause Acetyl-CoA depletion,
deacetylation, and autophagy.
And you can have the same effect by
inhibiting the protein acetyltransferases,
some of them have been identified. Like,
EP300 which appears extremely important
for autophagy regulation, and specific
inhibitors of EP300 such as spermidine and
natural compound or C646 which is a
pharmacological compound specifically
designed for this function. They can
also cause deacetylation and autophagy.
And finally, it is possible to activate
deacetylases or enzymes that remove acetyl
groups from proteins and cause
hypoacetylation and autophagy. And one
example that is well known is resveratrol
contained in red wine that induces
autophagy through this pathway. So,
all these agents, caloric restriction
mimetics, have different molecular
targets, but it activate autophagy by a
final common pathway.
acetylation seems like that.
Okay. So, with some of the
major ones that you’ve worked with,
spermidine. I’ve read quite a bit about
spermidine. I know it’s found in high
concentrations in natto, the Japanese
fermented soybean that doesn’t taste like
great. But I’ve seen studies about aging,
you know, giving it to even aging mice or
something, can then extend their lifespan.
Is that true?
So spermidine, to come to
the source of spermidine is contained in
the nuclei of all kind of cells. So, in
the nucleoid of bacteria but also the
nuclei from yeast cells or from
plant cells or animal cells. So,
all food items that contain nuclei cells,
are actually containing spermidine.
Also, there are large variations in
the content. So we have to know,
on this spermidine, it also is volatile
and accounts for the smell of sperm.
So, it is frequently found in food items
that have some kind of smell like natto or
durian fruit or fermented cheese, when it
is generated from non pasteurized sources
and very rich in bacteria and fungi that
are contributing to the fermentation
process, which is, of course, smelly
cheese. And it’s also quite abundant in
some vegetables and food where the scent
is more agreeable to most people because
it is complex to other molecules that
reduce its volatility. So spermidine has
the capacity to induce autophagy
when it is taken up with fruit or with
the drinking water when we take mice.
It can also be injected,
of course. It is produced by our
microbiota. So, one-third of the
spermidine in our body is probably
produced in the intestine, and you can
manipulate a microbiome to increase
its production of polyamines,
Probiotics? Or through…
So you know what
strains of bacteria… ?
Yeah, there is a Japanese
group that has been publishing that
specific bacteria overproducing polyamines
can be used to reduce the development of
colon cancer or to reduce aging.
Wow, fascinating. That’s very
interesting. And you’ve shown with the
spermidine, I know we have a limited time
here, with the spermidine that’s been
shown to… Was it spermidine or
hydroxycitric, I think, that was shown to
synergize with, like you were mentioning
before, the chemotherapeutic…
both of them actually.
So the mechanism is that
when you combine chemotherapy with caloric
restriction mimetics, all the caloric
restriction mimetics that I mentioned,
including spermidine and hydroxycitrate
and resveratrol, will enhance the
anti-cancer immune response that
makes the therapy durable. So,
we have been able to show that inhibition
of autophagy in the malignant cells or
destruction of the extracellular ATP that
is released as a result of autophagy is
sufficient to abolish the favorable
interaction between caloric restriction
mimetics and chemotherapy. And similarly,
actually, it is sufficient to remove
T-cells from the system. And you will lose
any kind of tumor growth reduction induced
by chemotherapy combined with caloric
restriction mimetics as it proves that the
cellular immune response is actually
decisive for therapeutic outcome.
Wow, that’s really quite
promising, I think, for you know at least
in the clinic, if you can somehow test
whether or not this caloric restriction or
fasting mimetics work in conjunction
with some of these immunotherapies,
that would be fantastic. But I want to ask
you one last thing too about some of these
fasting mimetics. Like, if I were to
just supplement with hydroxycitric,
for example, or if there were spermidine
supplement, and I was still eating a
normal, you know, healthy diet, but not
caloric restricted and not fasting.
Do you think that would be
sufficient to induce autophagy?
Well, I can
respond for mice cell…
How about in mice?
…that this is certainly
the case. I don’t know about humans
because we have no clinical
studies in this field.
So in mice, it does?
It does. And in mice,
you actually can give a combination of
high-fat diets, that usually would cause
obesity, with spermidine to reduce weight
gain through mechanisms that we don’t
understand and that we believe to be
autophagy-dependent but have to
elucidate in some molecular detail.
That’s very cool. So I’m going
to look up those strains of bacteria,
you know, to see that can I eat fermented
foods to increase that population and get
more spermidine. You know, things like
that would be very interesting to me as
useful little tools. Do you have
any practices that you do? Do you,
for example, do intermittent
fasting or any type of fasting
or time-restricted eating?
So, usually in my
ordinary life, when I’m in Paris and
working, I only have one meal per day.
So I have dinner with cheese and wine
containing spermidine resveratrol. It’s an
ideal combination because inhibiting the
acetyltransferase and activating the
deacetylase will have, of course,
a synergistic effect.
Oh, you do?
you can show this in mice
and this is an excuse for me to…
Yeah, wine and
cheese go great together.
…profit from a banquet
combination. And I do also practice some
fasting twice per year also when I
don’t eat anything for five days.
Oh, wow. So you do
a prolonged fast twice a year.
Well, it’s not so
prolonged, so you know that Gandhi,
for instance, has been doing fasting
exercises for 20 days or more,
which is the time at which a normal
individual could have some long-term
consequences on the health. So, 20 days
is some period that usually is could be
easily supported by a healthy
that has no underlying pathologies.
I have a couple of friends
that have done. One of them is used to
very prolonged fast, like 20 days. He’s
morbidly obese, and he’s lost now like 200
pounds over the course of a year by doing
just several rounds of these prolonged
fast. But he’s got a lot of fat, you know,
to supply energy. I’m not sure I would
still subject myself to a 20-day water
fast, but it’s cool to know that you do
these five-day water fast, twice a
year, and eating one meal a day.
and doing exercise.
And doing exercise.
And when you’re eating… Sorry,
throughout the day, when you’re
not eating, do you drink coffee?
And there’s polyphenols
in coffee that actually…
Yes, we published a study
in mice giving them a non-toxic dose of
caffeinated or decaffeinated coffee
with a drinking water continuously.
And we could show that this was
magnificently inducing autophagy.
Right, totally independent
caffeine, so just the polyphenols.
And this was like, I think, I read your
study, it was like after four hours or
something in the mice. So, in
humans, potentially, maybe the fasting
plus the coffee…
Well, coffee abuse has
been linked to major health-promoting
effects. So, most cardiovascular and
neurodegenerative diseases are actually
reduced in heavy coffee drinkers as an
independent link between lifestyle and
pathology. So, of course, it’s an
association that obviously can be
criticized because it’s just a study in
which you find a statistical correlation.
It would be much more interesting
to do a randomized clinical trial
on coffee intake.
and more expensive, yeah.
Great. So fasting,
coffee, wine and cheese, one meal,
exercise, you’re doing it all.
Well, really enjoyed this conversation,
Guido. If people want to find you,
you have a lab website, kroemerlab.com.
That’s K-R-O-E-M-E-R lab, L-A-B .com.
So, thank you again for this wonderful,
very illuminating conversation.
It’s a pleasure.
I hope you enjoyed
learning about all things autophagy,
both macro and micro from Dr. Guido
Kroemer. Dr. Kroemer in my mind is the
consummate expert in this field, and also
is amazingly prolific with somewhere in
the neighborhood of a thousand
publications to his name. Quite the feat
to say the least, and all of the more
reason why it was an enormous privilege to
get to talk to him.
Do you love these geeky, deep biology
podcasts? You can be a part of the
community that nurtures the existence of
this podcast and other useful creative
offshoots that can be found at
FoundMyFitness by heading over to
to buy a solar-powered wood chipper or a
gargantuan vat of gee? If you
get them from Amazon, make sure to click
the FoundMyFitness affiliate link first.
Any time you click our Amazon affiliate
link, it actually makes sure the podcast
will get a small cut for any orders you
make in the next 24 hours. It literally
doesn’t affect your transaction at all.
And the best part is you probably
already shop at Amazon anyway.
No big deal, right? You can find the
Amazon affiliate link, as well as other
great ways to support the podcast, such
as the Pay What You Can subscription,
either direct or through Patreon,
by heading over to
Crowd Sponsor. Finally, are you a genetics
geek? If you’ve done your 23andMe,
and already have your raw genetic data
available to you, you can run it through
the genome analysis tool on my website
That’s foundmyfitness.com/genetics. What’s
that? You may ask. This tool is a way for
me to tell you about interesting
single nucleotide polymorphisms,
that’s SNIPs for short, and to quickly
check your raw data for them.
Maybe you’ve heard me talk about a
polymorphism that affects how you
metabolize saturated fat or vitamin D or
any number of other polymorphisms that
I’ve talked about that usually share the
common theme of being diet or lifestyle
modifiable. Since I’m still learning,
my report is still learning,
and I’ll be adding more and more
interesting genetic polymorphisms to the
report as time goes on. Definitely worth
checking out if you have your 23andMe data
or, in the near future, ancestry.com
or any of the other genetic providers.
We’ll be adding those soon.
Learn more by heading over to
Thanks so much for listening
and I’ll catch you next time.