Microbiome, Brain and Behavior
Ted Dinan explores the importance of the microbiome and its affects on brain and behavior. What we eat potentially affects our brain and behavior and those microbiomes have many potential links in the communication with the brain. One study demonstrated–the first of its kind– a probiotic cocktail altered activity in regions of the brain that are associated with emotional processing.
Microbiome: “There are numerous potential immune mechanisms by which the microbiota could communicate with the brain.”
be here, and what I want to do in the time that’s available to me is to look at some
of the evidence out there to support the view that the microbiome does influence brain development
Now, assuming that the microbiota is capable of influencing the brain, I suppose one must
ask how or what pathways might enable the gut microbe axes to actually communicate with
the brain, and there are a variety of potential pathways. I mean, the vagus nerve is obviously
one potential pathway; the spinal cord. There are numerous potential immune mechanisms by
which the microbiota could communicate with the brain. And there are endocrine mechanisms
of communication as well. There is an accumulating volume of pre-clinical evidence looking at
some of these mechanisms of communication. The reality is that in humans we know very
little about the potential most important roots of communication between microbes and
If we are to study the influence of the microbiota on the brain, what are the techniques used?
Now this is a summary slide that John Cryan and I put together, and it looks at the most
common methods that are used in the literature at the moment. Germ-free studies are commonly
used. Infection models, where you put an infective agent into the gut and look at the impact
on brain function; that’s obviously one way of looking at the — or exploring the microbiota
influence on the brain.
We’ve heard about probiotic studies, and indeed, you know, prebiotic studies as well. There
are a number of antibiotic studies in the literature looking at how antibiotics acting
on the gut might influence brain behavior — brain and behavior. And, of course, recently
fecal transplantation has become big in relation to C. diff, but, of course, it’s an approach
that could be used to look at the impact of the microbiota on the brain.
If we start off with infection studies, back in the 1990s and Mark Lyte did a very interesting
study where he did — he essentially gave mice a subclinical dose of Campylobacter jejuni
that didn’t actually cause immune activation, but he showed that acting through the vagus
nerve and through the nucleus tractus solitarius that it actually caused anxiety in mice. And
since then there have been a number of studies that have suggested that infective agents
can cause marked alterations in mood or anxiety.
The Walkerton study is a study that many of you may be familiar with. It’s a naturalistic
study in so far as the town of Walkerton in Canada had its water supply contaminated with
- coli and with Campylobacter. And most of the population of the town suffered a gastrointestinal
infection, and GI symptoms as a result of infection. Now it was quite interesting to
follow the actual patterns of illness that emerged in the years following the actual
acute infection. Most of the focus has been on post-infective irritable bowel syndrome,
which occurred in a large proportion of the population. But as a psychiatrist, what interests
me is the fact that at the end of the first year, a very significant proportion of patients
had developed major depressive illness. So it suggests that the infective agent, through
whatever mechanism, resulted in the emergence of depression at the end of year one, and
was sustained in some of these patients for considerable periods of time.
Now, minocycline is an antibiotic, a tetracyclic, that is fairly widely used for all sorts of
things, including treating acne, and it acts on gram-positive and gram-negative bacteria.
Now, clinically, I treat many patients with depression, and some of these patients will
have treatment-refractory forms of depression. And there is now emerging, though all of the
studies are rather small sampled, but there is emerging evidence that minocycline, when
given in patients who are anti-depressive resistant, that you see a response to treatment
with the addition of minocycline. There is also a more recent study suggesting that minocycline,
when given to patients with schizophrenia, has an impact on the negative symptoms of
schizophrenia. So a suggestion that modulating the microbiota with an antibiotic might have
positive mental health benefits.
Fecal transplantation really, in psychiatric populations, has not really taken place. Steve
Collins and his group have had some very interesting — have done some very interesting studies
where they’ve taken animals who have either an anxious phenotype or a depressive phenotype,
and they transplanted the microbiota into germ-free animals. And what they’ve shown
is that transplanting the microbiota from an animal with an anxious phenotype into a
germ-free animal gives that animal an anxious phenotype; and likewise, the transplant of
a depressive phenotype, or the microbiota of an animal with a depressive phenotype into
a germ-free animal renders that animal having a depressive phenotype. There has been some
speculative evidence put together to indicate maybe transplantation of the microbiome would
be beneficial in things like Parkinson’s disease, but it is rather speculative, to say the least.
So what about germ-free studies? Well, when we look at germ-free animals, there are a
variety of alterations in such animals, certainly aspects of cognitive function are altered,
memories — aspects of memories are altered in such animals. There are alterations in
serotonin centrally, and we’ve been looking at altered social behavior patterns in such
animals as well. This is some of our data here, and it shows that if we look at germ-free
animals, if we look at brain-derived neurotrophic factor, which is exceedingly important in
areas of the brain involved in cognitive processing, for example, hippocampus, you see that germ-free
animals have lowers levels of brain-derived neurotrophic factor. You also see that in
response to stress they have altered cortical stearin output, they have an exaggerated cortical
stearin release, which is interesting because behavior, actually, these are not anxious
animals, but their endocrine response is one of an exaggerated cortical stearin release.
When we look at the hippocampus also we see altered levels of 5HT in the germ-free animals;
they have higher levels of 5HT in the hippocampus, and they’ve increased 5HT turnover in the
Now recently, we’ve been looking at social behavior, and some of you may be familiar
with three-chamber sociability test. In this particular test what we do is we place the
germ-free animal in a three chamber, where there is one — there is another animal in
one of the chambers, and we look at the amount of time the animal spends — the germ-free
animal spends with the other animal, and the amount of time it spends on its own. And what
one sees essentially is that in germ-free animals, here you have the germ-free relative
to the controls, you see that the germ-free animals will spend more time in the empty
chamber than they will with the other animal. However, when we colonize them, when we give
them a normal microbiota, you see that they will spend more time with the other mouse.
So the behavior can be altered, provided the colonization takes place early on. I mean,
if it takes place later on in development, you will not see normalization of behavior.
Now, we also look at another factor, and it’s where you have a germ-free animal, and you
have two animals in chambers with them. Now one of the animals is a novel animal that
the germ-free animal hasn’t come across before, and the other is a familiar animal. Now most
animals, most mice will spend more time with the less — or the non-familiar animal than
they will with the familiar animal. But it’s interesting, when you look at the germ-free
animal, you see, in fact, here is the conventionally colonized, and they’ll spend more time with
the novel animal than they will with the one that’s familiar. When you look at the germ-free,
there no real discrimination. If you colonize the germ-free, they will spend more time with
the novel animal.
So this is just two aspects. Basically, if you have — these animals will spend more
time with objects than they will with other animals, so their sociability is significantly
altered; they have autistic patterns of behavior.
Now Mark Lyte has put forward a concept that probiotics, in many ways, are delivery vehicles
for neurochemicals, and I suppose the question is, you know, what sort of neurochemicals
do these putative probiotics actually produce, or what are they capable of producing. And
I was actually quite surprised, I suppose, as a psychiatrist, you know, and one who works
in neurobiology, I’ve worked a lot over the years with norepinephrine, and serotonin,
and so forth, and I was quite surprised that various microbes are capable of actually producing
norepinephrine, or serotonin, or dopamine, very key monoamine neurotransmitters in the
And — but not only are they capable of producing neurotransmitters, but they’re capable of
modulating neurotransmitters as well. There’s studies with lactobacillus acidophilus showing
the capability to alter cannabinoid receptors. We’ve worked with Bif. infantis, we’ve done
a lot of work with Bif. infantis, and we’ve been able to show that Bif. infantis increases
tryptophan levels in the plasma, which, from a psychiatric prospective, may be important
because tryptophan is obviously the precursor of serotonin. And worked with lactobacillus
rhamnosus, which suggests that it alters central GABA receptors, and it’s that I want to focus
Now this is a study that John Cryan and I did in collaboration with John Bienenstock
and Mack Master [spelled phonetically], and John had a strain of lactobacillus rhamnosus
that he felt was anxiolytic. And we decided that we would look at it in a variety of behavior
models, and we would look at its impact on GABA receptors centrally. Now we chose GABA
receptors because they’re very ubiquitous in the mammalian brain, and they are potently,
they have a very significant or potent impact on levels of arousal and anxiety. Now GABA
receptors bind things like benzodiazepines, and they bind a variety of anesthetic agents.
GABAB receptors, again, are — well, they’re actually G-protein coupled receptors unlike
the GABA, which are ionotropic receptors. So when we fed lactobacillus rhamnosus as
opposed to a broth, an inert broth, we essentially found that when the animals were stressed
who were given lactobacillus rhamnosus, that their cortical stearin release was decreased.
We also found that in behavior tests, like the open field and the elevated plus maze,
that lactobacillus rhamnosus increased exploratory behavior, so it seemed to be having an anti-anxiety
effect in much the same way as the benzodiazepine would have. When we looked at GABAA receptors
expression in various brain regions, if you look here at the infralimbic cortex, you find
that lactobacillus rhamnosus decreases expression of GABAA; in the prelimbic, you get a decrease
as well. If you look at the amygdala — or, sorry — yeah, the amygdala, which is an important
area in terms of anxiety processing, you get very major changes in the expression of GABAA
with lactobacillus rhamnosus. If you look at the hippocampus here, you can see significant
increases in GABAA expression there. What about GABAB expression? Again, very significant
alterations in GABAB expression in a variety of brain regions. So lactobacillus rhamnosus
in a placebo-controlled study in rodents alters GABAA and GABAB receptor expression.
So we wanted to explore how exactly it did this. So we decided that we would do another
placebo control trial, this time in vagotimized animals and in sham-operated animals, to see
if, in fact, the effects we were witnessing with lactobacillus rhamnosus occurred in the
presence of a vagotomy. And what we essentially found was when the animals were vagotimized,
their exploratory behavior was not altered by lactobacillus rhamnosus. The increased
exploratory behavior that lactobacillus rhamnosus brings about was not seen in the presence
of a vagotomy, and the changes in GABAA and GABAB expression, which you can see here,
these are hippocampus slices, and it’s a in situ hybridization study, the changes in GABAA
and GABAB expression that we had previously seen with lactobacillus rhamnosus treatment
did not occur in the present of a vagotomy. So the effect of lactobacillus rhamnosus were
being mediated through the vagus nerve, and the animals needed an intact vagus nerve before
one saw anxiolytic effects from the lactobacillus rhamnosus, and alterations in GABAA and GABAB
Now, this is a study that Emeran Mayer has recently published in his group at UCLA, and
I think it is an extremely important study. And it’s important because it’s the first
real attempt to look at, in a systematic way, the potential impact of probiotics in humans.
And what Emeran and his colleagues did was they took a group of female subjects, and
they treated them either with a fermented milk product for four weeks, which contained
a number of probiotic organisms, so it is a cocktail of probiotic organisms, or a non-fermented
milk product, or no treatment. So you have a group of 15, 12, and 14, and they’re treated
or not treated for four weeks, and they have an FMRI scan pre and post treatment. Now,
the — basically what the study demonstrated was that treatment with the probiotic cocktail
altered activity in those areas of the brain that we associated with emotional processing.
So the probiotic cocktail was capable of altering activity in those brain regions associated
with emotional processing. Now it’s the first study of its kind; you know, there are many
studies similarly conducted in animals, but this is the first study in humans.
We’ve just published a paper in Biological Psychiatry where we defined a psychobiotic
as a class of live organism that, when ingested in adequate amounts, produces a health benefit
in patients suffering from psychiatric illness. And I think there is an accumulating volume
of evidence that, in certain conditions, probiotics can have a positive health benefit. So this
is a psychobiotic, to be distinguished from psychedelic substances, or snorting cocaine
or whatever, but we do believe that probiotics, they obviously need enormous more research,
but there is emerging evidence, I think, to indicate potential from a mental health prospective.
And when one looks at the areas where the best evidence is available, undoubtedly irritable
bowel syndrome, which I suppose gastroenterologists will say is an entirely gastrointestinal disorder,
whereas liaison psychiatrists will say it is a psychiatric condition; I suppose that’s
the product of subspecialization in medicine. But whatever the approach one takes to treat
irritable bowel syndrome, there is, I think, an accumulating volume of placebo-controlled
evidence that probiotics are effective in treating irritable bowel syndrome. If one
looks at the data, for example, with bifidobacterium infantis, there are at least two good placebo-controlled
trials in the literature indicating efficacy in irritable bowel syndrome, and there are
other studies with other organisms as well.
There is less good data for something like depression. The Messaoudi paper is a paper
about 18 months ago published which suggests that in a healthy population, a physically
healthy population, that a combination of L. helveticus and B. longum can reduce depression
scores and decrease 24R urinary free cortisol output. Now I think the endocrine aspect of
this I find actually a bit more convincing than the measurements of symptoms, but it’s
an interestingly preliminary study. There’s also a study in chronic fatigue syndrome,
which is such a difficult syndrome to treat, if anyone has been involved in treating it,
and there’s a study looking at L. casei, and finding quite mark reductions in anxiety scores
in patients with chronic fatigue syndrome. So there are some of the kind of clinical
areas where psychobiotics, or probiotics, might be of benefit from a mental health prospective.
So where are the major gaps? Well, one would have to say that when one looks at the impact
of the microbiota on the brain and behavior, that there’s a never-increasing and very good
quality literature from a pre-clinical perspective. But we are seriously lacking good clinical
studies, and they’re not clinical studies that will be necessarily difficult to do.
And if one takes the pre-clinical literature we’ve published, and one or two other groups
have also shown, that if you take an animal model of depression, for instance, that the
microbiota is altered in such models. Steve Collins’ group have shown similar changes
in an animal model recently, so the microbiota is altered in animal models of depression.
But there has been no attempt so far to profile the microbiota in patients with major depressive
illness. We’re about to start such a study, but there’s nothing in the published literature.
Likewise, when it comes to probiotics, or psychobiotics, as I like to call them, there
is lots of data pre-clinically in animal models of depression and anxiety that certain strains
of bacteria or certain psychobiotics are effective. But we have no clinical studies. Now, I think
one question that is exceedingly important, and I haven’t really focused on it because
for time restraints, but it’s the issue of, are all probiotics the same? And do they all
have an impact on behavior in the brain? And the reality is, no, they are not. I’ve showed
you one or two studies where we’ve done where the outcome was positive, but I would say
that 99 percent of the bacteria that we have tried to profile from a behavior prospective
have absolutely no impact whatsoever on behavior. They probably don’t even get through the acid
in the stomach. They’re totally ineffective; they have no impact in animal models. But
there are some that clearly do have a very dramatic impact in animal models, and I think
we really do need proper, well-controlled clinical trials in patient populations in
patients with depression or anxiety, and those studies are sadly lacking in literature at
the present time.
So I’d like to conclude by acknowledging my colleagues, John Cryan, Jarrett Clark [spelled
phonetically], and my other colleagues in the Alimentary Pharmabiotic Centre in Cork,
and thank you all for your attention.
Male Speaker: Thank you. We have time for questions.
Female Speaker: I have one. Over here. Hello.
Ted Dinan: Hi.
Female Speaker: So, very interesting, and you briefly mentioned
autism, and there have been a couple of recent studies that have shown that there is altered
microbiome in children with autism – –
Ted Dinan: Indeed.
Female Speaker: So I wonder if you would care to elaborate
on that any more?
Ted Dinan: As you say, there have been one or two studies
that have suggested an altered microbiota. There are also studies suggesting that one
or two antibiotics can significantly improve the behavior in autistic children, and there
is some suggestion of altered barrier function, gut barrier function in autism. So it may
very well be that in some autistic children — I think autism is a heterogeneous entity;
I’d be very reluctant to say it’s a homogenous entity — but I do think that there are some
autistic individuals who have altered barrier function. You get a seepage of LPS and other
molecules across the defective barrier, you get a pro-inflammatory phenotype, and you
do get central — you do you do get central changes.
So would probiotics be beneficial? Perhaps they might. I do think that there is convincing
evidence that where there’s altered barrier function, that some probiotics are capable
of improving the barrier function, so it may very well be that autistic subjects might
benefit from probiotics as they do from antibiotic therapy. But I think we need really very good
studies. Autism is an area where there is so much rubbish in the literature.
Male Speaker: My question is perhaps as much to you as,
Dr. Goodman, with the notion of psychobiotic beneficial to psychiatric health, and yet
all the indications that you’ve listed are clinical entities, like depression, and schizophrenia,
and irritable bowel syndrome, and these clinical trials are really there to treat the disease,
and that is a different rigor of what is required for such a product.
So how do you draw this boundary for yourself? Where does something that is just the beneficial
thing which can be taken care of by marketeers and commercial entities transition to become
a drug in much more hardcore intervention?
Ted Dinan: Oh, indeed. I — you raise a very interesting
point, and a very important one. It might very well be that there are probiotics or
psychobiotics out there that might be of benefit to people’s general mental wellbeing, but
might be totally ineffective if you put them into a psychiatric population. And I think
the only way we can really — I’m a psychiatric, so I’m interested in treating depression and
anxiety, and that’s why I focused on that — but we really need to, you know, employ,
you know, rigorous methodology, whichever group. I mean, even if we’re not going to
deal with psychiatric populations, I think we do need proper placebo-controlled studies
if we are going to even demonstrate positive mental health benefit in the general population.
I mean, that — those studies are not too difficult to do, and so far, most of these
organisms have grass [sp] status. It’s not — they’re safe, so one isn’t going to have
to go through the burden of the toxicology that one would with a drug and so forth. And,
you know, I think we do need to do the studies both in the general population and psychiatric
populations, and they have to be rigorously controlled studies.
Male Speaker: Perhaps not toxicology, but you still would
have to be held to a rigor of what’s actually in there.
Ted Dinan: Oh absolutely, absolutely. I agree entirely.
Female Speaker: Thank you, Ted, that was an eloquent talk.
And probiotic infusions are plausible from the purpose of this meeting within the microbiome
framework of actually looking at restocking a habitat for keystone species for microenvironment,
if you really look at it as restorative or preservation. But for all the reasons as you
said, until we really understand the safety in all of the necessary pathways. I think
what’s intriguing is you presented for anxiety and depression, and that is a particular phenotype,
but you also generalize it more to, “Let’s look at this from the standpoint of a microhabitat
for the general population.”
So I would just like to hear a little more comment from you on the direction in the field
for what we might see of probiotic in the future, and, you know, beyond just the anxiety/depression
disorders, but why you really think — and do you think it’s more likely to be just in
the live biologic sera [spelled phonetically], or will it encompass diet and foods?
Ted Dinan: Indeed. I think one of the problems in this
area with probiotics is that there are so many people making claims out there, which,
in Europe, EFSA are being exceedingly rigorous and trying to stamp out these claims, but
until such time, as companies cannot make claims, spurious claims for which they have
no data, there is no — there is little, really, to encourage a legitimate company to actually
do the necessary studies to demonstrate the benefits of a probiotic because they’d be
up against numerous companies that are marketing, very effectively, products that often have
absolutely no impact.
You know, I would stress the fact that, you know, the bulk of putative probiotics that
we had anything to do with in our lab, and I’m sure others will attest to this as well
who are interested in behavior, they do absolutely nothing. And yet there are all sorts of crazy
and grandiose claims being made by companies out there, and I think it’s only when that
is stopped that there will be, I suppose, an incentive for companies to do the necessary
studies, be it in the general healthy population, you know, from a mental health perspective,
or in psychiatric populations.
Female Speaker: Thank you.
Male Speaker: I think we need to move on as we are running
a little bit late, so let’s thank Ted.
Okay, the next speaker is Marty Blaser from New York University Medical Center, and —
Published on Aug 8, 2013