Combating Antibiotic Resistant Superbugs | Gautam Dantas || Radcliffe Institute

Combating Antibiotic Resistant Superbugs | Gautam Dantas || Radcliffe Institute


[MUSIC PLAYING] – Hello, everyone, and welcome. I want to thank those
intrepid souls who actually crossed
over to Radcliffe today despite the weather. Winter is fully upon us. Thank you very much. I want to welcome you to
our science lecture series. My name is Sean O’Donnell. I’m the associate director
of academic ventures, and I help to oversee
the lecture series. I have the opportunity
and privilege of working with Immaculata
De Vivo and Alyssa Goodman, our science faculty
directors, whose vision has really helped launch a year long
exploration of this idea of, what is the
undiscovered in science? There’s clearly so much
more for us to know. But in practice, of
course, for scientists, that often means that they are
setting off to find one thing, and what happens when they find
something completely different? This notion of serendipity
is not just romantic, but it’s also a very
practical experience. And we are really honored
today to have Professor Guatam Dantas with us
today, whose story about his own surprising
and unexpected discovery really redirected
his research and led to a whole decade worth
of a new research program. I will let Professor Dantas tell
that story himself, of course, but I do want to
briefly introduce him. And I’d like to do
so, if you don’t mind, by reading his Twitter handle. He is a bow tie enthusiast. He is a husband, a father,
a microbial ecologist, genomicist,
computational biologist. He’s also a home brewer,
a gardener, a foodie, and pretentious
cocktail mixologist. I hope he’s serving up some
wonderful cocktails for us today in his stories, and
should also point out, perhaps, that he is, at the same time,
the professor of pathology and immunology in
biomedical engineering and molecular microbiology at
Washington University School of Medicine in St. Lewis. And he is no
stranger to Harvard. He was here in
George Church’s lab. And he met with
our students today, and he clearly has a desire for
mentoring students, and leading them, and directing them in this
process of becoming scientists and inquisitive minds. And with that, I hope
that you will join me in welcoming Professor
Dantas to the stage today. Thank you. – Well, thank you, Sean, for
that very kind introduction. And thank you to the Radcliffe
Institute for this opportunity to come back to an old home. Before I get started, I want to
say something that I personally believe, and I think
reflects the opinion of most of my good colleagues, and
that is all good science is team science. And if there’s anything of
importance that I say today, it’s because of these
fine folks that I have the privilege of working with. I’m really here as
their spokesperson. Before I get to the
science aspects, I do want to mention
this role of serendipity in everything that
you’ll hear today. So I was a postdoc in George
Church’s lab from 2006 to 2009, and I came here to ostensibly
study microbial engineering for biofuels. This idea that we could
convert plant sustainable matter into replacements
for petroleum. And the approach that myself and
this grad student Martin Summer took were to try
to enrich microbes from the soil that could
survive on plant-based toxins as their sole source of carbon,
with the hope of understanding their genetic machinery. And we thought we’d
be good scientists and set up an appropriate
negative control. And luckily, our naivete
led us to this idea that, OK, the type of compounds
that no microbe should be able to consume as
a sole carbon source would be antibiotics. So we went to off
the shelf drugs that we could find in George’s
lab, set up the experiments, and lo and behold, a week
later, we got very poor results on the plant-based toxins, but
we had gangbuster growth on all of the antimicrobials. We were surprised. Most reasonable advisors
would have said, that’s great. Why don’t you focus
on what you’re doing? George is wonderfully
not that type. He got as excited
as we did and said, stop the biofuels work for
a while and figure this out. And the rest is history. Basically, everything
I tell you about today, every part of our interest
in antimicrobials, even the ones that will
seem like they were designed from scratch with this
oppression [INAUDIBLE],, really was because we decided
to follow up on that slightly weird experiment. So let’s step back
for a second and think about what we’re actually
going to discuss, and that is antimicrobials
or antibiotics. So the definition
of an antibiotic, just to get the
language straight, is any chemical that
either inhibits the growth of or kills microbes. And today, I’m
going to exclusively be talking about the group
of antimicrobials that are antibacterials,
these compounds that can be used to kill bacteria. One thing that you should
note is what they’re not is they’re not antivirals. And so this is a CDC Warning
that very clearly tells you, and you really should
heed this, that anytime you have a viral
infection, you certainly would be doing no good
if you take antibiotics for those viral infections. And I’ll show you pretty
soon that you’re probably doing something bad. So one thing to recognize as to
why these particular drugs work is because they target the most
key and conserved processes of bacterial life. They destroy the bacteria’s
ability to have a cell wall. So the guts bleed out
if they are impacted by things like penicillins. Or they gunk up the ability
for them to replicate. And the reason I bring
this up, rather than really to discuss molecular
mechanism, is the mention that Paul Erlich called
antibiotics magic bullets, but we really should be thinking
of them more as magic shotguns on magic nuclear bombs
because every bacterium has this target. So when you ingest an
antibiotic, thinking that you are going to be
targeting a specific pathogen, you’re actually impacting
every single microbe in that ecosystem. For instance, in your gut. And so we realize that even
though it might be warranted use of this really
important chemical, you’re going to have
potential collateral damage. And it’s through that
lens that, really, I’m going to be describing
work that people in my lab have done to understand what
that collateral damage might be, due to both warranted
and unwarranted use of antimicrobials. One thing that we have observed
in that particular vein, especially by looking at
retrospective studies, is anytime that
antibiotics are used in, for instance, the
human population, that those antibiotics do have
impacts on all of the good bugs as well as the bad bugs. So down here you look
at this particular plot, and you could observe that
virtually at any point in life, at any time point,
you’ll notice that there are some kind of malady or
some sort of problem that occurs on the human side,
even if the anti-microbial was warranted. Now, perhaps the
greatest response that a microbe might
have to be trying to be killed by
an anti-microbial is to respond to
become resistant to it. And this idea of
antibiotic resistance is something you’ll
see as a common feature that I’ll discuss through
the rest of this lecture. It turns out the history
of anti-microbial use in the human population
is intricately linked with antibiotic resistance. So we’re here to talk
about serendipity that the whole field
of antimicrobials is really thanks to a
serendipitous accident at the hands of Sir
Fleming, who happened to have one of the
greatest opportunities to win a Nobel Prize because
he got a little bit sloppy when he went on vacation, right? So he left out these plates
that we all have heard about, found the penicillin
mold, wiping out this potential pathogen. But the interesting
part for this is that this natural product
anti-microbial penicillin first discovered by Fleming
in the 1920s– the first actual
treatment because of the amount of development
it took was about in the 1940s. But the first reports of
resistance to this compound predated that by over a decade. And this is something I’ll
keep coming back to is that anti-microbial resistance,
even though we heard about it in these scary
stories in the media as being associated
with the clinic, is actually a natural
feature of all microbes. You also take the example of
the sulfonamides discovered by this German chemist,
Gerhard Domagk. The sulfonamides were a
synthetic class of compounds. Their first treatment
was in the mid ’30s. And this is a little
bit more intuitive that the first
reports of resistance about four or five years
later, though we’ve discovered that that was,
again, an underestimate. Resistance to those
particular compounds was well before humans
came on the scene. And so this
particular plot shows you the sort of zoomed
out view for every class of antimicrobials. So what you’re seeing
on the x-axis is time. You’ll see along
the particular plot every single important
class of antimicrobials used in the clinic. And when blue shades to red
is the very first reports of clinical resistance. Another way of
interpreting what you’re seeing on this particular
plot is antibiotic resistance and pathogens is
not a question of if, it’s just a matter of when. Death, taxes, antibiotic
resistance, right? All of them are
predictably going to come up very, very fast. And this encodes an incredible
burden on the human population. Now, some of these numbers
have been challenged, so I’m not going to focus
too much on the specifics, but we do know that a
large number of people, almost a million people
on the planet right now, die due to drug
resistant infections. And this might
scale up if things don’t change by about 2050 based
on this UK prime minister’s report from a few years
ago to maybe one person dying from a drug
resistant infection every three seconds
on the planet. Now, of course, the human
life is the most important in this equation, but it
also costs a lot of money. So the US economy is estimated
to lose about $55 to $60 billion per year because
of the treatment of drug resistant infections. If that 10 million
number is to be believed, it also scales to
about a cumulative hit to the global economy by
2050 of $100 trillion. So $100 trillion,
for those of you, like me, who don’t have
access to that many zeros, is about six years of
the US GDP wiped out. And I will warn you,
I’m a little bit sorry, but this is going to be a
slightly Bad News Bears talk. That exactly at this time, when
we’ve got this massive increase in drug resistant infections,
exactly when we need new drugs to be coming to
market, we don’t. That’s a lots of
reasons for this. So what we’ve decided to do
as a group is focus not only on trying to understand
how we might come up with new
antimicrobials, but take the parallel complementary
view of saying, if we were to
discover key features about antimicrobial
resistance that we might be able to counteract that could
be part of this particular pie. This gets a little bit
molecular in these details, and I’ll try to keep this
at the level that is still relevant to this
discussion, and that is to ask, now that we know that
resistance is a problem, how is it that bacteria
evolve resistance? When they see these
antimicrobials, how do they make
changes in their genomes to become resistant
to these compounds? And to do it in the way that
bacteria evolve any property. And it’s in one of two ways. One way is the way
that we evolve. That is we have– we pass on our genes
over to a progeny. Bacteria do this by dividing. And then mistakes occur during
those replication processes. And if a particular
mistake, a genetic mutation, happens to give your progeny
a selective advantage, it’s going to now take
over the population. And say that selection happens
to be antibiotic resistant. Now that particular resistant
mutant, that resistant kid, if you will, in the bacterial
population will take over. That’s called vertical transfer. But bacteria and archaea, the
other two domains of life, can do something that’s
much more spectacular. Something that, unfortunately,
so far we can’t do as humans. And that is, in one fell
swoop, exchange large chunks of DNA, the genetic material,
with their neighbors. So they don’t have to
wait until they have kids. They can pass on traits to other
members in their population. And as it turns out, if those
particular pieces of DNA encode antibiotic resistance,
in very short timescales, you can take an entirely
susceptible population of microbes and convert them
to be multidrug resistant. And analysis that many
groups have done now suggest and show that because
of the huge selection pressure of surviving under
antibiotic stress, microbes then use this
particular method. This idea of horizontal
gene transfer as opposed to vertical transfer to
become antimicrobial resistant. And so what I wanted
to talk about are a few stories that we’ve
used modern genomic methods and computational
methods to understand the burden of horizontal
gene transfer in terms of who’s the source of
these resistance genes, and how do they
eventually end up in the clinic to cause
the type of problems I just talked about? Now, one thing that we
as biomedical scientists have learned over the
last 10 or 15 years is if you’re a scientist
worth your salt, you have an “ome” at the
end of what you study. Genome, the epigenome. Not to be left behind,
we as the group that study antibiotic
resistance were fortunately given the resistome– actually a very elegant
concept put forward by Gerry Wright from McMaster
University about 12 years or so ago– to describe all of the genetic
material and any arbitrary microbe that will
allow you to be resistant to a
particular antibiotic. So how do you study
the resistome? How do you study all of
these resistance genes? Well, you could certainly
turn to the foundation of microbiology, this idea
of domesticating the microbes by bringing them into the lab. And we should certainly
happy that this occurs because if a clinical
microbiologist wanted to diagnose your
infectious agent, this is what they would do. In the modern era
of genomics, you could then sequence the
genome of that organism in further resistance property. The problem, though,
is we’ve really realized over maybe the last two
decades that most microbes live in microbial ecosystems
all over– in the soil, in our bodies. Virtually anywhere we look. And most of those have
not been cultured. So how do you study things
that have not been cultured, or that we’ve not been
able to domesticate? Well, we can turn to sequencing. This idea of now looking at just
the DNA, the genetic material in the organisms, and infer
what type of functions they might have. In this case, the function
of antibiotic resistance. So at the practical level,
this is literally just smashing up all of the microbial
cells, harvesting their DNA, putting them
through a sequencer, and then inferring what
the resistance might be. And this might seem like you’ve
got the whole part of the pie. Unfortunately, you’re only
iterating on the known. So what does that mean? It means that all
of our databases of known antibiotic
resistance genes have been inferred from
that small, tiny minority of bacteria that
have been cultured. And so then when you
use sequencing methods, you’re just relearning
the few tiny ones that you’ve known before. So what we’ve done
is try to compliment that part of the cryptic
resistome, that part of the pie that is the undiscovered, to try
to improve our ability to not only harvest out and
bioprospect genes that might be a
problem in the clinic, but also to try to be sort
of prescient about what might eventually occur in the clinic. So what is this complicated term
called functional metagenomics that is the method that we use? So the method itself
was codeveloped by Jo Handelsman,
who’s at Wisconsin, and John [INAUDIBLE],, who was
here at Harvard Medical School, back in the late ’90s. They discovered that much
like the natural process of horizontal gene transfer,
if you could come up with a method that forces
that to occur in large scale, then you might be able to use
model organisms like E. coli and test their ability to
express random chunks of DNA from any organism to
encode resistance. And the steps are the following. You extract the DNA, you clone
it in to a library in E. coli. This is an easy to
use microbe bacterium. You get millions of these
random chunks of DNA in E. coli, and then you turn
back to culture. So in this case, much like
a clinical microbiologist would do, you use Petri dishes
that have antibiotics in there. The bacteria that grow now
have a new chunk of DNA that is a resistance
gene, and then you figure out what’s going on. What are those
genes by sequencing? What our lab has
basically done is taken that basic engine, married
it with sequencing technology, and effectively reduced the cost
of doing this by about 1,000 fold. Now, let’s say for
the sake of argument that you completely zoned
out for those last two slides and don’t care about any
of those technical details. The only thing that you need to
take away from what I just said is that with this
method, you can start out with any arbitrary
group of microbes, whether they’re cultured or
not, and after pumping them through this particular
pipeline, which is now a lot cheaper to use,
you’ll get a catalog of all of the genetic material that is
compatible with horizontal gene transfer to make a
particular bug like E. coli resistant to antibiotics. So we apply a
method of this type to then interrogate all sorts of
different microbial communities for their potential
to contribute antibiotic resistance
to the clinic. So this is one way in
which we view the world. This cartoon schematic
of interacting habitats. And the reason I put this
up, and you’ll see this a couple times, is to recognize
partly that microbes don’t care about our definitions. We might define something
as a habitat boundary. Microbes can traverse those. And so what we’d really
like to understand from an ecological perspective
is that even though resistance really is most acutely
problematic in the clinic, because microbes in
different habitats could themselves
transfer over eventually into the clinic or their
genes could transfer over, we would like to understand
all of the interactions that are shown here. So I’ll walk through
a couple of habitats that we’ve
particularly focused on to discover resistance genes, to
assess their risk of getting it in the clinic, and
hopefully also to mitigate those particular risks. And I’ll start out by not
going into too much detail and just giving you a couple of
highlights that we’ve published on from our investigations
of the so-called commencal microbiome. So these are the good
bugs, for the most part, that live in and on our bodies. The one that’s been studied
the best is the gut microbiome. So, again, this is the trillions
of microbes that live in us. They’re basically another organ. And because they’re
constantly subjected to antibiotic pressure,
it makes sense that they might be a source
through which pathogens could pick up resistance genes. So one of the
things that we were surprised to find in
collaboration with Rob Knight’s group as well as Maria
Gloria Dominguez’ was that through
them, we got access to fecal samples
and oral samples from an Amazonian tribe
at first contract. I can’t tell you, how do you
convince a group of people who’ve never seen the
outside world to give you a fecal sample? I think that’s probably how our
first alien contact would be. They’ll be taking
microbiome samples from us. But Erica Pehrsson, a genetics
graduate student in my lab, accessed those samples, used
the methods I just described, and we were shocked to
find in these people who had no history of any
antibiotic use resistance genes against
modern antibiotics. And you’ll see when I get to
the soil why this makes sense, but we thought this is an
important thing to recognize– that resistance
against these compounds is not something that we
invent through our use. It already exists as a natural
feature of these ecosystems. We just give selection pressure
by using the antibiotics for them to amplify. Now, I’ll give you another
example of something that’s at the other extreme
of the spectrum [? will ?] [? be ?] no invulnerable
populations. They get a ton of antibiotics. And that is, for
instance, in this case, preterm human infants. So we know that rates of
prematurity are going up, and especially when we look at
very low birth weight infants. So these are infants that are
born about 10 weeks too early. These kids are highly
immature in terms of their immune systems. And because of their
risk of infections, we give them a ton
of antibiotics. Well, not me. I’m not a physician. But the field. And so, for instance,
in this cohort, which is describing kids born
in St. Louis at our Children’s Hospital, virtually 100%
of these kids born 10 weeks to early get antibiotics
as soon as they’re born for the first
couple days of life. And as you’ll see in the
plot next to the histogram, the top two drugs that
they get while they’re in the neonatal ICU
are antibiotics. Four of the top eight drugs
that they get are antibiotics. So we step back again from
an ecological perspective and you think about the
fact that these kids require this commencal
microbiome to set up. What you’re doing with each
of these insults is you’re carpet bombing an ecosystem that
should be setting up normally. So we wanted to assess what
the collateral damage might be. And so one thing
we found– and this was work led by Molly
Gibson, a graduate from our computational biology
program in collaboration with a couple of pediatricians
at Wash U, Barb Warner and Phil Tarr. We had access to a large
cohort of fecal samples. They actually have about
75,000 fecal samples bank, which is probably the
largest collection of poop in one particular location. And we access only 400 of
those in two groups of kids– ones that had lots of
antibiotic exposure and ones that didn’t have very much. And the thing that I
just want to emphasize for the non-clinical
microbiology aficionados, the main thing that
Molly discovered that was surprising
to us was the identity of the bacteria that
was surviving all of these insults in these kids. And so, again, don’t worry
about the specific names, but you have to take
my word for the fact that the bugs dominating
the guts of these very vulnerable kids are the bad guy
bugs on the CDC list, right? These are not bacteria that we
associate with good gut health. They’re the bugs that usually
cause nosocomial hospital acquired infections. And perhaps now that you think
about the selection pressure, it’s not all that surprising. Who else can survive
those insults except the really hardcore
drug resistant bugs? But the silver lining there
is that through a series of computational approaches– again, the aficionados
call it machine learning. That’s partly because
it sounds fancier. It’s really pattern recognition
using computer algorithms. So what Molly did was
she took all of this data that she had gathered in
terms of who was there in the microbiome and
their resistance genes and trained the model, and was
fortunately able to predict how the microbiome
would change based on just a couple
of these features, with about 85% accuracy. Why is that something to
consider as being important? Because it suggests
that a potential future of personalized medicine,
where each of us, anytime we get sick, we might be able to
get our microbiomes sequenced. As these type of
methods improve, your physicians
might be able to take the information about your
specific microbiome state and tailor your antimicrobial
therapy in a way that not just kills the infection,
but protects your microbiome. So we’re hoping that this
is where this will go. Now, a picture tells
a thousand words. And I can’t think of anything
besides this particular questionable parenting
event in really telling you that resistance cross
habitats, right? And so we decided to
take this concept– and here’s the slightly more
boring version of that slide, where, fortunately, through
a number of international collaborations led by Erica,
who I already mentioned, and Pablo Tsukayama, another
graduate student in my lab, we accessed samples from
a village in El Salvador in a slum outside Lima, Peru to
look at potential interactions between microbes and resistance
genes between people– so we have fecal samples– and as many
environmental samples we could get at the same
time with GPS coordinates. So soil samples,
and sewage samples, and subsistence
animals and pets. And, again, I won’t belabor
all of that incredible data analysis that these
folks did, and just jumped to the
conclusion that we think has impacts on public health. And that was they
were able to identify using this analysis key hot
spots for resistance gene exchange. So what this means is they
identified particular features, particular habitats
between these people in the environment that have
clear evidence for having the same resistance
genes in those habitats in humans as well
as the environment. And we found two such
key habitats, again, which could be really impactful
for public health in El Salvador and in Peru. So across the top panel, you’ll
see pictures of chicken coops in El Salvador in the
village, and that appeared to be one of these hotspots. So this is the droppings
from the chickens, who are subsistence animals. Had exactly the
same resistant genes as the humans, as
well as the soil, implicating those
chickens as a route between moving between
humans and the environment. The reason this is
important to consider, especially here
in the US in terms of its translational
impact, is the folks in El Salvador in this village
don’t use any antimicrobials in growing their chickens, where
almost any chicken that you or I would eat has
probably been grown under the pressure of
lots of antibiotics. If you think of
what’s happening here, consider what might be happening
when we’re eating chicken under antibiotic pressure. The story was a little
bit different in Lima in terms of what we found there. There, we were able to implicate
the sewage treatment system as a hotspot for such exchange. So this is a massively
crowded ecosystem. All of the effluent sewage
goes through a single treatment plant that’s cleaned up based on
pretty sound civil engineering principles. And then once that
water is clean, it has to be dumped somewhere. Now, in Boston, that’s
dumped in your harbor. Far out, but still
in you harbor. In Lima, being a
desert, that water is used to irrigate every single
field and every single place that requires water. And so you might now
consider that even though the civil engineering
system was working to destroy certain
types of microbes, because it was allowing
other microbes to thrive, that it picked up
resistance genes from the human microbiome,
they are now disseminating those genes all over Lima. So this is where, again, these
kind of bioprospecting methods to hunt for resistance
genes comes into play to have potential
public health impacts. So I’m going to leave
the human microbiome and spend a little bit
more time on a habitat that maybe a lot
of people may not consider as a really key part of
antibiotic resistant exchange, and that is the soil
or the soil microbiome. So why look at the soil? As it turns out,
there’s a long history, decades worth of really
important work, implicating the microbes in the soil,
perhaps the most diverse ecosystem of microbes
on the planet, in not just the
production of antibiotics, but also in
antibiotic resistance. And I’m only going to give
you three examples of work by other people that
really inspire our work. One was this beautiful story
by Gerry Wright and colleagues where they went into the
Canadian Beringian permafrost, cored out samples that they
could carbon date to be 30,000 years old, sequenced the DNA
there, and then interrogated that DNA for whether they could
find antibiotic resistance. And lo and behold,
they found resistance to modern antibiotics,
providing clear genetic evidence that resistance in
environmental microbes vastly predates any
human use of antibiotics. So kind of putting the nail
in the coffin of saying, resistance existed before
we came up with antibiotics or discovered them. Another piece of
evidence that’s important is put forward by kind
of the grandfather of the antibiotic resistance
field, Julian Davies, who now, more than 40 years
ago, recognized that almost all
of these compounds that we call antibiotics
are natural products of soil bacteria. So all of those
Nobel prizes that were given for the
discovery of antibiotics came from people hunting
in the environment– This is the case with amino
glycocides, as an example– where you found these
extracts from the soil and realized, whoa,
they’re able to kill these particular
pathogenic bacteria. That’s because of
these small molecules. And so what Julian posited,
and it’s been supported since, is that these producers must
be the original evolutionary progenitors of
antibiotic resistance. And why is that? Because these antimicrobials
target, as I said, key processes of bacterial life. So if they didn’t
have resistance at the same time as production,
they would commit suicide. So it would not be a
terribly interesting evolutionary process
to figure this out without having
resistance just in time. Since all of this happens
millions of billions of years ago, they have also given
all of their neighbors the pressure to
evolve resistance, hence soil is the original
source of resistance. And then finally, to show
that we as humans can actually have an impact in
a bad way, this is evidence from
European archival soils of about 70 years, where this
group led by David Graham was able to measure the
abundance of resistance genes and find that over the
years of human use, their abundance has gone up. So based on all
of this evidence, the one thing that
you might speculate is that for this
reason, every resistance gene, every resistance element
that you find in a pathogen, a disease causing microbe
must be the same as the one– they must be identical to things
that you find in the clinic. Sorry. In the soil. And bizarrely, that’s
not what we found. So when we started
working on this and looking through
databases, virtually all of the resistance genes
that people have worked hard to discover in
soil microbes were vastly different from
the type of microbes that we see causing infections. So this seemed like
a bit of a conundrum, and our hypothesis
was maybe we’ve just been looking at the
wrong slice of the pie. Now, one thing that
Gerry Wright’s group had done to really emphasize
that the bugs in the soil are highly resistant was
they went and they isolated, in the seminal paper from
2006, a whole group of microbes that there are no other
producers of antibiotics. So all they did was
culture them up, noted the produce
of antibiotics, and then tested to
see how resistant are these nonpathogenic
bacteria to antimicrobials? And shockingly, on
average, these bacteria were resistant to seven
or eight different classes of antibiotics, even though
they’re not pathogens. So that’s actually more
resistant than most pathogens. So now this should be used
as evidence, probably, that these guys were the
guys that were producing the resistance genes, right? But, again, we stumbled
across this problem where when people analyze the
genes that these producers have, they look completely
different from the genes that the pathogens have. So how do you reconcile
that particular disparity between the phenotype
of high resistance and the genotype of
being very different? So one way in which
you can reconcile that is to say, OK, I
buy the argument that these guys are the
original donators of resistance. But it happened so long ago,
millions to billions of years ago, that it’s not really
relevant in terms of us treating clinical problems. So that’s one hypothesis. The other hypothesis
is maybe there’s still another missing link. Maybe there are microbes in
the soil that are contributing, we just still haven’t
found them out yet. And so this now jumps over
to this crazy experiment that we did when we
were in George’s lab where we said, OK, what about
these weird bugs that are able to eat antibiotics, right? So you’ve got the producers,
you’ve got the resistors, and now you’ve got the eaters. So in this slightly, again,
ridiculous experiment, we tried to enrich microbes in
the soil that could– again, not just resistance antibiotics. They were using them as
their sole source of energy. It wasn’t a terribly
elegant experiment. It was just a lot of culturing. And when we were able
to successfully identify all of these microbes that
could eat antibiotics, we then tested their resistance. Now, what I’m showing
you in the bottom in blue is the same data
from Gerry Wright, where his producer organisms,
on average, resistant to about seven or
eight antibiotics. Now you look at the eaters
on the right hand side. They’re, on average, resistant
to 17 or 18 antibiotics out of 17 or 18 antibiotics
of which we could test. Basically, they don’t care
about the antibiotics. Actually, they do
care about them because they’re eating them. So we think now we have finally
stumbled across that missing link. The potential that
these eater organisms might be contributing to
resistance in the clinic. Now, before I get to showing
how we actually did that, I’m going to jump ahead 10 years
from that original discovery, which is how long it
took us to figure out how it was that these bizarre
bugs were eating antibiotics. How were they surviving
on the antibiotics the way that most bugs
survive on glucose? And so this was work that
was led by many people, but primarily by Terence Crofts,
a recent postdoc in my lab. We went to about five
of these soil bacteria and used a battery of techniques
to really pin down how the heck they were using those compounds,
which normally kill bacteria, as their source of food. And you almost brush past
it if I don’t point it out. One of these bacteria
can eat penicillin as its sole source of carbon,
but it can’t use glucose. Doesn’t like sugar,
like antibiotics. And so I’m not going to belabor,
again, the molecular biology. All I’ll say is that through
a lot of different people’s contributions, we were
able to figure out that there are three key steps
of how these bacteria eat antibiotics. And the important part is
that the committing step is an antibiotic resistance gene. So the very, very first thing
that these bad bugs need to do is to detoxify this
toxic compound. So they break it open exactly
the way the pathogens do, and then they digest it using
a couple of different methods. And just to show
that we really, truly understood one way in which
to show that the genes that underlie this
particular phenomenon are sufficient for
this particular trait is you could transfer that
trait to another bacterium that doesn’t have this. So that’s something else
that we did, led by Terence, was we were to take all
of the inference he made from these soil microbes,
transfer those genes into E. coli, one of these benign
microbes in the lab, at least, and he was able
to convert E. coli into an eater of penicillin. Now, we’re not doing
this because we’re crazy. Someone might ask,
why the hell are you turning E. coli, which can,
you know, some versions, make people sick. Why are you making
it eat antibiotics? Well, in this case,
our motivation was partly to prove
that we really understood how it
worked, but also because it might be
used in the future for clean up of antibiotics. Not with this basic
work, but perhaps in the future where if you
could have microbes that have these abilities to destroy
these antibiotics, which can contaminate systems, like we
had seen in the sewage systems, maybe use as a next
generation approach to clean up, much like
we clean up our sewage, using microbes to
destroy these antibiotics before they go
out into the wild. OK. So now let’s step back to say– so that showed that
we can figure out how to eat the
antibiotics, but what about this fundamental question? Are they the missing
link between the soil and the clinic? And so to do that,
we step back and– this was work done
by Kevin Forsberg, a graduate in our
genetics program, and Alejandro Reyes, who
was a graduate student in Jeff Gordon’s lab. So they applied that
method many, many slides ago, where
we go and discover new antibiotic resistance genes. And they applied it to
these antibiotic eating bugs and said, OK, using
this method that doesn’t require a priori
knowledge of what resistance genes are, looking for
cryptic resistance genes, can we discover
the missing link? And sure enough, they did. In a single experiment,
they’ve now increased evidence by an order of magnitude
of the number of genes in the soil that were exactly
the same as resistant genes and pathogens that
cause disease. And, again, I won’t talk
about the specific details of the resistance genes in
terms of what the names are. But the key features that
certainly made us a little bit scared was the fact that
between those 10 genes, they knock out four or five
classes of antibiotics. We find them not only in those
bugs, but also in pathogens. And in that map that you see,
any country that’s shaded dark is a country that is deposited,
or someone in that country has deposited at a pathogen,
with a resistance gene that was identical to those genes
we discovered in US soils. And to emphasize that this was
particularly problematic, what you’re seeing here
is a comparison of all of those genes
compared to genes that have been found in pathogens. And you’ll see that there
are many genes there. So across the bottom are the
genes that come from the soil; across the top are the genes
that come from four or five important pathogens;
and anywhere you see gray shading,
that’s identical. 100% DNA identity. And what we observe here
is that not only did we find that they’re the same
single genes across these bugs, but actually they’re
clustered together. So anything in red that you
see is an antibiotic resistance gene. Anything in yellow is a gene
involved in horizontal gene transfer. So they’re the vehicles
that move genes around. So what we’re seeing
here is evidence that bugs in the soil and
bugs that cause disease can move large chunks of
DNA between themselves that can knock out
three, four classes of antibiotic resistance in
horizontal gene transfer events that can happen on
the order of minutes. So, again, now, taking
all of this together, reconcile with the
day that was before, it suggests that we have
discovered the missing link, and these are the
bugs that we might want to focus on as we consider
challenges to the clinic. Now, I will again
preface all of this I guess by saying that
we stumbled across this. This was not something we
originally set out to do, because only someone insane
would think about looking for antibiotic eating bugs. But since we had this, and
now we had this data in hand, we wanted to ask,
how much of this is this the exception
versus the rule? That is to say, because we have
gone to this bizarre experiment of culturing bugs that
could eat antibiotics, which we discovered were
these multidrug resistant proteobacteria, now
Kevin, who led that study, teamed up with Sanket
Patel in the lab to ask, again, when we stumble
across those weird bugs, are they representative
of all bugs in the soil, or are they sort
of these weirdos? And basically, the short
answer to that story is they are weirdos. And, really, the
only important part of this sort of
complicated slide was we now collaborate
with a couple of ecologists in Colorado, Rob
Knight and Noah Fierer, got access to a much
larger number of soils, about 20 soils; applied
the same methods; discovered thousands of new
resistance genes in the soil. But luckily, most of
those resistance genes are not ones that we
found in pathogens. it appears, really, it’s only
that tiny weirdo minority there of proteobacteria
that are the ones that are the missing link. The other thing we
discovered is all of those genes we found
from these bugs that are not the same as in pathogens,
they don’t have those same linkages with
mobilization elements that cause those
genes to hop around. So we know that in
pathogens, for those genes to move between
different bugs, they require those accessory
elements to actually move the DNA back and forth. These guys in the
soil didn’t have them. So how do we put all of this
together towards something that could be impactful
for work in the future? We think what we’ve done is
we’ve discovered, indeed, that the soil is rife with
antibiotic resistance genes, probably reflecting this
evolutionary pressure of antibiotics being
produced in the soil. But, fortunately,
most of those genes are not at risk for being
acquired by pathogens, but there are these
very specific bugs, but now we know the identity,
that we can focus on so that we can prevent
them from traversing these particular boundaries. But, as it turns
out, there is still something interesting to learn
from these particular bacteria from the soil that have these
cryptic resistance genes. So one hypothesis that we came
up with at the time was to say, could these be
bellwether events? Could these genes
that we have seen that haven’t popped into
pathogenic bacteria yet, could we stack [INAUDIBLE] them? Could we evaluate their
risk across those thousands for the few that we
think might emerge into the clinic in a few years? So why would you
want to do that? Why would you want
to know what is going to happen in the future? It’s because if you know
which ones are at great risk, those are the ones you can
focus on for mitigation. The ones that you
might try to destroy, the ones that you
might try to diagnose. And this is the story
I’m going to tell you about one such class
of enzymes, and that’s against the class of antibiotics
called the tetracyclines. The tetracyclines are a
very, very important class of antibiotics. They were discovered
in the late ’40s, and they continue to be
one of the major classes of antibiotics used in the
clinic, in agriculture, and aquaculture. And the way they work is they
gunk up a bacterium’s ability to make proteins by targeting
this thing called a ribosome. So big surprise, because they’ve
been used against pathogens, pathogens have figured out
how to be resistant to them. And the two main ways
that they do that is they either pump the drug out or
they protect their ribosomes. What hadn’t been
observed very much, though, was a third
mechanism of resistance, which a lot of bacteria
use, and that is destroying the chemical itself. And that seems to be
a pretty good idea if you’re a bacterium in terms
of wanting to be resistant. If you destroy the
chemical, it can’t hurt you and it can’t hurt
your population. That’s how, for instance,
those beta-lactams, those penicillins,
that’s they’re destroyed by most pathogens. There are enzymes that cleave
them and break them apart. So we found evidence of
only one such instance. There was a nonpathogenic
bacterium from human guts that had been shown to have this
particular enzymatic activity. So we decided to say
we’ve got this treasure trove of newly discovered
resistance genes in the soil– could we find evidence that
these guys might actually be lurking somewhere beneath
the detection threshold? Motivating this was,
again, as I mentioned, the fact that the tetracyclines
are a unique class where not only are they old,
but they’ve kind of gone through this Renaissance
of being modified and getting newer and newer
versions of them. Including– and this is
something that’s really important to keep in mind– the last three drugs that were
approved by the FDA in the US as antibiotics are
all tetracyclines. One in August of 2018 and
two in October of 2018. They’re a really important
emerging class of antibiotics. One other thing to note
is that those drugs that are coming to market have not
been considering, obviously, a resistance mechanism that
has not been seen in the clinic yet. So we want to make
sure we can enable the protection of these
particular compounds by seeing whether there,
again, is this lurking danger against them. So to do that, Kevin, who
had done the soil work, basically discovered
this new class of enzyme across the soils. I won’t walk you
through all of the data. All I’ll say is that
he was able to show that about 10 such
enzymes existed that can destroy the
tetracyclines and a whole bunch of soil bacteria. They were completely novel. They were not
related, for instance, to the ones that had been
seen in the human gut before. And the only gene or enzyme that
we could find in any database that they were similar
to was a gene legionella. So legionella causes
legionnaires’ disease. It’s a soil derived
pathogen. And sure enough, this was something lurking
in the genome of legionella that no one knew was a
resistance gene, that we were able to now show can cause
legionella to no longer respond to tetracycline therapy. So that’s all well and good. What is the clinical relevance? So it turns out Drew Gasparrini,
who recently graduated from my lab, went through
all of those selections that had been performed in my
lab across many, many, many different habitats from human
fecal samples, and soils, and sewage samples from
El Salvador and Peru, and started looking to see
whether maybe more of these are there. So just a quick
cursory survey, he found that there were 70 more
such enzymes that existed just in our freezers. And in this case, good for
our funding and really bad for humanity, just in
the time that we started working on this,
those resistance genes started popping into pathogens. Exactly what we had
speculated would occur. Really hoped that it
wouldn’t, but it did. So this is why we’ve now decided
to take a very concerted effort to understand those enzymes,
understand the vulnerabilities to hopefully destroy them. And so that’s exactly
what we’ve been able to do, turning
back to something that we’ve learned
from the penicillins. So one way in which
the chemists have helped rescue the penicillins
a little bit is to recognize that, yep, the penicillins
or the beta-lactams are degraded by this
particular professional enzymes called beta-lactamases. So if you can come
up with chemicals that break the resistance, you
can reactivate the antibiotic. So this is the idea for
a beta-lactam, which is the antibiotic, and a
beta-lactamase inhibitor combination, so
breaking the resistance. So we said, hey,
why not try that? Let’s try to find
compounds that kind of look similar to
the tetracyclines that might be able to gunk up those
tetracycline disrupt cases. And, again, this is where
serendipity comes into play. You may not believe me, but the
very first compound we tried had that property. And it happens to be a compound
called anhydrotetracycline. The name is not
all that important. But again, what
we’re able to show is that if we add
anhydrotetracycline into the mix when
these enzymes are working to degrade
the tetracyclines, they don’t work anymore. And what you’re seeing
here is something that’s a little bit more
relevant to the clinic. There’s a way in which to
measure the resistance that is encoded in our
organism by looking at how well it will grow in
the presence of the antibiotic. So on your left hand
side, you have seen what’s called an Etest strip. This is basically
a piece of paper that has an antibiotic
across a gradient that you put on a lawn
of growing bacteria. If the bacteria can’t grow close
to it, they’re susceptible. They’ve being killed
by the antibiotic. If they grow really close
to it, they’re resistant. You can see in the first
panel the bacteria growing really, really close to it. So this is E. coli with
one of these enzymes growing close to tetracycline. They don’t care that
tetracycline’s there. We add this newly discovered
inhibitor, and lo and behold, tetracycline can work again. And so we are now really
heartened by this discovery. And through collaboration by
a couple of other researchers at Wash U, an organic
chemist, Tim Wencewicz, who trained with Chris Walsh
here, and [INAUDIBLE],, a structural
biologist, we have now been able to take this initial
sort of prediscovery almost, make new synthetic analogs
to that originally discovered inhibitor, and now
we’ve been able to make even more potent inhibitors
of these particular enzymes. And, again, now that we
were spending more time looking at more and
more of these genes, what we’re seeing here
about 20 or so of the genes that we’ve been
working on so far. The ones in red are
the ones in the soil. The ones in blue are ones
that now we’ve discovered live in the guts of healthy humans. One of the things
that’s, again, motivating us to sort of work
fast on this is, I already told you,
one of these enzymes in the soil we
found in legionella. Just last year, we
discovered that one of these enzymes that destroys
every single tetracycline we’ve thrown at these
particular enzymes– including the compounds that
were just approved by the FDA, they destroyed those too. That gene or that enzyme we
found in a cystic fibrosis patient sample in pseudomonas. So that’s really scary. But the good news is our
hypothesis came true. By focusing four or
five years earlier and trying to find
those inhibitors, we now have inhibitors
that can resensitize that pseudomonas isolate
to those latest generation tetracyclines. OK. So I’m going to end my
talk now in the next couple slides with a simple question–
why are we not dead yet? With everything
I’ve just shown you, why are we not
dropping like flies every time some highly drug
resistant bacterium comes around? And, certainly, the CDC has
this almost Pokemon card deck of bad guys that tells
you you should really be scared, you know, in
terms of their hit points. You’ve got acinetobacter
and carbapenems, resistant antibacterial pseudomonas. These beautiful pictures of
things that if they get in you, they’re going to
kill you, right? So what can we do about this? I showed you one strategy. Do we have anything
in our arsenal to be able to fight back against
these multidrug resistant organisms? So it turns out there are
some strategies outside of just coming up
with new antibiotics that we might be able to use. And this is where we turn to
some evolutionary biologists and systems biologists
to think sort of out of the box a
little bit to figure out, what are general
strategies we might have to do to kind of flip the
table on the drug resistant organisms? And so one idea is called
a selection inversion. So what does that
term actually mean? It means trying to come up
with some kind of magic sauce that transiently allows
the drug’s susceptible version of the organism,
the one that you can kill, to outcompete the drug
resistant version. Why would you do that? Because as you’re shown
here, normally you’ve got the susceptible and the
resistant guys mixed together. In this case, the yellow
bugs are susceptible, the blue bugs are resistant. Normally what
happens is you bring in the antibiotic, sure
enough, the susceptible ones, by definition the yellow
ones, get wiped out; the blue ones survive. And now you’ve got a drug
resistant population. Selection inversion,
this magic sauce, is somehow going to suppress
the blue resistant guys to let the yellow
ones take over, and then you kill them
with the antibiotic. So how do you do that? There are a few strategies
that been proposed. One is drug cycling. So if you were
able to find drugs that work in very different
ways, what you might do is use drug A until you get a
resistant population to drug A, and then much like crop
rotation, you switch to drug B. Because those resistant
mechanisms don’t necessarily talk to each other, you can go
back and forth, back and forth, and kill off these differently
resistant populations. This is only,
unfortunately, going to work for a short
period of time because, as I showed
you in the soil case, sometimes multiple
resistance genes can transfer at the same time,
and then you’ve lost this particular ability. So another strategy
is to recognize that drugs of different classes
sometimes can combine together to be better than the
trivial sum of their parts through a property
called synergy. So what you’re seeing here
in this particular panel is normally you take a
population that if you hit it with drug A, it doesn’t care. Individually hit it with
drug B, it doesn’t care. But when you combine
drug A and B together, there’s something magical
about that combination where they potentiate each other,
and they can synergistically, in combination,
kill those bacteria. So that strategy two. And then strategy three,
which I think is actually the most elegant, is
this concept called collateral sensitivity. So we try to explain
what that term is. Imagine for a second that
as a bacterium becomes resistant to one class
of drugs, class A, the mechanism that it
uses to become resistant opens up an Achilles heel. Something about what it’s doing
to become resistant to drug A opens up a vulnerability
against drug B. That’s the collateral
sensitivity. Because of resistance one, you
become susceptible to drug two, if you will. A St. Louis summer analogy,
for collateral sensitivity, is imagine for a second you want
to go out into the St. Louis summer and you don’t want to
be hit by mosquitoes or bitten by mosquitoes, and you also
don’t want to drop dead because of the heat. Now, imagine that
what you decide to do is to make sure that
you don’t want to get bitten by those mosquitoes, you
put on a big moon suit and you walk out into
the St. Louis summer. You’ve certainly protected
yourself against that mosquito, but you’re going to drop
dead because that moon suit is going to suffocate
you in terms of the heat. That’s the kind of weak analogy
for collateral sensitivity. So I’m just going
to end by saying we were able to implement this
against one of these scourge pathogens, methicillin-resistant
staph aureus, or MRSA. So MRSA is a deadly bug. It kills something like
11,000 Americans when it gets into the bloodstream. It’s resistant to the important
class of beta-lactams, these penicillins. And so we thought, could we
use some of those features that I told you about
before in combination to try to knock out MRSA? And so in this story, Pat
Gonzalez, a graduate student in genomics in my lab, was able
to find a combination of three beta-lactams, so that same
penicillin category, completely FDA approved generic
drugs that on their own are completely
useless against MRSA. But somehow in combination,
they work together to overwhelm MRSA’s
defenses and are able to wipe it out through
this process of synergy. But perhaps more
importantly, they had this process
or this property of collateral sensitivity. That is to say, if MRSA
becomes resistant to any one of those drugs, it
now becomes more susceptible to the other two. And why is that important? Because if you imagine you’re
a bacterial population, you’ve got these three drugs
around, you think you’re clever and you become resistant to one. Now you’ve got two
other drugs still around that are much better at
killing you, which locks them against resistance. And we tried very hard to
make these guys resistant, and we couldn’t. And that was fortunate. And then, finally,
in collaboration with Mayland Chang and Satish
Adusumilli, researchers at University of
Notre Dame, we tested to see whether this really
interesting test tube result could actually translate
into an animal model. So they happen to have
an animal model of really severe bloodstream
infections with MRSA, and were able to fortunately
show by working with them that this completely
cleared that infection, this triple antibiotic
combination, within 24 hours of treatment. And so we’re pretty jazzed
about this because, again, this is composed of
three generic drugs, and so we have a
lot of information about how they work. Fortunately, they all co-deliver
in the same way currently. And so we’re hopeful that this
might be one particular quiver in the arsenal against
this deadly infection, but also sets up
a paradigm of how we might be able to repurpose
other existing drugs to come up with at least stopgap measures
against infectious diseases. So with that, I’ll
end by not addressing this particular
elephant in the room. And that is to say,
everything I’ve said is focused kind of on
human use of antibiotics, but in this country
still, it’s estimated that something like 80% of
antimicrobials by weight are still used in the
rearing of food animals. And I would argue, without
getting into that debate just right now,
that that’s perhaps not a great use of
this natural resource. And there are different ways
in which to mitigate that, but I just want to say that
this is something I have not discussed today,
but I think it’s very germane to this
particular discussion. With that, I also say that
sometimes I look like this, but sometimes I also look
very Missouri with my lab. And I’d be remiss not to,
again, say that, really, anything of importance that
I might have said today is because of the
wonderful people that I have the privilege
of working with. Both the people my
lab; alumni who’ve gone on to bigger
and better things; people at Wash U and
around the world; and, obviously, the
funders that are generous to support this work. And if there’s one the PSA
thing that you remember when you leave today,
it’s this– please don’t use antimicrobials
for viral infections. I’m happy to take any questions. [APPLAUSE] – Thank you very much. That was great. I have a question
about teaching. So part of the reason that
we started this Undiscovered series was to talk about
how we have to teach science and how we have to
acknowledge these kind of unexpected findings. So I’m just curious,
in your teaching now, genetics is a constantly
evolving, relatively new subject. How do you incorporate this
kind of thinking when you teach especially undergraduates? – Yeah. That’s a very good question. And I think there are already
great measures in place. And doing this at
Wash U where part of our undergraduate curriculum
has a discovery element right off the bat. So there’s this amazing
thing called phage hunters, which a lot of universities
around the country have committed to. And anyone interested,
even if they’re not declared biology majors,
they have this sort of experience based
class where you’ll still learn the fundamentals
of microbiology, and bioinformatics,
and sequencing, but they’ll do it by going out
into the field, for instance; harvesting soil; getting their
own phages, the viruses that infect bacteria; going
through the isolation process; going through the sequencing
process; and the annotation. So I think sometimes
the best way to teach the sort of wonders of
science is to actually perform the science yourself. Not everyone has the
opportunity to immediately join a research lab. So I think engaging
students very, very early on in discovery based science– it’s, of course, exceptionally
important to teach them the value of hypotheses. But also to have that
kind of exciting moment to say, no matter what you
do in a phage hunters class, you’re going to
find something new. And then once you
have something new, you can test a particular
biological question. So that’s one aspect that
we find pretty useful. The other is my lab always
hosts a pretty large cadre of high school students
and undergraduates all through the year. Sometimes the people in
my lab complain about this because we have a dozen students
that come in additionally in the summer. But I try to convince
people that you’re doing this for multiple reasons. There’s no other way to maintain
the steady pipeline of students interested in science
without allowing them to do the science themselves. And also, the only way to learn
to be a mentor is to mentor. And so that’s why I strongly
encourage every single person in my lab, independent
of what career stage they’re at, to take on an
undergraduate or a high school student to be able to see
what it is like to teach. I think those are two
ways in which could work. – Thank you again. That was fantastic. So the general mechanism
of drug resistance or the way the bacteria
has become the superbug is pretty well understood,
thanks to you and others. So what is the
practical solution other than to tell people not
to take antibiotics and sort of sit out the cold is something
that most people are not going to, in practice, follow. So, in general, what
would you suggest? – Yeah. There are a few things. Actually, I would say
that education campaigns in countries that have tried
to pretty openly discuss when it is appropriate
to take an antibiotic– Canada is an example– have been successful. It’s just that the
timelines of turning people’s opinions around are
not on the order of weeks or months. They could be years. And so I think
reframing antibiotics as a natural resource, as
something that’s finite and as something that really
requires a sort of conservation mentality, and then tie
that to the fact that there is collateral risk when
you sort of willy-nilly take an antibiotic. I think a large part of
being able to turn that table to sort of preserve
those antibodies for when we need them is education. When I say education,
I just don’t mean to the consumers
of the antibiotics, but also to empower
physicians, pediatricians, for instance, to
say, here’s why– ultimately, you’re
still going to have to respond to the really
aggressive parent who says, I don’t really care. I want this drug for my kid. Fine. But if you can spend maybe
a few minutes to explain to the parent, here’s why. Here’s a brochure. Here’s something that
you should consider. Here’s the risk that
your kid, five years down the line because of this
unwarranted antibiotic, has a greater chance of
something bad happening. I think that is a large
part of change in practice, I think, for turning
the tide there. I think the other also is sort
of better reporting structures. So we, in clinical fields
and biomedical fields, have to report every
single antibiotic use. We still have no
legislation in this country, and it’s never been
the case, where any agricultural
use of antibiotics needs to actually be reported
on quantitative scales. And I think that’s a problem. – Thank you for your nice talk. I just wonder if, given
the current technology of human manipulation of
genes, especially gene editing, so convenient and widely used
by humans, I just wonder, does this in any way
affect the antibiotic in a gene’s evolution,
the speed of changes? – Yeah. So let me just restate your
question so I understand it. If you’re asking whether
the ability to gene edit in a human genomic context
has impact here, I don’t know. – Not only human,
but all the other– – Right. So the same methods
that people have heard in the media,
these CRISPR systems that can go out and very precisely
edit various aspects of any genome, there
is a lot of promise to use methods of that
type to perhaps go after the resistance genes
themselves in the bacteria. So one way in
which to counteract the emergence of
resistance might be directed targeting of
those resistance genes and to blow them up. And there’s been some
really, really promising work from a lot of researchers
here in the Boston area to show a decent proof of
concept that that can happen. And the reason that those
might be really elegant methods is because the reason resistance
has a really high potential of arising is because
you’re challenging the life of the bacterial population. It’s a life or death decision,
so there’s immense selection to be able to become resistant. In principle, if you were to
destroy the resistance gene, then you might be able to
get nonresistant versions of those bugs to survive,
and the selection pressure might not be as high. So, yes, absolutely. I think those gene
editing methods might be very useful as part
of the arsenal in the future to help mitigate this problem. – Hi. Thank you for the talk. It was very interesting. I have a bit of a technical
question about the strategy that you use for discovering
potential new antibiotic resistance mechanisms by
transferring them into E. coli. Some genes will definitely
be and have been discovered in this way, but
there definitely could be genes that even
if you put them in E. coli, they’re not compatible with
the machinery of E. coli, they wouldn’t actually protect
it from the antibiotics, but they would some
other bacteria. Do you have any
way of estimating the proportion of how
much you’re missing out versus how much you’re finding? – Yeah. That’s an excellent question. And so even to expand
on that, E. coli is in this group of organisms
called gram-negative bacteria, and there are a whole class
of really important pathogens that are gram-positives. And the reason this detail
matters is those drugs don’t work against the
gram-negatives at all. So E. coli is intrinsically
resistant to those antibiotics. We would never discover
any resistance genes against those
important antibiotics like vancomycin and linezolid. So one way of which to
address that problem is take exactly the
scheme I showed you and swap out E. coli for
another host bacterium that has the properties
you care about, and do the same type of screen. But then the direct
answer to your question of the estimate of how
much we might be missing is, again, we think
that as biased towards particular bacteria. So the way we’ve
estimated that in the past is we’ve done that
same experiment, but with a DNA
material that goes in from defined bacteria of
particular types, for which we already know what the
resistance genes might be, and then see which
ones come through. And, again, what we see is that
there are two layers of bias. One layer of bias,
obviously, is against any of the gram-positive bugs. Those mechanisms just
don’t come through. And then we also have a little
bit of bias towards, again, that kind of GC-content
and codon bias that could have things
that appear to be more distant from E. coli. But the good news
with E. coli, and I think we got kind of
dumb lucky with this in terms of it being
this model organism, it is pretty promiscuous. So its transcriptional and
translational machinery, the ability of DNA into
RNA, RNA into protein, is sufficiently promiscuous that
a lot more genes come through than if you were to, say,
use just by arbitrary nature some other microbe. – Thank you for your talk. Could you explain more
about the difference between drug cycling and the
collateral susceptibility? Because it seemed like,
from what I understood, the collateral susceptibility– one drug resistance open
susceptibility to another, so wouldn’t you just
be cycling drug B or C? – Yeah. Great question. I should have said that none of
those are mutually exclusive. So, in fact, people
have suggested that when you do
drugs cycling, you should use collaterally
sensitive drugs. It’s just that you could also
then use collateral sensitivity in combination. So you can combine any one
of those different methods to be able to come up
with novel ways in which to repurpose drugs. So drug cycling is just
a very general concept of using multiple drugs one
after the other like crop rotations. Whether use collaterally
sensitive version or not, it might still work. If you happen to
use collaterally sensitive versions, it might
work a little bit better. But you could also use
collateral sensitivity without any cycling, basically,
by using it in combinations. That’s really the clarification. – Thank you. It was a great talk. I had a question, however,
about the inhibitors of the mechanism that
inactivate the tetracyclines. Because, by analogy, with the
beta lacatamase inhibitors, all you do is you select for
mutations in the resistance gene. So presumably, that
will also happen. – Yeah. That’s a great question. And probably the most honest
answer to your question is, eventually, the bacteria
will win with any single drug, right? They are just evolutionarily
smarter than we are. So how do we try to extend the
lifetime of a new compound? One thing we’ve done, and I
didn’t show the data here, is we focus on our
structural approach. So we’re solving the structures
of all of those enzymes with both substrates
and inhibitors bound, and trying to rationalize why
the inhibitors are working. And by rationalizing
that, then we can use more sophisticated
structural methods to say, hey, maybe we can lock these
particular positions in these inhibitors through
synthetic design, such that it’s less easy for those
particular enzymes to mutate. So, in a way, and
this is something that’s a great part
of biology, anytime you think you’ve found something
good, you try to break it. And once you break
it, you understand what the risk of the later
on medication might be. So we’re doing two things. We’re adaptably
evolving those enzymes to be even better not because
we want to release them into the environment,
but because we want to know what
the risk might be, trying in an amino-acid-specific
way in the structure to understand how
mutable they might be, and then designing
our inhibitors through chemical synthesis
to allow that not to happen. Again, that’s never going
to be the end of the story, but it’ll give us a longer
therapeutic window, we think. – Thanks very much. I’m just rather sad that my
era as a working scientist happened before all of
these modern techniques were available. But anyway, it’s great
that you’re using them. – Thank you. – OK. Thank you so much
for the presentation. I really appreciate it. My concern during
the oral presentation was that the antibiotics,
while fighting diseases, they actually fight
even our bodies. And we cannot even go
back to the evolution. We’re just moving forward. And my concerns are two, though. The first concern is
that with evolution– I mean for bacteria
and all the diseases, evolution is the
same for a human. I mean, not practically or
specifically, but I mean, it exists. And when antibiotics
go to our body, they fight our body as well
as they fight the bacteria. And I’m concerned that– do you believe that we should
[INAUDIBLE] antibiotics, or we should try a way
to find a [INAUDIBLE] that something more biotic
into our body that instead of fighting our body,
support our body and together with the
body to fight the disease? – I think that’s
a fantastic point. We’ve been kind of stuck
in this warfare mentality with microbes because
that worked back when the antibiotics were found. And they worked
really well, we’re happy that they
worked when they did, but they’re going to
have collateral impacts. Now, it’s true that there
are some toxicities involved with antibiotics, but
for the most part, the doses that we use
a lot of antibiotics are relatively safe. But we still have
this issue that we are taking these
chemicals that are going to have all of these
collateral impacts at least on the
other good bacteria. So the neat part of
microbiome science, all of this focus on microbial
ecosystems and communities, is beginning to let us learn
from that in terms of saying, how is it that microbes
naturally interact? They interact in communities. They work with each other. Sometimes benignly,
sometimes in conflict, but it’s still at
a community level. So can we learn from
those principles to come up with new therapeutics
of the future that are not purely these offense molecules? And there’s amazing work being
done right now in basic science and also in industry to
do things like design– either design a probiotic,
specific microbes that can be engineered to
deliver particular products or to modulate other
microbes, and also other types of chemicals that
are not offense molecules, but they might allow
one particular microbe to grow a little bit better
than another type of microbe so that potentiators
or maybe good bugs, they can crowd out bad bugs. So I think you’re
absolutely right. The therapeutics of the
future will not simply be offense molecules
of the types we’ve had. They will include
them to some degree because no matter what I say
about collateral damage and all of that stuff, if
you’re infected with the plague
bacterium, you shouldn’t worry about wringing your
hands about whether you should take antibiotics or not. Your [INAUDIBLE] is
not going to respond to some touchy-feely
FMT for instance, right? You’re going to have
to kill that bug. But for the future, for
other types of infections that require a little
bit more moderation, I think there are
incredible new avenues. And I think, for instance, one
of the really cool features of the microbiome
that I’m really attracted to as a
future therapeutic is its direct interface with
the human immune system. So we can modulate microbes
even at a community perspective, be able to very carefully
tickle the immune system, to then trigger it at
the appropriate time to be neither underreactive
nor overreactive, and go after specific pathogens. May be an incredible therapeutic
avenue for the future. And I didn’t have time
to talk about this, but we have a big arm in
our lab of our engineers who are working on
probiotic engineering, both bacterial probiotics as
well as fungal probiotics, to do exactly the type of
things I just talked about– deliver enzymes
that are specific for metabolic disorders,
but also to, for instance, display very specific
tags on their outside, on their surfaces,
specific to bad bugs to be able to take them
out of the ecosystem. – Now, my only
concern was that we were so focused on
destroying the problem and we were not talking
much about safety, and that was my concern. So thank you so much. – Hi. Regarding what you said
about the use of antibiotics in agriculture,
obviously, that’s in large part due to
the fact that it just makes process of animal
husbandry much easier and more efficient. However, there is a
lot of evidence showing that there is a lot of
deleterious effects emerging because we are using
antibiotics for that, and it’s hurting our ability
to fight disease in humans. So do you see any
potential ways that we could reduce the
use of antibiotics in a way that makes
sense economically for these companies, or do
you think legislation is going to be the major mechanism? – No, I think that’s a
brilliant question also because the agricultural
industry in the US is an exceptionally important
part of our economy. And I think just
wantonly banning the use antimicrobials
is not going to work. In fact, we’ve seen
experiments in this ilk in Europe, for
instance, where we’ll ban the so-called nontherapeutic
use of antibiotics. You know, wash your hands of it
and think everything is good. And then what’ll
happen is the industry is incentivized to change
the concentrations of what’s called a therapeutic use. Then the drug use goes back up. So I think working
with the industry and coming up with
solutions to mitigate and come up with
alternatives, actually, I think the best
alternatives right now are back to microbiome science. Why are these
antimicrobials working to make these animals
that much better in terms of their productivity? It probably has something
to do, eventually, with something on the host or
something on the microbiome. So there’s some really
exciting work already going on, certainly, in academia, but
also in a few small biotech companies, where they’re trying
to understand whether you can deliver probiotics or particular
features in terms of, again, compounds that are not
antimicrobial, but might give boost to particular microbes
that allow the animal to be a little bit more productive. And now you’ve given
your young farmers an option that
allows them to still have meat that costs less,
and so it’s more productive, but then saves the
antimicrobials for us. So I’m not saying that’s a
panacea in terms of a solution, but it is one way
in which we can use all of this incredible
work that’s going on in microbial ecosystems,
understand the mechanism of why the antibiotics work
in the first place, and then replace them with
the microbial solution. [MUSIC PLAYING]

Author:

Leave a Reply

Your email address will not be published. Required fields are marked *