Cystic Fibrosis: From Gene Discovery to Basic Biology to Precision Medicines

I think the festive
atmosphere is telling. This is truly a– it’s truly an afternoon
of celebration. It’s going to be– it’s going
to be a great afternoon, because we’re celebrating– we’re celebrating the promise
of science and medicine, a series of discoveries made by
five relentlessly passionate, brilliant individuals,
who have together changed the course
of cystic fibrosis, a rare and devastating disease. Their discoveries span
the fields of genetics, physiology, pulmonologist,
biochemistry, pharmacology, and have culminated in
life-altering precision targeted treatments for CF. The hallmark of
cystic fibrosis– as I’m sure we all know– is a genetic mutation that
impairs the cell’s ability to transport chloride. This malfunction leads
to the accumulation of very sticky mucus in
the lungs and other organs and causes a
constellation of symptoms, including recurrent
lung infections, progressive scarring, and
loss of lung function, as well as pancreatic, liver,
and other gastrointestinal problems. It’s a true
multi-organ condition. Cystic fibrosis has sadly cut
short many, many young lives, and has confounded physicians
and scientists for ages. In the 1960s, a baby
born with cystic fibrosis was unlikely to live
past the age of 10. But thanks to
advances in diagnosis, care, remarkable
advances in treatment, the median age of survival
has increased steadily, and is now estimated to be
in the 40s, and we see more. Until recently, the treatment
approach to cystic fibrosis remained largely symptomatic,
palliative, but no more. Collectively, the
discoveries made by the five pioneers
we’re honoring today have resulted in small molecule
therapies that can actually correct the underlying
malfunction that is responsible for the range of
pathologies that we see in CF. Thanks to these discoveries,
an infant born today can receive treatment
that restores healthy function in
the faulty protein responsible for the devastating
symptoms and natural course of this disease. So new therapies we believe
will add healthy years to the lives of those born
and those living with CF, and continuing to emerge from
the remarkable work of those whom we honor today, Francis
Collins, Paul Negulescu, Bonnie Ramsey, Lap-Chee Tsui,
and Michael Welsh. Unfortunately, Professor
Tsui is unable to join us. Their achievements
represent one of the most elegant and instructive
protein to person trajectories in the history of biomedicine. Lap-Chee Tsui and Francis
Collins and their teams conducted the
foundational work that led to the discovery of the
cystic fibrosis gene, CFTR. Their efforts elucidated the
gene’s molecular structure and pinpointed its location,
providing an entry point into understanding the
genetic defects that drive CF. Michael Welsh and his
laboratory observed that the cells lining the
organs of patients with CF lack the ability to
transport chloride. Dr. Welsh and his team showed
that correcting the protein product of CFTR could restore
defective chloride transport. And this proof of
principle demonstration linked CFTR mutations to
actual protein malfunctions and showed how they fueled
disease development. Now building on these key
fundamental discoveries, a team led by Paul Negulescu– originally working in a
small biotech company Aurora, and later at Vertex– identified compounds that
modulate the function of the CFTR protein. Their work led to the
development of the only CF treatments that
are available today which actually correct the
underlying protein defect and restore cell’s ability
to transport chloride. Pediatric pulmonologist Bonnie
Ramsey played a critical role in the translation of these
therapies into the clinic. She was the architect of
the clinical trial network and the seminal studies
that led to the approval of the first small
molecule treatment for CF, as well as later
combination treatments. And Dr. Ramsey I am
proud to say is an alumni of Harvard Medical School. The trailblazing
efforts of our honorees were propelled by the
Cystic Fibrosis Foundation, whom we also honor today. The CF foundation partnered
with private industry to fund early stage research,
and thus catalyze translation of basic biological insights
into the therapies that are changing patients’ lives. The foundation’s brave,
mold-breaking move marked the birth
of something that is now referred to as venture
philanthropy, a model that has been emulated by many
other disease organizations. We also have to thank
Joe and Kathy O’Donnell. After losing their 12-year-old
son Joey to CF in 1986, the O’Donnells transformed
sorrow into action by launching the Joey Fund. Partnering with the Cystic
Fibrosis Foundation, they’ve raised critical
early stage support to the tune of $175 million. And that kickstarted
the work that culminated in the
development of ivacaftor, the first c-f treatment
that corrects CFTR function. Now together, the efforts
of those we celebrate represent a triumph
of collaboration, one that combine the
acumen of academia with the audacity of a
young biotech company, and the drive of a
nonprofit organization. Their story is remarkable. And it serves as an
opportunity for all of us to reflect and to learn
so that we can replicate that success again and again. We live in an era that is
marked by a dizzying pace of new scientific
discoveries, by the emergence and proliferation of
unprecedented new technologies. It took the [INAUDIBLE] and
Collin’s Labs the better part of a decade, I believe– we’ll learn more about that– to locate the CF gene. And what would it take us today? Hours? Days? Maybe weeks? But clearly, the
pace of discovery has accelerated dramatically. We now have structural
tools, such as cryoelectron microscopy, that
can dramatically accelerate our ability to make
the kinds of insights that translate fundamental
insights into structure, into function, and the
promise of new therapies. Chemoinformatics, computational
approaches, computer-based drug screening, and so much
more, all of these tools are transforming the pace and
impact of drug development. And they’re helping to
make the painstaking clinical development
carried out by Doctor Ramsey and other colleagues more
efficient and better targeted. We are now able to interrogate
the biology of disease and health in ways
that really were unimaginable even a decade ago. And imagine, just since
this award was announced a few months ago, there have
been already major new insights that offer prospects for
new and even more targeted therapies for cystic fibrosis. This past August,
two research teams– which were co-lead by scientists
from Harvard Medical School– announced their discovery
of a new cell type in the human airway. The cell is called
pulmonary ionocytes. They appear to be the
nidus and epicenter for activity of the CFTR gene. And insights like this
will set the stage for a new wave of therapies. And we’re going to be hearing
about some of this work from Jay a little later. So I think we all
come together today with a sense of excitement,
a sense of satisfaction, a sense of having
delivered on promise, and a great sense of more
to come in the future. Because we still have to
learn how to better innovate at each point of the
therapeutic development path, because we still
need to figure out how to develop treatments
better, faster, and ultimately, cheaper. And we must learn how to
work better with each other, because we know that we can
achieve tremendous things when we reach across the
barriers that might often otherwise divide us. So let’s remember that the
importance of a discovery is measured not by the number
of citations that a paper earns, but by the ability to
translate that discovery into a new treatment
that actually counts in the number of lives saved. That’s the number that we as
a community want to emphasize. That is our contract
with society as biomedical scientists. That is our reason for being. Now few organizations have
understood that better than the Warren
Alpert Foundation. And its mission has
been to recognize scientists whose work
has altered and reshaped our understanding of and
ability to treat disease. This symposium is made
possible by the vision of the founder of the
Warren Alpert Foundation, and by this wonderful
organization that supports us. Their history–
I’ve told it before. I enjoy telling it every year. It’s the stuff of legend. So it turns out in
1987, Warren Alpert came across a news
article that described the work of Sir
Kenneth Murray, who is the British scientist
credited with developing the hepatitis B vaccine. And right there and then,
he picked up the phone and he cold called
Murray and told him that he had won the Warren
Alpert Foundation prize. [LAUGHTER] A jury of one. The only problem was that
the foundation didn’t exist. So I guess what
happened was Warren approached Dan Tosteson– who was then the head– the
dean of Harvard Medical School– asking him to convene
a panel of experts that would legitimately choose
the future award winners. Subject, of course, to the
ascent of the foundation board. So here we are with
30 years later. And the Warren Alpert
Foundation prize has been awarded to 64
scientists, 10 of whom have gone on to win
the Nobel Prize. Two of those scientists– Jim Allison and Tasuku Honja– who were recognized with last
year’s Warren Alpert prize you will know this past
Monday were awarded the Nobel Prize for Physiology or Medicine
for their foundational studies in cancer Immunology. So I’m thrilled that we
will have today visiting with us members of the
Warren Alpert Foundation’s board of directors. On behalf of the
scientific community, I’d like to give a
special thank you to the foundation for
its support of science and discovery. I want to especially
congratulate the recipients of this
year’s Warren Alpert prize, to express my deep
admiration for their remarkable
achievements which have brought hope to families
and patients around the world. And my thanks to all of you here
today who have come together collectively to help
us all celebrate these momentous achievements. Now I would like to
invite to the podium our symposium moderator,
Dr. Vamsi Mootha. Vamsi is an alum of
Harvard Medical School. He completed his
clinical training at the Brigham and
Women’s Hospital. And he’s currently a
professor of systems biology at the Massachusetts General
Hospital and the Harvard Medical School. And he’s also a member
of the Broad Institute and an investigator of
the Howard Hughes Medical Institute. Vamsi’s lab combines
a constellation of disciplines, computational
biology, genetics, physiology, to understand the molecular
mechanisms of the mitochondria and their role in
disease to define the origins of
mitochondrial dysfunction that fuels a number of
devastating diseases, and in the hopes of discovering
new therapies for them. Dr. Mootha is one of the
shining lights of our community. And it’s great honor to welcome
him to moderate this symposium. Vamsi. [APPLAUSE] Thank you, George. It’s a real honor to be
moderating today’s session. I want to just begin
by welcoming everyone to the celebration. I’ve been at Harvard
Medical School for 25 years, and I have to say, this is
often the very, very highlight of each year as we get
together to celebrate the very best of biomedical
research across the world. As Dean Daley indicated, this
foundation– the Warren Alpert Foundation and
Harvard Medical School have collaborated for about 30
years or so to honor scientists whose breakthroughs have helped
us to understand and to cure human diseases. And it’s very
exciting every year, and then this year,
it’s awesome that we’re celebrating fundamental and
translational discoveries in the field of cystic fibrosis. Now this year’s award is
special in a lot of ways. It’s really celebrating
research that took place over a 30 year period. And as George said, it’s
going to be recognizing the discovery of the
root genetic cause of cystic fibrosis. It’s going to be
honoring discoveries that helped us to understand how
those root genetic causes lead to defects at a molecular level. It celebrates the development
of small molecule therapeutics that for the first
time, actually helped to correct those
molecular defects. And finally, it celebrates
innovative clinical trials that help to get these genotype
specific medicines approved. This entire arc is
only possible because of the creativity,
the persistence of the investigators. And one of the most
important themes that I think we’re going to be
celebrating today is teamwork. All of the individual
recipients today are amazing in their own right,
but they represent huge teams, national, international
students, professors, academics, industry
scientists, and it’s only because of all of these
collective efforts that we’re here today. This arc of precision medicine
specifically as applied to cystic fibrosis
is an exciting one. And it’s been so exciting
that even the former president of the United States,
Barack Obama, he referenced in his 2015
State of the Union how exciting this is when
he introduced the Precision Medicine Initiative. In 2015 he said, and I quote,
“in some patients with cystic fibrosis, this approach”–
referring to precision medicine– “has reversed
a disease once thought unstoppable.” So today, we’re going to get
to celebrate this 30 year path. And we’re going
to hear firsthand from the investigators that made
so many of these discoveries possible. We have a real
star-studded lineup, and it’s going to
include four out of the five prize recipients. And book ending these
four prize recipients, we’re also honored to have a
leader in academic medicine that’s going to help to set the
stage for these discoveries. And at the back end,
we’re going to hear from one of the rising stars
in the Harvard Medical School community that’s going
to share with us some of the future of lung biology
and lung disease biology research. So to kick things off, I’m going
to introduce the first speaker. We’re very fortunate to have
Dr. Pamela Davis with us today. Dr. Davis is currently dean
of medicine at Case Western Reserve University, where
she also serves a senior vice president for medical affairs. Dr. Davis is a graduate
of Smith College. She received her M.D. And PhD
degrees from Duke University. And she then pursued her
clinical and research training in cystic fibrosis at the
National Institutes of Health. She’s a seasoned basic and
translational researcher specifically in the
field of cystic fibrosis, and formerly worked as the
head of pediatric pulmonology at Rainbow Babies Hospital
where for years, she focused on the care and
translational research for CF. She’s received a
number of awards, including election to the
National Academy of Medicine. And today, she’s going to help
to provide a historical context for understanding the
great discoveries that are being awarded this year. [APPLAUSE] Well, it’s a true
honor to be here to celebrate the success in
the field of cystic fibrosis. When I chose to work in the
cystic fibrosis lab at the NIH lo these many years
ago, I decided to do that for all
the wrong reasons. I thought this disease would
be solved in my lifetime, and I wanted to be
close to something that was going to be put
into the history books. And I think I was right
for the wrong reasons. I want to tell you today not
only about the perspective on solving the problems
of cystic fibrosis, but also the implications of
this process for what we do now for the future of
genetic diseases. And the term genetic
diseases is now broadened from the autosomal
recessive or the autosomal dominate to a large
number of diseases where the genetic errors
profoundly influence the risk. So let me tell you a
little bit about all this. I am going to talk a
little bit about what the specific
implications of CF are for the classical
genetic disorders. It’s been estimated
that there are about 10,000 genetic
disorders, but most of them affect fewer than one in
1,000,000 individuals. Cystic fibrosis on the other
hand, in the United States affects about one in
3,500 white live births. And that calculates up
to about 30,000 patients in the United States. That’s a fortunate
number, because it’s enough to do studies with. But it’s also a fortunate
number because there are 60,000 parents out there who
really care about this disease. And there’s the multiplier of
the sibs and the grandparents who came together, all together
to try to solve the problem. As Dean Daley told you, we’re
doing much better with CF. We’ll talk about
the arc of this. But as of 2016, the
predicted median survival– that is for a child
born in 2016– was approaching 50 years. And the mean age of death– median age of death in
2016 was almost 30 years. And I’ll show you how
amazing that really is. Just to briefly tally
a review for you, the cystic fibrosis
gene defect turns out to be an impairment
in chloride transport across epithelial surfaces. This is most clinically
challenging in the lung where the dehydration of mucus
and other metabolic problems lead to vulnerability
to infection, massive inflammatory response,
and progression of this problem until death. In the liver– the liver is the
second leading cause of death, way behind the lung– but similar obstructive
processes and cirrhosis can affect the patient. The digestive system is impaired
because the pancreas, because of plugging of the
pancreatic ducts, fail to secrete the
digestive enzymes. And indeed, initially,
cystic fibrosis was thought to be
pancreatic disorder, and it was thought to be a
disorder of malnutrition. And the gut itself is affected. An intestinal
obstruction is a problem. We recognized cystic
fibrosis for many years by its excessively salty sweat. And that turned out to be
important in recognizing the chloride transport defect. This disease was
actually first described by a pathologist in 1938, and
it was called cystic fibrosis of the pancreas. And it was thought that the
victims of cystic fibrosis survive to about
six months of age. In the early 1950s, the sweat
defect was discovered by Paul di Sant’Agnese following a
heatwave in New York City, where he noticed that many of
the children coming in with heat stroke had cystic fibrosis,
a disproportionate number of them. And that led to the discovery
that the salt content in the sweat of patients
with cystic fibrosis was about five times that
of healthy individuals. But it was almost
75 years to 2012 before we had a
drug that actually addressed the basic defect
for cystic fibrosis. So how did we get there? And how should we be
thinking about this for the next disease, and
the next, and the next? So I think that one of the
key things that happened was that we collected patients
at centers so the clinicians got experience with them. And there was some inspiration
of the basic science researchers at that
center to do something about those little kids who
were suffering in the hospital. I think collection and
formation of the centers network was absolutely critical. I think another critical
event, which occurred actually in 1955, was the establishment
of the Cystic Fibrosis Foundation. This has turned out to be
one of the massive success stories among voluntary
health agencies. And I hope I’ll show you as we
go through this the importance of this coordinating
factor and this guiding genius of the field. You can see on this graph that
the median survival age has increased markedly over time. In 1955, once the sweat
test came into use, we were able to recognize milder
cases than Dorothy Andersen was able to see in 1938. But still, the survival was
about five years of age. It was very distressing. But the foundation established
a strong collaborative network for care, basic research,
and clinical research to facilitate not only the basic
understanding of the disease, but also treatments
and clinical testing. Now that sounds all wonderful. I started working for the
Cystic Fibrosis Foundation– this is going to date me– in about 1975 when
I was a fellow. The budget of the Cystic
Fibrosis Foundation was about $6 million
a year back then. And back in those
days, the clinicians were concerned that
they needed more to take care of the patients
because they were suffering. And the researchers
were concerned that if they didn’t
get the support, the clinicians 20
years later were still going to be dealing
with the suffering. So not having
enough to work with, having not enough to
put into all of this was very distressing
all the way around. Shortly thereafter, Bob
Dressing, Doris Tulcin and Bob Beall came on the scene. Bob, as the president of the
Cystic Fibrosis Foundation, was an absolutely
incredible catalyst. He came from the NIH, so the NIH
can take a little bit of credit for this imagination and drive. But basically the concept was
we have to raise more money and we have to have a plan. So as soon as the dollars
hit around 40 million or so, all the competition between the
clinicians and the researchers disappeared. The clinicians didn’t have
everything they wanted, but they were able to take
care of their patients in a better fashion. And the researchers
saw a real opportunity to go after this
challenging disease. Now the Cystic
Fibrosis Foundation saw the need to catalyze
collaboration in various ways, and in the mid-1970s established
the research development program. Eventually that became a network
of centers, each of which focused on its own
particular area. They weren’t cookie
cutter centers, but they were able to
capitalize on the expertise of each institution
that could be brought to bear on different aspects
of the cystic fibrosis problem. And this greatly
enhanced the ability to conduct cystic
fibrosis research. In addition, the NIH stepped
up and created a network of P30 centers, which
are core centers, and therefore was able to
supplement and complement the activities of the CF Center
and the CF Research Development program and establish
its own broad portfolio. In about 1995, the CF
Foundation established the Therapeutics Development
Network, which I think was a critical feature
in the later success of cystic fibrosis. So originally the concept was to
have a network of centers which had comparable and uniform
capacity to conduct the kinds of clinical
research that are going to be necessary for cystic fibrosis. So there were some
kinds of procedures that you were going to
have to be able to do. There was a team that
had to be in place so you didn’t have to assemble it each
time there was a new project. You had a team who was used
to dealing with the patients. And the centers provided
the patient base, the clinically-based connection
to a patient population that was willing to participate
in patient-based research. In any given year at the
center at Rainbow Babies and Children’s Hospital, we
had about 50% of our patients in any given year participating
in a clinical study. That’s extraordinary. The NCI is struggling to
get 10% of cancer patients on clinical trials. But this population is
linked to its physicians and ready to do that. We wouldn’t have that if we
didn’t have the center network. So initially, the TDN
started with, I think, it was seven centers. And that nucleus
grew to be quite a skilled clinical research
group of investigators. Subsequently, there was another
ring of about 17 to 20 centers that had significant
capabilities, but not necessarily
all of the ones that were in the core group. And then the 127 centers
scattered around the country were primed for conducting
Phase 3 trials where the specifics of doing alternate
markers for improvement were not quite as important
as the measurement of clinical improvement. So this was really
a brilliant stroke and was absolutely critical. I’ve pictured up there a number
of the drugs that came out between actually,
probably, the early ’80s and the discovery of the
specific precision guided medicine. And you can see
that these drugs, with no specific reference
to the basic defect, actually improve the ability
of the patients to live, and live comfortably,
for many years. So this was a very
valuable proposition. It also primed the
network to work together and to do these
studies so that when the drugs came that might be
gene directed, they were ready. Very important. Find the. Gene now we had the
suspicion that it was, in fact, a single
gene that caused most cases of cystic
fibrosis, as opposed to some genetic disorders
where the phenotype is created by multiple different
kinds of genetic disorders. The clinical
observations on genetics presage the fact that it would
be, in fact, a single gene. But it wasn’t so easy
to find the gene. And Francis, I think,
is going to tell us a little bit about
what kind of a tour de force it required in the 1980s. There were many groups
that started this. I remember three of
them particularly who were in it, hammer and
tongs, almost to the end. One of them was in Great
Britain, one of them was in Utah, and the third
was a collaborative group between the
University of Michigan and the University of Toronto
and sick kids in Toronto. And they turned out to be
the folks who found it first. And Francis is going to tell you
all the ins and outs of that. But I have to say that from the
point of view of Case Western Reserve, I felt very fortunate
because Francis’s lab was a great training
environment, and I was able to recruit his
first graduate student, who was part of this study. And that is Mitch
Drumm, who has been a major success in cystic
fibrosis subsequently. So I have a special fondness
for Francis’s success in all of this. Now one of the critical
things about all this is that all these programs
needed clinical materials to work with in those days. And particularly they
needed special families. They needed families with
cousins and multiple siblings in the same family. That turns out not
to be as easy as you might think because
sometimes, when parents had a child with cystic
fibrosis, that was the conclusion of the family. So in order to find
those families, that network of centers,
and the willingness to find those families, convince
them to donate their DNA, and collaborate turned out
to be very, very important. It was also
necessary to identify the true underlying
functional abnormality. And I have to give some credit
in this to Paul Quinton, himself a victim of
cystic fibrosis, who chose to study a
difficult organ, but one that was uncomplicated by
infection or inflammation. And that was the sweat glands. So Paul DeSantis had
recognized that there was a vast difference in the
salt concentrations in sweat, a five-fold difference. So he had a big
signal to work with. And he went out to
try to figure out what was causing this difference. And he, in 1989, published
a paper which I thought was pretty definitive showing
that it was a chloride transport defect,
and it all came out of the humble sweat gland. I have to say that
Paul himself took much of his clinical
material from himself. And someday, if
you talk to Paul, ask him to take off
his shirt and show you how people biopsied his
back over and over and over. There were multiple candidates
for the basic defect until we focused down
on the ion transport. They extended from
the structure of mucus to the regulation of
secretion to sodium transport. But chloride transport was,
in fact, the key issue. In 1989, the Quinton
paper came out, and the CF gene was discovered. And mirabile dictu, the channel
that was predicted by the gene was lined by positive charges,
which made it an anion channel. And so the work fit together,
and that was greatly to be delighted at. That meant that we could
now say that in most cases, because the CF protein
was destroyed before it was able to reach its
intended site of action at the surface of
the cell, chloride failed to be excreted from the
cell, at least in the airway, and failed to be
resorbed in the sweat. It is a bi-directional channel. But the consequence of that
was that sodium was reabsorbed to an excessive extent,
and water as well, leaving the airway
surface liquid dehydrated and the mucus sticky
and entangling the cilia. So there we had our
problem outlined. How were we going to fix it? there are lots of
ingredients that go into finding a
potential therapeutic. And there are many reagents and
materials that were important. You needed cell lines
well-characterized. You needed probes. You needed primary cell
cultures because you didn’t want things to change
in the course of making a cell line. You needed animal models. You needed patient samples. And you needed
combined expertise. And you needed to be
able to share them. And I think that that
was absolutely key. The Cystic Fibrosis Foundation
was putting enough money into the academic
labs so that they could put a lot of
pressure on the labs to share, even when they got
a little bit huffy about that. And they were able
to get them to share not only with other academic
labs, but with the industry. It was a very interesting time. The Cystic Fibrosis Foundation
supported the production and characterization of
cell lines, the development of appropriate primary
cell culture techniques which would replicate
the key features that you wanted to correct. They supported the
development of animal models. Now the easiest animal model
back in those days to make, which wasn’t very easy like
it is now, was the mouse. Turns out the mouse
doesn’t have a lung a whole lot like
cystic fibrosis. But it was a CS mouse,
and it was what we had. But subsequently, as you
may hear from Mike Welsh, the ferret and the pig
models came forward. The ferret’s lovely because
it’s a very long trachea. Lovely. The pig is the
closest to the human. Think about that for a minute. We in the community
created these things, and we shared patient samples. And as this was going on, the
Cystic Fibrosis Foundation put money not only into
the academic community, but also into various
industrial partners. It was the concept
that we were going to invest in
anything that looked good to cure the disease. But they also had a few little
asterisks associated with it. They had scientific
advisory boards associated with almost
all of their investments. But the scientific
advisory board didn’t report, for the most
part, to the company the way most of them do. They reported to the
Cystic Fibrosis Foundation. And this did two things. Number one, they didn’t pay us
as much as a regular industry scientific advisory board. Saved a lot of money. And number two,
the advisory boards helped the Cystic
Fibrosis Foundation assess the progress of the various
programs in industry, and it also helped
them identify what reagents the industry might
need and get them to them. So it really, I think,
did accelerate everything. It was important to have
this kind of acceleration, and I have to credit the
foundation for most of that. There was ongoing meeting,
discussions, evaluation of data. The Williamsburg
conferences were legendary. They had about 100 people
at them, give or take. I still have the
polo shirts that were given out at a few of
them right around the turn of the 21st century. And they were no holds barred
conferences and the promise that you wouldn’t be
quoted on your data. So people were willing
to show early data. They were willing to show
data that they might not be able to replicate the
third or the fourth time. But they were willing
to stimulate people in the course of discovery. And they were great
catalysts for collaboration and they were great
catalysts for assessment of where the whole field went. So I think that that
was absolutely key in the entire process. I think it took a lot of people. Not everybody was at every
conference every year. But there were enough
people from each center so that the word
could be carried back. And we all learned something. And I think that this
exchange was really very, very important. It took a lot of people. I don’t know whether any
of you saw the analysis by Sandy Williams on
Scientific Discovery. They picked two
drugs, one of them a CLTA for inhibitor and cancer,
and the other one, Ivacaftor, the first drug released
for cystic fibrosis. And they did the
analytics on what it took. Now for cystic
fibrosis, they came out to looking over
a 59-year period, looking at what went
into that discovery, and analyzing the
papers that were cited and the branches from that. Almost 2,900 scientists
at 2,500 institutions contributed to this activity. We are honoring five of the
most prominent here today. And they are, in fact,
the prime movers. But you have to
remember that it really did take a village to get
to the point that we’re in. In fact, this really is
team science at its best. The analytics that were produced
by Sandy Williams and his team cited seven institutions
and 33 investigators who wound up accounting
for 49% in their analytics of the actual discovery, the
basis of the actual discovery. And it started a
long, long time ago. So remember that
everybody contributes in a variety of ways. So what can we learn
from the remarkable arc of cystic fibrosis 75 years
from discovery and report, roughly, of the
disease, 23 years from the discovery of the gene. It took a very long time. I think most of us
in the field thought that the discovery of the gene,
it would be downhill from there and we’d be done. It turned out not
to be that way. And it turned out to be
much more complicated. It turned out to be
much more difficult. It turned out to have
many more fits and starts. But we actually went
through it, we did it, and we need to think about
what it is that got us there. In the 1980s finding a disease
gene took many, many years. Nowadays it’s days,
weeks, months. If you have a toughie it’ll
take you a couple of months. you know it’s
really accelerated. Creating disease models,
including animal models, with the new CRISPR
technology is facile. You can make a
mouse very rapidly. It depends on how rapidly
the mouse will breed for you. But that is a boon
that is incredible. It took us a long time
to make the CF mouse, to validate the
CF mouse, and then to discover that maybe
the CF mouse wasn’t the most terrific model for it. Test thousands of drugs, not
to make an initial assumption as to what kind of
a drug was going to help, but to be able
to do an unbiased screen. Back in the 1980s
that wasn’t possible. Nowadays it is possible,
and you can test thousands to millions of drugs
in a period of months. Not only that, but
you can test a library of drugs that are already
safety tested in human beings. And many people will start
there first nowadays because we didn’t have that earlier. But now we’ve got libraries of
compounds that have actually been through clinical testing. And some of them
have been discarded. There are how man, 200
discarded drug hypotheses for Alzheimer’s? Maybe they work
for somebody else. So we got some of that. Bringing drugs to patients
still takes a lot of time. The clinical trials
are still extensive and the requirements of the
FDA are still substantial. Genetic correction, back in the
1980s, was really not possible. Now it’s being developed. Whether we correct with
actually correcting the DNA, the native DNA,
whether we correct by supplying a copy of the effect of the
good DNA, it’s being developed, and it’s going to
be there for us. So I would say
that we’ve learned a lot from cystic fibrosis. And you’re going to
hear the details of this from the honorees. But I would say that collecting
the population of patients and utilizing not for
profit health agencies to collaborate and to
coordinate the activities was really crucial for CF. To establish a strong
and collaborative network for both care and research was
fundamental to the discoveries that were made. Finding the gene and identifying
the true functional abnormality was really the threshold of
finding strategies to cure it. Create and share the
investigative materials. The sharing is really crucial. It does take a village. People prepare things. They need to be shared. And take advantage of the
different kinds of expertise that exist in the
biomedical community. Invest in the best
research and development wherever it may be found. Maybe it’s found in academia,
maybe it’s found in industry, maybe it’s found in the US,
maybe it’s somewhere else. But we have to go after
it if we really want to do the best by our patients. I think it’s key to
support collaborations. If we don’t work
together, we are not taking full advantage
of the opportunities of well-educated, imaginative,
creative individuals in the community. So we have to be open
to collaborations and we have to figure out better
and more facile ways to do it. I have always been a believer
that the bedside informs the bench and the bench
informs the bedside, and it’s a bi-directional
improvement. I think that the cystic
fibrosis world epitomizes that. We have gone back
and forth from bench to bedside a number of
times, and it’s really been a very strong
collaborative effort. And cystic fibrosis now stands
on the threshold of cure, maybe not for every single
patient right now today. But we stand on the
threshold of being able to call this disease cured. I had hoped that that would
happen in my lifetime. I think there’s a
real opportunity here to learn from
cystic fibrosis and to be able to take our
lessons into other diseases and accelerate the
progress for them. I think the future is here. It’s just not widely
distributed yet. Thank you very much. [APPLAUSE] Thank you, Dr. Davis, for those
very inspiring and instructive comments that really help to
contextualize this afternoon’s symposium. So our next speaker and
first award recipient probably requires
no introduction. He may be one of the
most famous figures in US biomedicine today. Dr. Francis Collins is currently
the director of the National Institutes of Health. He was first appointed
by Barack Obama in 2009, and then again in 2016 by
President Donald Trump. We were joking earlier
that he’s living proof that these two presidents
can occasionally agree on some things. Dr. Collins is a physician
scientist with PhD training in physical chemistry followed
by clinical and postdoctoral training in human genetics. He’s made numerous biomedical
research contributions, including directing the
human genome sequencing project, which has
often been analogized to the landing on the moon. In his role as NIH director,
he has been a tireless champion of both basic and
applied research, and he’s also been a champion
of precision medicine. He’s received a number of
awards, including the Gairdner Award, the National
Medal of Science, the Presidential
Medal of Freedom. In the 1980s, Dr.
Francis Collins helped to lead an
international effort to map the root genetic cause
of cystic fibrosis that led to the identification
of the CFTR gene, which we’ll hear about shortly. He’s a truly gifted scientist. We’re lucky to have
him as a leader. And he’s also a pretty
darn good musician, which we may hear later tonight. So today we’re
going to be hearing from Dr. Collins on what it
was like to discover a disease gene when it was actually hard. [APPLAUSE] Thank you, [INAUDIBLE]. Yes, it was it really hard. And I’m going to force you to go
through that experience with me so that nobody will think
we should have been so much faster about getting this
answer because in the 1980s, this whole idea of
finding a gene that’s responsible for a well-behaved
a disorder was almost impossible to imagine
being achievable. I’m deeply honored to be
amongst the five recipients of the Warren Alpert Prize. In case there are any government
ethics people listening, I am required, as a
government employee, to decline the cash component
of the prize and have done so. But I’m glad to
be here with you. And maybe I’ll get a
wooden plaque to take home. We’ll see how that turns out. But its price has to be
within a certain limit. So nothing fancy, OK, George? And I’m aware that, as Pam
so elegantly described, this has been a team effort
of the most remarkable sort. And while five of us are
having the chance here to be recognized by the
awarding of the prize, there is a very long
list of other highly deserving individuals who
could just as well have been on that list as well. And so I will try to mention a
couple of them as I go along. But I’ll be no doubt
inadequate in that effort, and hopefully some of
those who follow me will also be sure to mention
just what a remarkable team effort this has been
from the very beginning. And at the end I’ll
say a little bit about where we
might be going next, although I think
that will mostly be the work of some of those who
follow me here at this podium. So I am going to take
you back to the 1980s. Some might say that
was the medieval period of biomedical research. You’d see the
alchemist over there. Well, you see the other
alchemist on the right. That’s me looking at a perfectly
godawful autoradiogram. I don’t know what
that data is that I was trying to make sense of. But I doubt that it ever
ended up in a publication. Yeah, that’s kind of the status. I was in Michigan then,
working with a wonderful group of talented graduate
students and postdocs and trying to figure out
how one could identify the cause of conditions which
I encountered as a physician in my genetics clinic, but
which, for the most part, had no molecular explanation. And that was certainly true
of cystic fibrosis, a disorder that I first learned
about as an intern, and then when I got to the point
of having my own research lab, really wanted to see what
could be done to sort it out. But there was a pretty
tough challenge here. As you heard from
Pam, it was 1938 when cystic fibrosis was
recognized as a disease, and that it’s recessive. And there seemed to be
pretty good arguments that there was probably
a single gene involved. But it had to be somewhere. We didn’t even quite
know at that point how big the genome was. By the time we got to
the 1980s, at least we knew the size of the target,
that the genome was somewhere in the neighborhood of
three billion base pairs. And we knew we were looking
for something that might be a single letter gone awry. So how the heck
would you do that? Well, some genes
had been getting discovered at that point
for hereditary disorders. But they went at it
in a different way, and they went at it in a more
straightforward way, where you had the disease and then
you figured out a functional abnormality, and then you
worked from that to what gene must be involved
to explain that. If you were trying to
understand hemophilia, well, you figured out
that there was a protein called Factor 8, or sometimes
Factor 9, that wasn’t working. And then you could
backtrack from the sequence of the protein, with some
artistic efforts involved, to figure out what the
gene must have looked like, and ultimately discover
Factor 8 and Factor 9 genes, and then you’d have
the story done. Wouldn’t you like to do
that for cystic fibrosis? Well, you would like to, but
at least in the early 1980s , before Paul Quinton came along
and made more sense out of all this, there was a very strong
amount of debate about exactly what was the fundamental protein
defect in this condition. And frankly, even
if we had known then it was a chloride
channel, we would not have come to the right
gene because it was not one of the ones that we knew about. We didn’t know much
about the genome, and we didn’t know all
that much about function. So you put the two together,
you have a hard problem. A very significant
development came about in 1980, which was
this paper in the American Journal of Human Genetics
by Botstein and colleagues, who proposed a strategy of
how one might actually succeed at this effort of finding
disease genes using the same approach
that had worked for other organisms like
yeast and drosophila, but which might actually
be applied to humans. Th diagram is a
little hard to follow. I just put it in there to
show you that in the old days graphics were even worse
than they are now in time to explaining concepts. But the idea was you need
variable parts of the genome. We’re all different. We’re 99.9% the
same, but actually that 0.1% makes for an awful lot
of variations in our genomes. If you could identify
places where people have those common variants and
develop probes to follow them through a family,
you would then expect that if you’re looking at a
disease like cystic fibrosis, and you have families with
more than one affected sibling, they should end up
sharing both the maternal and the paternal copy
of anything that’s close to the gene. Otherwise they wouldn’t
both be infected. That was the idea
about linkage approach in a recessive disease. You could do it in a
dominant condition as well. And this kind of got
people interested, although I think a
lot of people thought, this is never going to work. It’s much too hard. For the students who
are here, how did you score those
particular probes? You had to do southern blots. And if you ever get asked by
your principal investigator do a southern
blot, please object and say there are better
ways to do almost everything than that very demanding
effort, which I think caused more people to
leave graduate school than any other
technology I know of. It was possible to
get them to work, but oh, man, what a
headache that was. I think we all got
excited, though, because in 1983, just three
years after this paper, Jim Gusella stood up
at a genetics meeting and said he’d mapped
the Huntington’s disease gene to the short
arm of chromosome 4. This actually was
possible to work. Well, OK, if it could work
for Huntington’s, couldn’t it work for cystic fibrosis and
a list of other conditions where people had
already done hard work of collecting the family DNA
and the clinical information? And again, the cystic
fibrosis and those centers, having prepared for
this, was critical for CF being able to get on the onramp
for this kind of strategy. And so the strategy
then became you couldn’t do the function part. Let’s see can you
map the disease gene using this linkage strategy. OK. The problem is you’ll get
in the general vicinity, but what you really
want to do is to get all the way to the gene. Can you do that part? Well, in 1985, Lap-Chee Tsui’s
group, and independently, Ray White’s group, were able
to use that linkage strategy to say the cystic
fibrosis gene is on the long arm of chromosome 7. And that was a big deal. That was narrowing things down. But still, that’s
a lot of territory and a lot of very
unmapped territory because most of the
genome was unmapped. This was still, in
1985, a problem that seemed almost impossible
to be able to tackle in a reasonable
lifetime, certainly the lifetime of a
graduate student. I was one of those
who felt like we have to press ahead this way. And my research group
certainly agreed and rolled up their sleeves to tackle this. I found I had to explain,
a lot of the time, to people who were not tracking
this why it was so hard. And so I would make analogies. And I even would get my
daughter to draw pictures, as you see here, of a haystack. And I figured out,
because I was in Michigan, how much a haystack weighs. It weighs about eight tons. And I got a needle and put
it on the Mettler balance, and it weighed a couple
thousand milligrams. And so the ratio, I figured,
of needle to haystack was kind of similar to the
ratio of, say, the hemoglobin gene to the whole genome,
which is my explanation for why I haven’t found the answer yet. That was supposed to be. And I even decorated
that with a visual. And those are real
chickens, yeah. And you can see how difficult
it was to justify your existence when this was really hard. But we pressed ahead and
built a strategy here to try to accomplish this. And the strategy
involved something called chromosome jumping. At least that was my
approach because if you knew you were in the
right neighborhood, but the neighborhood was
still millions of base pairs, and you wanted to
cover that territory, the way that was
done in the 1980s was something called
chromosome walking. You’d get a cloned
fragment of DNA. You’d get the end of it. You’d go back into
your library, find another fragment that might
step you along a little further, recycle that. We figured to cover the region
where the cystic fibrosis gene was known to be would take
about 10 years in the strategy that was available. But think about this. If you could jump, and
then each place you landed, you could expand
from that, you could have multiple simultaneous
expansions going on, that would be great. Well, how do you do that? And this was a strategy
which I started working on as a postdoc with Sharon
Weissman, my remarkably brilliant postdoc
advisor at Yale. And the idea was to basically,
as you can see in the diagram, take very large,
long DNA fragments and circularize them, but insert
a recoverable selectable marker at the point of the junction. And those junctions,
then, are interesting, because if the fragment had
been 100 kilobases long, by the time you’ve made the
circle and marked that circle, you could use that as a jump. And the idea was if you made
a whole library of that sort, you could travel along the
chromosome more quickly. That’s what chromosome
jumping was all about. Believe me, you don’t want
to do that at this point. There’s much better ways, with
your mouse on your computer, to accomplish that. But at this point, it was
actually somewhat useful. So this was, in fact, then
applied after the mapping of the gene to chromosome 7Q. And we, by 1987, had started
to make some progress here, publishing a paper in
Science, which I was proud of and which led to that first
phone call from Bob Beall. Many phone calls followed
after that saying, oh, you’re working
on this disease. We need to talk. And again, already I think
I’ve said how critical it was that we’ve had
such a remarkable team effort in this space. My own team has been a wonderful
experience in this gene hunting search, as in many others. Already mentioned once
but I’ll mention it again, a absolute rock in my
group, at that point, my first graduate
student, Mitch Drumm, who is noted here working
at his desk at that point. And Mitch is here
with his wife, Mary. And I’m happy to say
that Mitch is now a full professor and an endowed
chair at Case Western Reserve. So one of the most joyful things
we get to do as scientists is to have the chance to mentor
other students and postdocs, and then see them
succeed and continue to contribute in
incredible ways to progress in an important disorder. And Mitch’s whole career has
focused on cystic fibrosis. OK. So we were pushing away here. We knew the gene was
somewhere in this two megabase interval between the oncogene
MET and a marker called J311. Other genetic disorders
like muscular dystrophy were beginning to have
successes because they had cytogenetic abnormalities
in rare patients that acted like neon signs
to tell you where to look. If you have a translocation
that actually interrupted the causative gene, you
can use the position of the translocation to
know what to look at. None of those turned up, nor
do I think they ever have, for cystic fibrosis. So the problem was really
exquisitely difficult. No deletions, no
cytogenetic abnormalities. What to do? Time for a new approach. This was mentioned by Pam. There were three groups,
four at that time– mine, Lap-Chee Tsui’s
group in Toronto, Ray White’s group in Utah,
Williamson’s group in the UK. And it became clear to me that
this was a really hard problem and that we all wanted
to see it succeed, and that maybe it
was a good time to think about collaborating
instead of competing. And I found a willing partner
in that argument sitting out in the sun at the
American Society for Human Genetics meeting
in San Diego in 1987. And that was Lap-Chee Tsui. Lap-Chee’s lab, highly motivated
and organized in the way that everybody would want to
see to go after this gene, but using the chromosome
walking strategy, where the chromosome
jumping that we were doing could clearly be
nicely synergistic. And of course, we
weren’t that far apart. It is a bit of a drive
from Ann Arbor to Toronto. And believe me, that
road got used a lot because we decided
that day, yeah, let’s just merge our labs together. Let’s not worry about who
ultimately gets the credit. Let’s just try to figure out
if we can find an answer here. And I went to Toronto, and
Lap-Chee came to Ann Arbor, and by the way, brought with him
a bottle of Canadian whiskey, which we agreed would
only be opened at the time that some success
had been achieved in terms of finding the gene. So it sat on the
shelf for awhile. I also had the great
privilege of working with the people
in Lap-Chee’s lab who were so devoted to this. And I have to mention Johanna
Rommens and Batsheva Kerem, and also Jack Reardon, who is
a researcher in his own right, very independent, and
incredibly effective researcher now
working in Toronto with Lap-Chee, who
became a critical part of the ultimate success of this
because of what he had been doing, particularly
with sweat glands and with making cDNA libraries. I’ll come to that shortly. Meanwhile, it was interesting. About this time my grant
application to the NIH to continue working on the
search for the cystic fibrosis gene got the worst score that
I’ve ever gotten in my life. Clearly unfundable
because there was a rumor that the answer
was already in hand and that one of the other
groups had found the gene. That rumor turned
out not to be true. And fortunately, the
advisory council for NIDDK rescued my grant, or
I don’t know what we would have done at that point. Fortunately, we also
had, at that point, the strong support of Bob
Beall and Bob Dressing and the Cystic Fibrosis
Foundation, who came alongside in
ways that helped us when we were stuck
at a particular place and made it possible to go
faster than we otherwise could have. There was a
consequence for that. Their support tended
to be attached to certain expectations
of progress. And I would get
notes like this one from Bob Beall of
February 20, 1989. “Dear Francis, my family
is now in a desperate way. There has not been any
income since June 1. My children are in the
streets and my wife is the sole breadwinner. I stay at the CF Foundation
awaiting your call every day that the CF gene is in hand. I hear tidbits now and
then about transfecting experiments and stories about
cDNA libraries and new jumps. But the truth is
there’s still no gene. The situation is
getting more desperate. Now Bob Dressing is thinking
about putting me out on the street,
taking away my office because I’ve not been
able to deliver the gene. And we both are in need
of some hope and optimism. Needless to say, my
desperate situation should not impart
any pressure on you. Remember, no pressure.” Boy, did I hear
that a few times. It’s great that Bob
is here with us. “Thank you for everything. But let’s hurry up a little.” Yeah, it went on
like this a lot. Well, we were hurrying up. We were hurrying up a lot. But you can imagine
what this is like. You’re traveling through
unknown territory. What are you looking for? You’re looking
for a piece of DNA that looks like it’s
evolutionarily conserved, for starters, because
that probably means it might be an axon of some sort. And then once you
have one of those, you try screening a cDNA library
to see if you get anything. And believe me, we
encountered every kind of artifact known to man and
chased after many of those. And here’s the diagram of
what that jumping and walking looked like. That’s a small
part of the effort. This is about a half a megabase
of a two-megabase enterprise. You can see those
arcs are the jumps. And then the horizontal
lines that are also there are the walks that were
accomplished at the same time. But again, you could do
this sort of simultaneously from multiple start points. And if you see in the
middle line there, there’s a Roman
numeral I followed by a bunch of other
bold vertical lines with Roman numerals. That was the moment when
things really got exciting because Roman
numeral I turned out to be something
that was conserved and that did hit a cDNA clone
in a sweat gland library from Jack Reardon. And then once you have
the cDNA start point, you can begin to
build what turned out to be a gene of 24
axons, and then you can begin immediately to map out
that genomic region even more quickly. And you can start, because
Jack had libraries, also, from cystic fibrosis
sweat glands, to look for something
that’s different. And it did emerge, along
about April of 1989, that there was something
really intriguing. And this is what it looked like. Again, those of you
who have not done DNA sequencing 30
years ago, you might wonder what you’re looking at. This is radioactively labeled
DNA sequencing ladders, the normal on the left of
the cystic fibrosis, cDNA on the right. And if you look
carefully at the pattern, you’ll figure out that
there’s a CTT that’s missing. And this is in the middle
of axon 10 of that gene that was never previously described. So that’s a three
base pair deletion. I got to say when we
first saw that, I thought, oh, boy, I wish it had
been a frame shift. Then I would have been a lot
more sure about the function. But now I’m glad it
was not a frame shift because that meant
the protein is there. And a lot of the
success you’re going to hear about in terms
of drug development would have been very
different if that had been a two or
a four base pair deletion instead of a three. When you go and look, then, at
the amino acid sequence, what turns out to have
happened here is you have deleted a phenylalanine
codon at position 508, the cystic fibrosis protein. Hence Delta F508, the
most common mutation, appearing for the first time
in this kind of sequencing gel in April of 1989. And I remember very well
the evening where we really were convinced that this was
right because, of course, this could have just been some
sort of random common variant. Doesn’t look all that serious. It’s not a frame shift. So a lot more had to be done
to genotype the families with cystic fibrosis. And you wanted to
be sure you never saw a normal individual that had
two copies of this three base pair deletion because
that would shoot down your theory immediately. And that data arrived in a
fax machine in a dorm room at Yale University in the middle
of one of those human gene mapping conferences
where Lap-Chee and I were working together
on somewhat other topics and going back to
the dorm at night to see what our
labs had produced. And as the fax paper
lying on the floor revealed that
evening, this was it. May of 1989, a rainy
night in New Haven. Of course, we had to
celebrate, so what did we do? Went to Toronto and
opened that doggone bottle of Canadian whiskey. This was 10 o’clock
in the morning. The technician who
took the picture was scandalized that this kind
of thing was happening. But you know, you have
to make the moment count. And of course, we
then had to write a lot of papers in a hurry
trying to get this together, trying not to have the
information spill out before it was
completely peer reviewed and everybody
agreed it was right. We almost made it. There was a little
leak that happened. As the papers had
been accepted and were in their final clearing up in
terms of a few tiny tidbits, the word got out and the
press spilled the beans. And on August 25 of
1989, the word was out. The gene for cystic
fibrosis has been found. My favorite way that
that was notated is a diary entry of
an eight-year-old girl who has cystic fibrosis. “Today is the most best
day ever in my life. They found a J-E-A-N– gene– for cystic fibrosis.” And I don’t know
where JH is today, but I hope she is doing great. We did publish these
papers, three of them in Science, regarding the
walking, the jumping, the cDNA and genetic analysis, which made
it pretty clear that there was a single common variant
namely, DeltaF508, and then a bunch of other variants. There were going to
clearly be issues here in terms of heterogeneity
at the mutational level, but no evidence then or now
that there was another locus. CFTR was it. And we had a
wonderful opportunity to tell the world about this. And Jack and Lap-Chee and I
took part in a number of moments to talk to families,
to talk to the press, and to have wonderful
recognition moments. Here again with the Cystic
Fibrosis Foundation, Bob Dressing on the
left, Bob Beall, looking somewhat younger. But we all were, too, in
this particular recognition. So there it was. And I guess at that point,
like Pam said, many of us thought, OK, that
was the bottleneck. Now things are just
going to fly along. My lab immediately got engaged
in trying to approach the gene therapy strategy. We showed very quickly you
could take a normal copy of CFTR and put it into
cells in culture, and you could correct their
chloride channel defect. So that was the functional
proof we had the right gene. And oh, it’ll only
be a few steps until you can figure out how to
do that for the human airway. Boy, that didn’t turn
out to be the case. The immune system is
not to be messed with. And every time we tried to
figure out a way to do this– Melissa Rosenfeld is here, who
undertook much of that at NIH back in the day– you could get a
transient effect. But then the immune system
would say oh, no, you don’t. You’re putting that virus in
my airway with that other gene. And it was very difficult
to get anything that was sustainable and significant. There was a lot of
work, of course, done about the mutational
allelic spectrum. And we now know of
more than 1,000 ways that the CFTR gene
can be misspelled. But that DeltaF508 accounts
for about 70% of the alleles. And about 90% of
patients, therefore, have at least one copy,
and many of them have two. That G551D was one that was
discovered fairly early on. That one’s important
for Ivacaftor, so I just flag it
there for a moment. A lot of work going on to
figure out the function of CFTR. I’m sure Mike Welsh
is going to talk a lot about the
elegant physiology that was done by his group
and others to try to understand why
those mutations do what they do to function. The gene therapy effort giving
us a lot of headaches– still, in many ways giving
us headaches. But indeed, the real
opportunity then for drug therapy emerging
in the later 1990s. And again, I don’t
think this would have made the progress it has by
any stretch of the imagination without the CF Foundation
putting a very big, high risk bet on the idea that small
molecules could work here because the revealed wisdom
was if you’re trying to treat a recessive disease where you’re
missing a function, that’s going to be really hard
for a drug to take care of. If you’re talking
about a disease where you have a function
that shouldn’t be there, like an oncogene, and
you want to block it, well, that sounds
like something where a drug strategy would work. But a missing function? Oh, boy. And it took a lot of courage
and a lot of resources to initiate that effort through
Aurora, which became Vertex, with a lot of academics also
feeding great ideas into that, and brought us to where
we are here today. And we’ll hear much more
about that from Paul. And here we are with
now the emergence of effective treatment based
on an intimate knowledge of the nature of the gene
and its protein product, leading to therapeutics
for most patients that are truly amazing. But we’re not done. There’s still patients who
have null alleles for whom the current strategies are not
currently going to help them. And we have opportunities
scientifically, because of the way in
which gene therapy is now advancing, because of the
opportunity for such things as gene editing, to fix the
problem at the DNA level. I think everybody
in this room has to be excited about where
that’s taking us for thousands of genetic diseases. And cystic fibrosis is still
there as a highly attractive target to go after. So watch this space. Before I finish, I do want
to say the cystic fibrosis search has also been a
critical one for motivating exciting science across a broad
array of other applications. And maybe first of
all, I would say I’m not sure we would have had
the momentum to get the Human Genome Project started
without both the success in finding the cystic
fibrosis gene and the fact that it was so darned
hard that nobody thought we could do that for
thousands of other diseases. We had to have a foundational
understanding of the genome if this was going to be
scalable to other conditions. And so this did, in
fact, help quite a bit. When the National
Academy was coming up with arguments about doing
the Human Genome Project, CF was noted in that regard. And so the Genome
Project happened. And that now is the reason why,
as you have heard, what we did, which took all those
years in the 1980s, could now be done by a
pretty good graduate student in the space of a
week, I would argue, if you had access to DNA
samples from families, a thermal cycler, and
access to the internet. And yet for us,
this was something that cost upwards
of $50 million when you add up all of the resources
that went into the gene hunting and took all those years. Other things I’d just
like to point to that CF has contributed to. There’s this
question about, well, not everybody who has
the same genetic disease has the same severity. And some of that is variations
in the specific mutation. And we know for CF, that there
are milder ones that spare the pancreas and
others that don’t. But even for people who are
homozygous for DeltaF508, there’s variability. And there are kids that are born
with Meconium ileus and others that aren’t. And what’s that all about? So there must be
other modifiers, and maybe they’re genetic. Well, you bet they are. CF has done a lot in that space. Just found this
diagram yesterday looking for an example of this. And there’s an international
CF Gene Modifier Consortium where people have
gotten together on this. This one’s particularly
interesting because this is SLC26A9, which
happens to be, guess what, a chloride channel. And it turns out that there’s
a common variant, which is a significant factor in
whether Meconium ileus happens or not. That green line
that you see relates to the statistical likelihood
that that particular region is contributing to
risk of Meconium ileus. But notice all those dots. This is data from
the UK Biobank study. They’re not studying people
with cystic fibrosis. They’re studying
mostly normal people, and they’re looking at their
particular expiratory flow in a pulmonary function test. And guess what? That same common variant,
at a statistical level of 10 to the minus
25, is, in fact, a major modifier of what
happens to normal lung function. How about that? Again, the ability to
take advances in genomics and expand from them, CF has
been right in that space. And protein structure. On the right there is
the medieval version of what we thought
CFTR would look like. It was very fortunate
that Jack Reardon had spent much of his
prior career working on the family of proteins
called ABC transporters. And that’s what CFTR is. It was almost, like, cosmic
that you hit this gene. And Jack goes, oh,
I know that one. That’s an ABC transporter. It would have 12
transmembrane domains. It would have two
nucleotide binding sites. And oh, there’s this
other funny thing in there that we didn’t expect to see. Well, we’ll just call
that the R domain. I don’t think that
was for Reardon. I think it was for regulatory. But we always wondered. And that was what’s
in the paper in 1989. And now on the right, you see
the actual structure of CFTR by cryo-EM. Not bad, actually. The yellow thing there
is that R domain. The green thing
to the right of it is one of the nucleotide
binding folds. Yes, all those
transmembrane domains. Not bad at all. But it is so elegant
and so satisfying to now see the real answer. And cryo-EM, which
has come into its own, now makes that possible,
even for proteins that you could probably not
crystallize because they’re transmembrane, this makes it
possible to see such things. Again, I think this structure,
for people who were not completely sold on cryo-EM, this
one really got their attention. And then– and you’ll hear
much more about this one in the final talk of the day– CF and single cell analysis,
where single cell profiling, which has just become possible
in the last four or five years, discovering this
rare cell type, now called the ionocyte, which
is the major producer of CFTR in the airway. I realize I made a
major blunder when I threw this cell together, and
I cited one of the two papers in Nature, but I
cited the wrong one. So you’ll hear this afternoon
from the Boston group, which, in the same issue
of Nature in August, reports lovely complementary
information that describes this new discovery. And who knew? I mean, I thought we understood
the airway pretty well. And all this time, this
particular amazing finding had eluded us. But the ability to do
analysis of single cells and look, do RNA seek and say
what genes are on, reveals. Here is this 1%
population that’s doing most of the CFTR work
with obvious consequences. Oh, sorry. Well, OK, congratulations
to everybody. It’s a Boston effort
all the way around. OK. So let me just finish, though,
by pointing out that what this has really always been about is
not just the exciting science– and it has been wonderfully
exciting science– but it’s about the
people whose lives we have aimed to try to assist. And that was supposed to
be clear in this original publication in Science in 1989. That’s Danny Bessette at
age 6 six on the cover. And we were very
clear we had to have a picture of somebody
affected in this, not just a science diagram. Well, fast forward. This is Danny Bessette 11 years
ago, looking pretty good, 2007. And this is Danny Bessette
today with his wife, Jackie. Danny is doing pretty well,
although in the absence of effective drug therapy
until very recently, he has gone all the way
through a lung transplant. The lung transplant
is serving him well. And again, what a
remarkable story to be able to have been part
of, as I have over these years, as he has been, in
many ways, the face of this particular
condition for many of us. But not the only face. And so as I finish
this presentation, and hopefully turn all
of our attention to what really matters, it’s
really about Danny and Jackie. It’s about Bill, who’s
there at the top, who has G551D and has had a
fantastic result from drug therapy. It’s about Lindsay
in the middle. It’s about Avalyn, one of
the younger kids with CF, for whom I think the future
now is so much brighter now because of what’s
been happening. And it’s about some
of you in this room, who I know are also affected
or who have family members. This is a time, I think,
where we can really celebrate what’s happened. I wish it had happened
in less than 30 years. But it might not have
happened for 50 or 60 years without the shared effort,
the resources, the commitment to get there. And where we are now, for me. Is truly gratifying. And I hope you won’t
think it corny, but I think, before I
sit down, that maybe we ought to have a little group
recognition of what this is all about. And I’m going to ask
you to join in with me with a song, which
I wrote, actually, for an early meeting of the
Cystic Fibrosis Consortium. It was in Nashville, so it
had to be a song that had kind of a country feel to it. And we’re going to do it again. And the chorus has words
that you can see there, so you can’t pretend
you couldn’t learn them. And I’ll sing the verses. I won’t expect you
to do that part. And I’ve modified
the last verse just a little bit because we’re
here on a special occasion. I hope you can hear me all
right on this microphone. Let’s try the chorus first. I’ll sing it for you first, and
then I’ll get you to join in. So pay attention. [PLAYING GUITAR] (SINGING) Dare to
dream, dare to dream. All our brothers and
sisters breathing free. Unafraid, our hopes unswayed,
till the story of CF is history. Come on, here we go. Dare to dream, dare to dream. All our brothers and
sisters breathing free. Unafraid, our hopes unswayed,
till the story of CF is history. As we gather here together
through many streets and towns, doing different kinds
of jobs yet all a team, every person knows the
story of struggle, hope, and love, every person
gathers here to dare to dream. Dare to dream, dare to dream,
all our brothers and sisters breathing free. Unafraid, our hopes unswayed,
till the story of CF is history. Those who join here together
have traveled many miles, some are hopeful, some
are down, some in between. Whether caring for a loved
one or working in the lab, this one thing binds us
all, we dare to dream. Dare to dream, dare to dream. All our brothers and
sisters breathing free. Unafraid, our hopes unswayed,
till the story of CF is history. There were many kinds of
hurdles that must be overcome, but the right stuff came
together in this team. I salute my friends
and colleagues at this Warren Alpert
Prize, [INAUDIBLE] at last we dare to dream. Dare to dream, dare to dream,
all our brothers and sisters breathing free. Unafraid, our hopes unswayed,
till the story of CF is history. Let’s turn the story
of CF into history. [APPLAUSE] Wow. That’s a tough one to follow. Thank you. That was really
remarkable in every way. And I have to say that it
makes me truly grateful for having the human genome
sequence as well at this point. I can’t imagine doing science
without it at this point. Our next speaker in
this path, in this arc, is Dr. Michael Welsh. Dr. Welsh is an investigator
at the Howard Hughes Medical Institute and Professor of
Internal Medicine, Molecular Physiology and Biophysics
at the University of Iowa, where he directs the
CF Research Center. He’s received a number of
awards, including the Walter B. Cannon Award from the American
Physiological Society, and he’s an elected member
of the National Academy of Sciences. In the 1980s, Dr. Welsh had
been doing basic research on salt and water
movement in the airways, and soon after the
cloning of the CFTR, he helped to discover that
it encodes an anion channel. And importantly, one of
his really key discoveries was the observation that the
common mutation that we just heard about is actually a
temperature sensitive mutation, which helped anticipate
that it may be possible that other small molecules
May one day be able to correct that particular defect. So please join me in
welcoming Dr. Welsh. [APPLAUSE] Francis, I’d like to
bring my horse in now. Thank you, [INAUDIBLE]. The First thing I want to
do is say thank you. Thank you to the Warren
Alpert foundation. While I’ve had the good
fortune to be here today, there’s so many other people
that could be here also. I’d like to thank my colleagues. They’re the best. They challenge me. They make the science better. I’d like to thank the staff. They’re hardworking,
they’re innovative. They just get things done
and make things work. The students and postdocs. Seems funny to
call them trainees since I’m the one who’s learned
much more from them than they have from me. My collaborators for the work
we’re talking about today. That’s Alan Smith and
his colleagues that were at Genzyme at that time. I’m incredibly privileged
to work with all these committed individuals. I’d like to thank those who
funded the work, the Howard Hughes Medical Institute,
the Cystic Fibrosis Foundation, people who pay their
taxes to keep the NIH going. And finally, the people with CF. They’ve participated in
so many of our studies, they’ve supported us,
and they’ve inspired us. So here I was, a young
assistant professor. Here’s an airway epithelia. And this is air side, this is
basolateral side, little cilia here. I wanted to understand
how chloride got into the cell and
then how it got out, how chloride was secreted. And with time, we
learned that this process was regulated by intracellular
cyclic AMP, probably by phosphorylation at
the apical membrane. And when we were
doing this work, we had the idea in
the back of our heads that maybe it had
something to do with CF. And then I read this
paper by Paul Quinton. And Paul wrote that this,
abnormally low chloride permeability in cystic fibrosis. And that was really
important because it made me think that
maybe the same thing was true in the airway. And indeed, collaborated
with Jonathan Widdicombe on CF epithelia, and we found
that there is a defect here in the apical membrane. Chloride couldn’t move
across the apical membrane. What that did was provide
a unifying hypothesis for the pathophysiologic defect. It was a chloride permeability
defect in the sweat gland, in the airway, in the pancreas. In the gut. The other thing it did was
set the stage for future work. And some of that
future work occurred when I was standing next to the
fax machine seeing this paper roll out of the fax machine,
this paper that Francis just finished talking about. And I took that and I
looked at that sequence, and I pasted it in the back
of every lab notebook I had for a long period of time
as we tried to understand what was going on. And the first
question, of course, was could we put CFTR into a CF
cell and correct this defect. And we tried for a while,
and it took a long time. And the problem was you
couldn’t put together the cDNA. And I think it was
Rich Gregory at Genzyme that was the first to
realize what the problem was. There were cryptic
bacterial promoters, and he put it together. And we had started collaborating
with Alan Smith and Rich Gregory and their colleagues. But it was hard. Finally, in the early
summer of the next year, after the gene was
identified, the postdoc came out to my house
on Sunday with a gel that looked like this. And there on that gel was CFTR. It was really exciting. The next morning,
Monday morning, I went to see this man. He was my department
chair, Frank Abboud. And I said Frank, I need
to have a sabbatical. He says, oh, well, that’s OK. When are you going to go? I said, next week. And he asked me, well,
where are you going? I said nowhere, I’m
going into my lab. And that was so important. He valued science,
he valued medicine, he’s a big supporter of
physician scientists. And I look back at
that paper, the one here where we first expressed
it several months later and reported it, and I looked
at the Acknowledgments section. Here’s the Acknowledgments
section in that paper in 1990. And right there I thanked
my colleagues James Hunter and Bob Fick for
covering the MICU. The next week I was
supposed to start a month of attending on the
Medical Intensive Care Unit, and that’s a demanding service. And so I asked my colleagues,
and they stepped up immediately. There was a camaraderie
that was really important, and they picked that service up. Let me point out two other
names, Phil Karp and Theresa Mayhew. Phil Karp was my first research
assistant, started with me a couple of months
after I started. He’s still there,
helps me so much. Theresa Mayhew was my
secretarial assistant almost as long as Phil,
and they helped me so much. The reason I’m mentioning
Frank and Bob Fick and James and Phil
and Theresa is because of their
individual contributions. But even more
important than that, they were emblematic of the
supportive environment I had at the University of Iowa. So back to the question. Oh, I’m sorry. I forgot Christopher, my son,
Christopher worked in the lab during the summer part time,
and my other children did, too, Teresa and David, and
I’m really happy about that. So back to the question. Could we do it? Could we complement the
defect, express CFTR? here’s how we went about. We took CF cells and we
put a fluorophore in there, and we also loaded those
cells up with iodide. We knew iodide could get out
of a normal cell but not a CF cell, and we also
knew that iodide inhibited the fluorophore so
the fluorescence went down. Then we put it in CFTR
with the vaccinia virus. We wanted a lot in the cells. We stimulated with cyclic AMP. Iodide could then leave
the cell, and when it did, the concentration would go down,
your inhibition of fluorescence would go down, and the cells
would start to fluroesce. Here’s some of the images
from that first paper. On the top we have CF
cells, and we put in CFTR with the mutation that
Francis just described, the Delta F508 mutation. Now this is CSL. You can see that
little bit of blue there, a little
bit of blue there. You wait a minute,
nothing really happens. But if you put in wild
type or normal CFTR, this is what we got. Look what happened. This guy’s pretty dark. It lights up, pretty
dark, lights up. This one’s already lit up a
little bit, lights up more. The presence of CF
allowed to see this. At the time we were doing this,
I had a postdoc, Deborah Rich, and she was doing the molecular
work and the cellular work. And then I’d pick up
the cells in the morning and disappear into
this dark room with a microscope and
old-fashioned fluorescence. And I would be trying
to develop this assay, and in the late afternoon
I’d come out of there. And I was starting to get
pretty excited because I knew something was going on. And I’d show it to
the people in the lab. And I wanted them to
be as excited as I was, but they were just sort of
nodding their heads at me, and they didn’t
seem too excited. And I thought, geez. And then one day I went to
the cell culture incubator to pick up the
cells, and instead of saying CFTR and
mutated CFTR, it said A and B. My postdoc decided
to blind me without asking me. I don’t know if that says
something about my ability to convince people
with data or not, but it made me a believer
in blinded studies. So what this did was
said that expressing see CFTR in CF epithelia
corrects the chloride transport defect. It doesn’t with the
Delta F508, and it indicates that a
fluorescence-based assay can report CFTR activity. So then the next
question, of course, well, what does CFTR do? Well, as Francis already
suggested, it was in a family, a family called ATP-binding
cassette transporters. And they have a general
structure like that shown here. If this is the
membrane, and here’s a membrane spanning domain. It’s usually about
12 sequences that go back and forth
across the membrane, and two nucleotide-binding
domains, and then something unique to CFTR, as
Francis mentioned, the R domain or regulatory domain. Now most members of
this family are pumps. That’s what they do. You can think about
them like machines. They’re a machine. If you have a machine,
you have a pump, and then you have to have an
engine to drive that pump. And in ABC transporters,
the engine’s down here, the nucleotide-binding domains
and the fuel they burn is ATP. So what about CFTR? This is the general scheme
for an ABC transporter. There’s the idea for CFTR. Here’s the membrane,
here’s CFTR. And probably the most
common idea at that time was that CFTR was
some kind of pump. That, of course, raised
the question, well, what the heck does it pump? Is it actively pumping
something out of the cell or into the cell, and how
does that relate to a chloride channel? If this is CFTR,
how is CFTR a pump or associated with a
channel to regulate it? Well, as I told you, I had
been studying airway epithelia, so we knew that the defect
in the apical membrane was likely a chloride channel. So we set out with the
very simplest hypothesis. And that is, actually CFTR
is a chloride channel itself. It doesn’t require
this complex systems. So how are we going to
go about doing that? How are we going
to test that idea? So if this is CFTR, and
here you have a pore now. It can let chloride flow either
direction across the membrane. And with some
informed guesses, we guessed that this pore, because
it’s a little bit nonselective, would be lined by the side
chains of the amino acids. And some of those would
be positively charged. So our idea was if we took one
of those, we could change it from a positive charge
to a negative charge, and that ought to affect the
way chloride goes through. Here’s how we did
the experiment. Took HeLa cells, or a
variety of other cells, expressed CFTR in them,
took a little patch pipe, had a little glass
patch pipette, set it down in the cell,
break the membrane. And now you can perfuse
the inside of the cell with what’s in
the patch pipette. And then we can look at the flow
of chloride through that CFTR and measure that electrically. And we can change the
composition, chloride, bromide, or iodine. This is an example
of what we got. Here you see the flow
of chloride or the ion we’re studying, on the
y-axis and the voltage here. And where these little
traces cross this line are called the
reversal potential. And they let us know
about the permeability. And we learned that bromide was
more permeable than chloride than iodide. Then we could, over here,
change one amino acid from a positive side
chain to a negative. And we did that with
changing lycine, which was positively charged
to glutamic acid, which is negatively charged. And that’s what we got. Iodide was the least
permeable in wild type CFTR, and that became the
most permeable here, and then bromide
and then chloride. This was really exciting. I remember Matt Anderson, who
was doing those experiments, running out of the
patch clamp room. I don’t mean walking out
of the patch clamp room. Running out of the
patch clamp room saying, I can’t believe it worked. I can’t believe it worked. It was really exciting
because it said that CFTR is itself a chloride channel. And it really said
more than that. It said it’s an anion channel. It’s not just chloride,
it’s also bicarbonate. And I’ll come back
to that later. What this did was
link the genetics and the molecular biology
here to the physiology in airway epithelium. It provided a strategy for
discovering and testing therapeutics, and then that
led to the next question. Well, how does CFTR work? How does this
complex thing work? We did a variety of studies. We looked at the
membrane-spanning domains. We realized that they formed
a pore, a watery channel– I won’t say water– a
channel through which chloride, bicarbonate,
and other ions can flow. We looked at the
nucleotide binding domains. We added ATP, and then we could
look at single channel traces. In a single channel trace we
look at ATP, and we now can say is the channel closed or open? And we did that. And with time, running
along this axis, we never saw the channel open. And then we could phosphorylate
the R domain with cyclic AMP dependent protein kinase. And now in the presence
of ATP, we saw this. The channel was open, and
then it flickers closed, flickers open, flickers
closed, flickers open. It’s so exciting. It was just thrilling
to see that flickering across the screen as we did it. And that helped us understand
the function of CFTR domains and how activity is controlled. So if you want to
understand this protein or do anything about it,
this provided the background information. Well, how do mutations–
people with CF have mutations in the CFTR gene. You just heard about that. How do those mutations
disrupt CFTR function? Well, geneticists around the
world were sequencing CFTR, and they found lots of
different mutations. And the strategy was introduce
those mutations into CFTR to discover what goes wrong. How do those
mutations break CFTR? And as the results from
our lab and the results from the lab of Alan
Smith and his colleagues, we began to see there’s
lots of ways to break CFTR. And in looking at that,
you began to see patterns. And I thought at that point that
we should have a classification scheme. This was 25 years ago. And here I’m going to show
you the classification scheme we developed at that time. This is a figure
from that paper. So here we have a cell. Here’s the apical
membrane up here. You have the genetic
material in the nucleus. You make CFTR in the
endoplasmic reticulum. It then travels down
the assembly line to the Golgi complex,
and from there it’s off to the cell surface, where
it can function as a chloride channel. In CF, the genetic defect,
you have your mutations here. And we realized there are four
general ways to break CFTR. One we called Class 1, and
that was defective protein production. You never made a
full-length protein. The protein’s
actually never made. . Lots of mutations like that. Here’s one example, W1282X. In some cases, the
gene is here, you make the protein in the
endoplasmic reticulum, but it’s somehow misfolded, and
the cellular quality control system says don’t send
it on down the path. Take it off the assembly line. That’s the Delta F508 mutation,
the most common CF mutation. Xang Chang and Alan Smith
and their colleagues were the first to recognize that
the Delta F508 mutant couldn’t get out of the ER. In some cases the channel
actually was made, it got polished up
in the Golgi complex, and then it’s off
to the cell surface. And it sits there,
but it doesn’t open. That little trace I
showed you, it just doesn’t flicker open very often. As a result, you don’t
have chloride current. An example there is G551D. And some mutants made a
channel, got to the memory, and they’d flicker
open, but something was wrong with the pore so
it didn’t conduct chloride and other ions very well. And an example there is R117H. What was the results of this? Well, it described how
mutations disrupt CFTR, say what goes wrong. it helped with genotype
phenotype analysis. Some people with some mutations
have milder phenotypes that lead to milder disease,
some more severe disease. It helped us with that. But probably the
most important thing was it provided a blueprint
for developing therapeutics. If you want to think
about going after one mutation or the other, it
suggested different places along here that you
might need to fix that. Well, that of course raised the
question can you correct CFTR? Can you correct the mutant CFTR? At that point we were
thinking about that, and Francis Collins’
lab published a paper, and they showed that
CFTR made current. And of course, my first reaction
was what did Francis do wrong? Sorry, Francis. But I read the paper again
and I realized, well, they had done their
experiment in xenopus oocytes, and that’s really important
because in xenopus oocytes, you grow those down at 18
degrees centigrade or so instead of 37 degrees
centigrade, where we live, humans live. And so that was really
an important paper because it then suggested
maybe mutations in CFTR with the Delta F508 mutation are
a temperature sensitive mutant. This is the experiment
that we did. Here we’re looking at
CFTR running out on a gel. This is wild type or normal
CFTR, 37 degree centigrade. I want you to
focus on this band. This band is the mature protein. This is the kind of protein
that goes out and sits on the cell surface and makes
a chlorine or bicarbonate channel. This is CFTR with the
Delta F508 mutation. And what’s important here
is what you don’t see. You don’t see any
mature protein. It never reaches out to
where it’s supposed to go. And then we did a
simple experiment. We just lowered the
incubation temperature. And what you start to
see here is Band C here, the mature protein
starts to appear as you drop the temperature down. And that was really
exciting, and it was particularly exciting
because we also saw this. We measured the
chloride current. This is CFTR with the
Delta F508 mutation. Here’s 37 degrees. You drop the temperature
to 30 degrees, now you could see current. It wasn’t normal. It only opened about 1/4
quarter or 1/3 as much. But boy, it was
really nice to see. It was a lot of current. So with that, we could
say defect in CFTR actually can be corrected. It said that distinct mutation
classes require distinct fixes. If you want to increase
the opening of CFTR, you may need one type of fix. If you want to fix a Class 1
defective protein production, you might need gene
therapy or gene editing. If you want to fix
defective folding, you may need a different
type of treatment, different type of medicine. And it also says
that some mutations affect more than one class. So for example, the
Delta F508 mutation not only couldn’t get out of
the endoplasmic reticulum, when it got here, it didn’t
work quite as well. And so you might need
more than one fix for a specific mutation. And last, let me
just briefly talk about how does loss of CFTR
cause the airway disease? So here’s the problem. I told you CFTR is
an anion channel, sits in the apical membrane
of airway epithelia. And Pam told you about what you
see when you go to the clinic. Here’s Dorothy Anderson, 1962. Here’s the airway. It’s filled with infection,
neutrophils, bacteria. And the question is how
do you get from here to what we see in the clinic? Well, there’s some
host defense defect. What is that? Well, it was
difficult because it was impossible to investigate
in newborn babies. And mice don’t get
disease like humans can. So what are we going to do? Well, being from Iowa, we
decided to go for a pig. And we wanted a CF pig, but how
are you going to get a CF pig? CRISPR wasn’t around. TALENs weren’t around. There were no
embryonic stem cells. There was no other
mammal other than mice in which it was possible to
mutate a specific gene to get a specific mutation. And so it was tough. And other people in the lab,
they were wondering about this, and they took it upon themselves
to do this above the bench where people were
working on this. They hung a little pig
up there that could fly, and the message
was pretty clear. We’re going to get that
CF pig when pigs can fly. But ultimately we
were successful. We used homologous
recombination and cloning. This is our first litter
from a heterozygous mating, and some of these
pigs have CF, and we did this in collaboration
with Randy Prather and his colleagues at the
University of Missouri. And when these pigs got older– here’s a two-month pig– you take a lung out, you
set it on the lab bench. Look at that. See how it’s sort
of whiter there? That’s whiter because
there’s air trapped in there. The air can’t get out normally. It’s trapped because here’s
a cut section of lung, there’s a blood vessel,
there’s a blood vessel. Here’s a small airway,
and it’s filled with mucus and infection. Here’s the trachea and the
infection with staph aureus. Here’s a five and
1/2 month old pig. Here’s a cut section. You see pretty bad
disease right there. Look, you see that
little slice of mucus. They were thick and elastic
and you could pull it out. And when you put
it on a slide it looked like this, with lots
of neutrophils and bacteria. In this case, it was
bordetella bronchiseptica. So now we could
think about this, and we could think about how
do you get from here to here. And we knew that it gave us a
big opportunity because the CF pigs, on the day they’re born,
they don’t have any infection. They don’t have
any inflammation. The airway’s not remodeling. So you can look in a
very clean, pristine set. And we knew that the pigs, with,
time, went on and developed disease that looks just like
Dorothy Anderson had described years earlier. And we knew that there
was a host defense defect that led to infection and
inflammation and destruction. And that said this
is a critical time. I badly wanted to know what’s
the first thing that goes wrong on the path to disease. And that’s
particularly important because of this, babies. We have universal
genetic screening so we identify babies at birth. It’s really important. We have an opportunity
and a responsibility to try and do
something about that. So we focused on that. Now I wanted to know what
are the origins of CF airway disease. So here’s a cartoon. I’m just going to tell
you about one study. Here’s the airway epithelia. The air would be up here. This is a submucosal gland. CFTR is in the submucosal
gland as well as in the surface epithelium. And this submucosal gland,
in the first 10 generations of the lung, makes
most of the mucus. So we could take
these pigs, and we wanted to do this without
instrumenting them, without going in or
anything invasive. And we take these pigs and
we put in little micro disks. And we put this pig in a
CT scanner and get a CT. And we ran that CT
scanner pretty fast. We got a high resolution
CT, every 15 seconds, and we could track down where
these little micro disks were, and we got this kind of picture. So here we have an
automatic reconstruction of the airway in
that individual pig. And here are these
little micro disks that we’d blown into the
lung just before this CT was started. And I’m going to show you this. These are larger than
the actual micro disk, but you can see them. I’m going to show you a
compressed video from 10 minutes. And I want you to
look at two of these. Watch that red guy, and
then watch the green guy. Here they go. Look at that red guy. He’s a racehorse, passes almost
everybody on the way out. The green guy just taking
his time, dwaddling along. We immediately knew the
textbooks were wrong. Mucus isn’t a blanket. Blankets don’t move like that. It suggested that
enormous heterogeneity. It must be individual
strands of mucus that were causing us to see this. We did the same
thing with a CF pig, and we didn’t see any difference
under basal conditions. So then we turned
around and stimulated the submucosal glands
to secrete mucus. And when we did this, we got
what I’ll show you right here. Most of those moved
completely normally. Two of them are stuck. They’re bouncing around because
of respiratory movement. Finally the red guy lets go, and
he moves completely normally. So that suggests there’s
something stuck here. And it did a lot of studies. And once these
disks were moving, their speed was the
same, the B frequency the same, pericillary
liquid depth was the same. What was different was the
percent of time moving. So on this axis is
percent of time moving. The blue is non-CF. So for example,
in this pig, this would be the average of all the
micro disks, maybe 10 or so, and about 80% of the
time they’re moving. The red is CF. And what you see is there’s
a lot of variability. But on average, they’re
spending a lot less time moving. We wanted to know what that was. And it suggested
those micro disks are stuck due to abnormal
mucus, so we went ex vivo. We took the trachea
out of the animal, pinned it out on
a dish, and now we can look at mucus
being produced. And when we do that, we
can add little 40 nanometer fluorescent nanospheres, and
they’ll stick to the mucus, and then you can see them. Here you see the
submucosal gland duct, you see the cilia beating. See that strand of
mucus coming out? It goes out, it stretches
out, and then it breaks and it goes flowing
across the airway surface. And you can look
down on that surface. There’s a strand of
mucus right there, and you can see
them flowing across. There goes a squeegie strand. It’s like a squeegee
sweeping across, picking up things in its path. This is done under
basal conditions. If we do the same thing,
and now we stimulate the submucosal glands
to produce mucus, look, it’s just lots of strands
and squeegees sort of sweeping across, picking
up everything in their way and clearing the airway out. I showed you this before. This is a little bit
similar to one I showed you. There’s a submucosal gland duct. You see this growing
and growing and growing. It snaps, breaks free, and
sweeps up across the airway. This is like what you see in CF. Here’s the submucosal
gland duct down there. And as I start this
video clip, what you see is it’s hanging on. When this hangs on,
it doesn’t break free. Things attach to
it, but they can’t get carried out of the airway. As what we were doing this,
we were having a lot of fun. It’s fun to watch
movies in the lab. And then we started to
realize what we really wanted to be doing was
looking at what’s not moving. So we developed an averaging
technique so we could do that. I’m going to show you this. This is a panorama. Here you can see pins
holding this tissue out, these black things. And this is up towards the head. And what you’re
going to see here is a video clip of this panorama
over the course of one hour. And here’s what we’re seeing. Right there we
added methacholine, so that now the submucosal
glands are secreting mucus like crazy. What’s important here? It’s what you don’t see. You don’t see any
mucus out there. There is a lot of
mucus secreted, but it all goes on up here,
and you don’t really see it in this averaging technique. Then we turn around and do CF. You already see little bits
of bubbles of mucus above some of the submucosal glands. And here, when we stimulated
submucosal gland secretion right here, you see
these enormous strands of mucus bunching
together, piling up, and not being cleared. We then thought, well, if
CFTR’s responsible for this– and this happened down
in the submucosal gland– then we ought to be able to
reproduce the loss of this by taking out chloride,
taking out bicarbonate. So now I’m going to show
you something from non-CF. This is from a non-CF pig. We got rid of bicarbonate. We got rid of
chloride secretion. And this is what you see. Again, same kind
of process I just showed you, one-hour vide clip. There we added methacholine. What do you see? You see all these mucus
strands building up out on the surface of
the airway epithelia. They’re stuck. They can’t break free,
and they cause that. So where are we Here
we’ve got CFTR down here in the submucosal glands. When it’s defective, the
pH goes down in here, the volume goes down in here. And now when we secrete mucus
up out here, it can’t break. It doesn’t break. It breaks sometimes. But it often doesn’t break,
and you get this held up here. And so in these newborn
pigs, before infection or inflammation, when
along comes bacteria on to the surface, they’re
not cleared out as well. That links loss of CFTR,
chlorine bicarbonate secretion, to defective
mucociliary transport. I need to give credit to Jeff
Wine, Steve Ballard, Paul Quinton. Their work informed
what we were doing in the lab in these studies. But it’s not that simple. It’s more complex than that. We also have an
antimicrobials up here. These are like
antibiotics that kill the bacteria that we inhale. So they kill a lot
of those bacteria. And we know that CFTR is
not just missing here. It’s missing everywhere, so
it’s also missing up here. CFTR is missing here. We’re lacking
bicarbonate secretion. Bicarbonate’s a
base, and so when you lack bicarbonate
secretion, the pH goes down. And when the pH goes
down, antimicrobials become less effective. They still kill some
bacteria, but not as many. That says that loss of CFTR
partially impairs two host defense mechanisms. And so neither one is complete. You got a partial defect in
antimicrobials, partial defect in mucociliary transport,
probably additional things. And these two partial defects– neither of which alone would
cause the kind of thing we see in CF– the combination of those two
cause a particularly vicious disease. Good news about that is if
you fix just one of those, you might have a
positive effect. It also suggests additional
targets for therapies. So these data link the
genetics, the molecular biology, physiology in disease
and enables therapies. I started out, I
just wanted to know how does chloride move
across these two membranes. Along the way we realized
that chloride and bicarbonate were moving through
ion channels here. And that began to explain
some of the normal physiology. We learned how mutations
disrupt function and how that can be corrected. And we begin to understand
how that causes disease. And that, altogether,
enables therapies. The last thing I’m going
to say, we’re not done yet. We’re not done yet. This is complex. We don’t understand well all
the things that are going on. We clearly don’t
have all the answers. We clearly don’t have
all the therapies. We’re not done yet. I started out by thanking people
here, some of my colleagues. Now I’m going to thank the most
important people, my family. Here we are 15, 20 years ago. We’ve grown. Here we are more recently. I’m really fortunate. My grandson, Nathan Welch,
is here with me today. Thank you. [APPLAUSE] We’re going to take a break. But before we take
a break, I do want to just let everyone know
that about an hour ago or so, there were some reports of some
possible shootings near Simmons College. It’s been an evolving situation. And since then,
apparently, there have been an all
clear signal that have been sent out as well. So I want to let folks know. As far as we know right
now, it’s all clear. It was unfounded. It was an unfounded. OK, great. So we’ll reconvene
INDISTINCT CHATTER] OK, we’re going to try
to get started again. If I can get
everyone’s attention, we’re going to get started again
with the second half of today’s symposium. And we have learned that
there’s an all clear on what happened earlier. So I think we can now relax even
more and enjoy the second half. It was an unfounded
alert, George tells me. So I’m going to transition
now a little bit. And before I introduce
our next speaker, I did want to make
a couple of remarks. The Warren Alpert Prize
formally recognizes individual scientists for
their research and discovery contributions. That’s the way that the Warren
Alper Prize is designed. And we’ve heard from two
of the recipients so far. And we’ll hear from two
more in just a few minutes. But today both Harvard Medical
School and the Warren Alpert Foundation did want to
provide a special recognition for another entity without
which none of this stuff really would have been possible. That’s the Cystic
Fibrosis Foundation. This is sort of– [APPLAUSE] I think Dr. Davis did an amazing
job of actually reviewing what their many contributions were. And I may be a
little bit repetitive with what she’s already said. But I think given how
important their contributions were, I think it’s
just worth reiterating. It’s an amazing foundation
that has actually participated, if I’m not mistaken, in
virtually every single aspect of even the discoveries that
we’re celebrating today. The foundation was
started in 1955. And it was started
by concerned parents who were determined to save
the lives of their children. Their early contributions
included the creation of these specialized
clinical care centers. And by 1978, there
are more than 100 different accredited
centers across the country and the world. Now, they made a number of
different contributions, not only to help in the
clinical care of these patients, but they also created what
are called patient registries. And patient registries
are extremely important for defining the
natural history of a study. And if you don’t know what a
natural history of a disease is, it’s very difficult
to develop biomarkers. It’s very difficult
to mobilize troops for getting clinical trials
accomplished as well. The CF Foundation
helped to support some of the early work
in the gene mapping. They also helped to support some
of the early basic molecular biology that we
just heard about. But by the late 1990s, the
CF Foundation leadership was appropriately
becoming impatient. They really wanted a cure. So in the 1990s,
they created what was called a therapeutic
development program. And they invested a huge
amount of money, tens to hundreds of millions of
dollars, in academic research. But also– what was
quite innovative and quite distinct– they
actually made investments in biotech companies in
the pharmaceutical sector to help encourage them to take
on a rare disease, which is traditionally not profitable. One of their early
investments was in a small fledgling
biotech company in San Diego called
Aurora Biosciences, which had been started by Roger
Chen and Charles Zucker, who were at that time
in the San Diego area. And this actually sparked Aurora
Biosciences to actually start working on cystic fibrosis. This company would
then get taken over by the Boston-based
Vertex Pharmaceuticals. And the foundation would then
make yet another very large investment in Vertex
Pharmaceuticals to help them to
continue to invest in drug discovery for
this rare disease. And this is a
pretty unusual move. It’s now called
venture philanthropy. It’s actually the poster child
for venture philanthropy. And this high risk
investment actually paid off in many, many ways. Because it actually led to
the discovery of the therapies that we’ll soon hear about. And it was also a huge
financial success as well. Harvard Business School has even
written an entire case study chronicling the role of the
Cystic Fibrosis Foundation and its impact it’s
had on this disease. And so I just want to take
another moment right now to really recognize the special
efforts of the Cystic Fibrosis Foundation. And we celebrate their
contributions to everything that they’ve done
alongside with all of the scientific contributions
of our award recipients. So if you’ll join me in
another round of applause for– [APPLAUSE] So this is a great segue
then for my introduction of the next speaker, who
is Dr. Paul Negulescu. Dr. Paul Negulescu is currently
the senior vice president of research at Vertex
Pharmaceuticals. Dr. Negulescu received
his bachelor’s and his PhD from UC Berkeley in physiology. And then he continued
post-doctoral work and epithelial biology
and in biophysics. Following his post-doctoral
training in 1996, he moved to San Diego to join
this company Aurora Biosciences that I just alluded to as one
of its earliest employees. He would end up spending the
next 20 years or so initially at Aurora than at
Vertex Pharmaceuticals, leading the team
that would eventually discover three different
classes of medicines that target the different mechanisms
that we just heard about from the last talk. And what’s really exciting is
that work that he has led– the team that he
has led has now led to these precise medicines
that are now given in a genotype-specific manner. So please join me in
welcoming Dr. Negulescu. [APPLAUSE] Wow. Thank you for that introduction. And I could not
be happier to hear the acknowledgment of the
Cystic Fibrosis Foundation. When I got a call 20-ish
years ago from Bob Bell, I was working in the lab trying
to set up some sort of piece of equipment. Someone came into
my lab and said, there’s this guy on the phone
from the Cystic Fibrosis Foundation. I can’t get him off the
phone until you talk to him. And so I went into
the conference room. And Bob was asking, I’ve
just been to this conference. I saw Roger Chen give a
talk about high throughput screening. I want to do high throughput
screening for CFTR. I want you guys to do it. Can you do it? I said, uh, we can try. So that got us started. And it changed
our lives, changed the lives of many people in
this room and many others across the country
and across the world. But I’ll never forget
that phone call and also the mentorship, Bob, that you
provided to me along the way. Thank you. [APPLAUSE] So this picture is from one
of the Great Strides walks. This is the San Diego crew. And I realize that this is
actually a pretty old picture. We take this picture
every year, but I realize now I’ve been using the
same picture for over a decade, about 15 years. And I can tell because– let’s see here–
that’s one of my sons. And he’s eight. And you want to stand up so they
can see what you look like now? [LAUGHTER] [APPLAUSE] So we’re going to
get a new picture. So I’m going to talk
about fixing CFTR. So I think the previous
speakers have set up this section of the
talk beautifully. Because in 1998,
when Bob Bell called, and he wanted to do high
throughput screening to try to discover drugs
to fix CFTR, we had already a very strong
foundation of knowledge about the disease,
what caused it, and perhaps some inklings about
ways to go about treating it. And we know from the work
that you’ve heard about earlier this afternoon that
CF was caused by mutations in the CFTR gene, CFTR gene
included a chloride ion channel, and that
mutations affected the function of that channel. The common mutation Delta-F508
caused a processing defect that Mike Welsh nicely
described, but also had this defective gating. So even when it
got to the surface, it didn’t function very well. So there were a couple
of problems to fix there. So we also knew by
1998 that gene therapy, the initial, perhaps,
most straightforward conceptual path, was going
to take a little longer. It wasn’t around
the corner in 1998. So alternative approaches
needed to be tried. And actually there were
many going on at the time. There were ways
people were thinking about bypassing
the CFTR protein, trying to activate alternative
fluoride channels that might be resident in
the airway of humans, or blocking other channels that
might regulate fluid secretion. But there was this
hesitation at the time around the small molecule
world to actually go at trying to fix CFTR protein. And I think it’s for the
reasons that Francis mentioned. There had never been
drugs that restored the function of defective
proteins, mutated proteins. Most drugs turn things off,
not fix things that are broken. So there was a
conceptual barrier there. And when Bob called,
I was 35 years old. I was really too dumb to know
how hard it was going to be. [LAUGHTER] So we were up for the
challenge of trying to fix that broken protein. And it was so important,
as Francis mentioned, that there was protein
in these cells. The Delta-F mutation allowed
the cell to make some protein. Not a lot, but some. So there was something there,
perhaps, for a small molecule to grab onto. And that was going to
be our starting point. So the team– I won’t go over this too much,
because we’ve talked about it already this afternoon. But the CF Foundation
drove this. We were a biotech company
located in San Diego. We had a service model. Our model was, come to
us with your problems, and we will try to create
screening solutions for you. And so he saw that
as an opportunity to approach us
directly and say, let’s invest in that
early stage biology and screening to try to find
compounds that might rescue the function of CFTR. Of course, the foundation
had a lot more around it than just the investment in
the early stage research. They had the R&D network
which was so important. I’ll talk about that later. And ultimately, of
course, the patient and clinical networks, which
Bonnie will talk about as well. So they had all
the pieces to allow for a transition between
research to development. So our part at the very
beginning as Aurora was around the early
stage of drug discovery. We were early adopters and early
innovators in this discipline, at the time, called high
throughput screening. Nowadays, it’s pretty common. But in 1998, it was
just getting started. And this was because
there was a confluence of new targets coming
in from biology, but also new breakthroughs
in chemistry that allowed large chemical libraries
to be synthesized very quickly. And therefore, the
question was, how can you assay all those chemicals? How can you test
for their activity? We needed ways to do that. And then there were assays
that were needed as well, tests that could be run
against hundreds of thousands or millions of compounds. So we were working
in that space. And we started working
with the CF Foundation in 1998 on assays
specifically designed to measure CFTR activity. When Vertex was acquired– when Aurora was acquired
by Vertex in 2001– I still have a psychological
thing about that. [LAUGHTER] When Vertex acquired Aurora in
2001, we continued the project. And actually Vertex added
a lot of capabilities that Aurora did not have
that turned out later on to be very important for
the further development of the program,
including expertise in medicinal chemistry, clinical
development, and actually a pathway to take something
that we discovered all the way to the patient. And of course, the CF community. I think you don’t have to
be long in the CF field to know that it’s much
bigger than any one person, and that it is supported by this
network of stakeholders, which include patients, of
course their families, but also physicians, friends. It’s like a family. It’s more than a
community, actually. It’s a family. And pretty soon you
figure that out. I remember my first
interaction with a CF patient, a person with CF. It was a four-year-old. Came to our site
with his parents. Didn’t come in on his own. And he was speaking
to our site staff the day before one of
these fundraising walks. And he basically
said, I just want to thank you guys for
working on my vitamins. And I think at that moment
we kind of figured out we had to kind of see this
through and commit to it. Because that little
guy was counting on us. And he told us that. So the community was there all
along and really inspired us. Continues to do
that to this day. So our goal. Well, simple. Let’s find a pill. Let’s find a
medicine that you can take by the mouth
that could treat all of the organs affected by CF. CF is not just a lung disease. And I remember also a
CF patient telling me at one of the Williamsburg
meetings, please don’t give me another
inhale therapy, because my lungs can’t take it. I took that back to
our team and said, let’s go for a pill
rather than inhaled. And everyone said, yeah,
that’d be a good idea. Let’s do that. So again, hard, harder. Perhaps in inhaled
therapy you have to do a lot more
optimization of the compound. But ultimately it
was those sorts of interactions with
the community that shaped what we did. So a pill. So how to make a pill. What should it do? Well, as we thought
about the problem, we obviously could
leverage all the knowledge from the previous
10 years of research about the underlying defect. We wanted to target initially
the Delta-F508 mutation. It’s the most common mutation
present on one or two CF alleles in 90% of people. So it was a logical
place to start. But we thought there might be
two types of medicines needed to maximize the effectiveness
of a small molecule therapy for CFTR. The first class of compounds
was called the CFTR correcter. And as Mike Welsh discussed,
this type of molecule would facilitate the
processing of that protein, get it out of the endoplasmic
reticulum and out to the cell surface. Maybe the channel would
function when it was on the cell surface, maybe it wouldn’t. Maybe it would retain
that gating defect that is intrinsic to the Delta-F508. Or maybe it would be worse
if we corrected and got jammed more to the surface. So simultaneously, we
thought, let’s look for something that can get that
Delta-F protein to function as high as possible,
as much as possible. And that’s a potentiator. That’s a second compound
that acts on the protein when it’s in the plasma
membrane and helps to enhance its function. So we started out looking
for two types of molecules. We needed a very
sensitive assay. To do high throughput screening
on hundreds of thousands of compounds, you
want a test that’s going to work every
time you introduce a compound to those cells. You have to be able
to scale those cells. You have to get the
same result each time you do the experiment. We screened many
different methodologies to select the one
we ended up with, which is shown here as a
membrane potential dye. Little different method
than the one that Mike mentioned where he had
an intracellular assay for concentration of the ion. We actually looked at the
transition of the ions from inside the
cell to the outside. Because when chloride
leaves the cell, it leaves behind
a positive charge. And if you have a
sensitive enough assay, you can detect that change
in the membrane potential. Your cell line had to
be very tight, though. It had to have no
other channels in order for the small amount of
current coming out of CFTR to change the
membrane potential. And through the help
of the foundation, we screened over 20
different cell lines expressing or not
expressing CFTR to find the right
kind that would be suitable for
this sort of assay. We ended up with one that
worked out of all 20. Fortunately, we didn’t
have to build a new one. And the one that we got was an
NIH 3T3 cell from Mike Welsh. So the assays, just to talk
a little bit more about them. They’re the foundation
of the program. I’m going to take a
minute or two just because I’m an assay person. And you’ll have
to sit through it. But the other reason is
that without a good assay, you don’t have a good screen. Without a good
screen, you don’t have a starting point for chemistry. Without a starting
point for chemistry, you don’t have a drug. So the assay is the foundation. And we spent the time
to try to develop one that was going to work
under the different conditions that we wanted to
test for modulation. So the potentiator
assay, very simple assay. Take 3T3 cell expressing
the Delta-F508. It’s inside the cell. It’s not at the cell surface. Now this is for the potentiator. Correct those cells. Take them out of the incubator. Put them overnight on the bench. And let nature, the
low temperature, move that CFTR to
the cell surface. That was the insight
from Welsh’s lab that allowed us to move
that Delta-F to the surface and then allow us
to add compounds on top of those
temperature corrected cells to look for potentiators. And so the assay was simple. If there was a
potentiator in the well, the chloride would
exit the cell, and it would change color. Now, to check for
correctors, we just changed the order of things. We added test compounds at
the beginning of the assay, incubated them overnight. And if they were correctors,
CFTR would move to the surface. And then to pick them up,
we added a potentiator for the second step to
maximize the flow of ions through the cell. So in this case, if the
compound is a correcter, the chloride exits, and
the cell changes color. So the same readout, just
a slightly different order of addition of the compounds. We basically run this
sort of assay to this day. We’ve run it for 20 years, keep
finding compounds using it. We’ve adapted it. We’ve moved it into
different cell types. But we still basically
use this format. All right, so we had
to take those cells, put them onto an
automated platform in order to screen the hundreds
of thousands of compounds. At the time, you kind of had
to mix and match robotic arms from the automotive industry. There weren’t these
little automation things that you can go buy
and put in your lab. So they were not really that– they were for bending
steel and stuff. So we dropped a lot of plates,
crushed a lot of vials, but eventually got them to work. And I’m going to show you this
vintage video from the year 2000. Here are the compound trays
being handled by the robots. We had adjust their
little hands so they could grab the little guys. Moving around, they’re
bringing the compounds to a transfer station
where they then get added to the cells that
are in these 384 well plates. The plates then get
incubated with compound. Here you see the
dye being added. The yellow liquid. Compounds are incubated
and then go into a reader, plate reader, which
was designed, actually, by our engineers at Aurora to
detect the very small signals coming out of these cells. So there it goes. Looks like a barbecue, but
it’s actually a plate reader. And those are the signals
coming off of one well of– one plate of test compounds. There weren’t any hits
in that one, by the way. We had to screen through
hundreds of thousands of compounds to find any hits. This shows you a little
bit of the screening flow. So compounds go in the top. A few come through. This screen had very– both these screens
had fairly low hit rates, especially
the correcter screen. So out of 10,000
compounds you might find one that had activity. And it was weak, often
very weak activity. We had to take those
through secondary tests to make sure they
weren’t messing up the cell in some other way. A lot of the hits, a lot
of the nonspecific hits would just change the membrane
potential non-selectively by punching a hole in
the cell membranes. Probably found a
lot of antibiotics, but that’s another story. And then other assays to
confirm the activity on CFTR. And we end up at the
end with a few compounds for further evaluation. From the first rounds of
screening that we did in between 1999 and
2004, we came out with six compounds for
further evaluation. Not a lot of compounds,
but enough to get started. And of those, there were one
of each class that kind of rose above the rest
based on initial chemical exploration of the
potency and ability to modify the activity
of the compound. I’m going to tell you the story
of how those two compounds became the drugs we have today. So we’ll start with
the potentiator. The screening hit
484 is on the left. The final drug, VX-770, which
became ivacaftor and then marketed as Kalydeco,
is on the right. You can see the chemical
similarity, actually, between the hit
and the final drug. It’s really clear that
left hand side was not changed at all through the
medicinal chemistry process. But through that
process we’ve made over 800 different
compounds testing each one for its activity. Some got better. Some got worse. Some improved the
bioavailability, which was important if
you’re going to have a pill. Some improved the
stability in plasma, so it didn’t get degraded
by the body too much. That’s the genius of
the medicinal chemists to pull all those
features together and make a compound that not
only works on your cell, but can be taken as a drug. And the final compound was
about 1,000 times more potent than the original hit. The efficacy was about
the same, but the potency was much improved. This allowed us to end up
with a 300 milligram dose. If we’d started
with our hit, you would have been eating, I
think, about 300 grams a day, which is too much. So we needed to get
that potency down. And sometimes we gloss over
this step of the optimization of the compound. So we’re going to
represent it here to you in a little video showing you
the first 400 or so compounds that we made. And on the right, I’m going
to show you the potency. Higher potency is up. And then the compound number
as it proceeds to the right. And keep in mind that each
compound you’re going to see was made handcrafted by
a chemist trying to do exactly what they wanted to do. Move a hydrogen there,
move a bond there. And this was 18
months of effort. Now, this is speeding 18
months into 18 seconds. See that left hand side? Some exploration, trying
to make it better. Nope, nope, nope. Keep it back. Bring it back. And then work mostly on the
right hand side leading to 770. One thing I like
about this slide is it shows that the most
potent compounded did not become the drug. In fact, we found the
most potent compound, but it had such poor
properties as a drug that we actually
had to backtrack. Go back to a lower potent
compound and build it back up. These are the twists and turns
of the drug discovery process. So here we were in
2004, five years after we’d started our
efforts, with a candidate that could advance to the clinic. Super, super exciting. But we didn’t have a correcter. So we couldn’t take it to
the Delta-F mutation yet. What were we going to
do with this compound? I remember sitting around the
boardroom table in San Diego. Preston and Bob are all
looking at each other. What are we going to
do with this compound? Wait for the correcter? Or take it into one of these
other very rare mutations that Mike and others
had characterized? And so that’s what
we talked about. Could the CFTR
potentiator by itself work on a fraction
of the CF population, the fraction of patients
that have either gating defects in the channel,
channels at the cell surface that doesn’t open
properly, or those where there is some
channel at the surface, but it’s partially functional? We call those residual
function mutations. Generally not associated with
the most severe forms of CF, but a significant fraction
of the population. Overall these two groups
make up about 15%. So we thought let’s
try it in those. So remember that all the
optimization, though, had been done on
the Delta-F protein. Could have been that
the potentiator was specific for Delta-F. So we had
to test if that was the case or not. So we got cells expressing
some of the other isoforms, including the G551D. We tested 770 on those cells. And this is showing you
patch-clamp data looking at the single channel behavior
of D551D plus and minus ivacaftor. And I’m showing you at the
top what normal CFTR looks like when it’s phosphorylated
and allowed to operate in this patch configuration. It’s open about half the time. The G551D under the
same conditions, very little activity. Dead. Confirming it’s a gating defect. Then we added 770, and we could
see the channels come to life, the channels start
talking to us. And interesting, there’s a
lot of information in this. Because if you look at
the pattern of the gating, it’s not exactly the same. So we’re restoring the gating. But the mean open
times are different. There’s a lot of rich
information in here about how this
potentiator works. I don’t have time talk
to you about it today, but I’d love to talk
about it sometime. So we also evaluated
whether 770 could affect the overall
chloride transport from human cells isolated
from lungs of patients who had this mutation. These are very precious
samples, as you can imagine. There are only about 2,000
or 3,000 G551D patients in the world. And the number of lung
samples available you could count on one hand. Joe Pilewski at the
University of Pittsburgh graciously offered to
give us one of two vials that he had so that we
could try this test. We expanded the
cells in the lab, grew them up, and measured
the chloride transport across these cells
as a monolayer. And you can see the dramatic
effect that the compound had, increasing the amount
of chloride transport by about tenfold. So I want to show you what
these cells look like. So this is looking down
on these G551D cells. This is under resting
conditions, no drug added. Now these are cells grown on
a permeable support in a lab. So they don’t have all
the complex structures that Mike Welsh just described. We don’t have submucosal glands. I don’t know if there’s
squeegees going across on these things or not. But these are the surface cells. That do secrete fluid. They also secrete mucus. And you can see the
little movement. Those are little
tufts of cilia moving slowly and asynchronously
in this culture. And you can see the little
blobs on top of them. See that? All those? Those are little globs of mucus. And the cilia are
trying to get them off. Just get that off of me. But it can’t. The cilia are flattened
by the lack of fluid. And they cannot move those
mucous blobs off of them. And they’re just crushed by it. So we added 770 to these
cultures for three days to allow time for,
not only the drug to act– although the
drug acts very quickly– but to allow for the
fluid to start building up on the surface of the cells and
see the physiological effects of the drug. I’m going to show you that here. So after three days of 770,
came into the lab one morning. Actually, I wasn’t in
the lab at the time. Someone else came
and said, I think you have to come down and
see what’s on the microscope. So we went down there. And this is what the cells
were doing after three days. And it was kind of like
the hallelujah moment, because we realized we were
probably on the right track. This was about a year before
we had clinical data for 770. 770 was actually in
the clinic at the time. So if this hadn’t
worked, I don’t know what we would have
told our clinical colleagues about– well, we have
this little problem. But this really
confirmed that we had activity on that
channel, that the activity on the channel resulted in
fluid transport increases, and that those increases
also seemed to normalize the function of these surfaces. So you can see the cilia
waving in synchrony. And you can see the little
particles moving around. So we were on the right track. So about a year later, we did
start to get clinical data. The first clinical data
came in from this paper. This was a phase
2 study of VX-770. First author, Frank Accurso,
senior author, Bonnie Ramsay. What I love about
this paper, though, is it has authors from
the CF Foundation. It’s got authors from Vertex. It’s got all the
clinical investigators. It was a small enough study that
we were able to put, I think, every clinical
investigator on it. Each investigator had
only one or two patients. The number of– G551D are concentrated
in one center. So it was quite a feat to
actually do this study. And the results–
I remember sitting in the office, because the
study director from Vertex, Claudia Ordonez, called me up. I was sitting with Peter
Gruten, who was the project leader at the time at Vertex. And Claudia is usually
a very reserved person. She doesn’t get
excited about much. And we were kind of
just sitting there listening to her kind
of do her introduction. And she said, well, you know,
I think we hit a home run. And that was also another
moment that I will never forget when you hear stuff like that. And this is what she meant. So this is from the first part
of the study, 20 patients. We did a dose response
in 20 patients looking for changes in sweat chloride. This is the untreated group. 25 milligrams, 75, 150. Keep in mind, we had
no animal models. All of this was based on the HBE
assay and the in vitro potency. And based on that
and human PK, we estimated what the clinically
effective dose was. And we hit the dose
response in 20 patients. As a scientist, it doesn’t
get more amazing than this. I hope, even though this was
eight years ago, that you’re– I’m still as excited as I
was the first day looking at this slide. Because it opened the door. It was the first
study that showed we could correct CFTR function
in vivo with a small molecule. 20 patients told us that. And we had a dose response. The other thing that the study
wasn’t really designed to do was show us whether we were
impacting lung function. At the time we thought,
maybe these modulators would slow the rate of
progression of lung disease. We had no reason to
think that, perhaps, it would produce acute
improvements in lung function. But indeed, even in
these 20 patients we saw improvements
in lung function, again, in a dose response. 20 patients in 14 days. So an amazing study. And Bonnie will talk more about
what happened after this phase 2 study. But it was, I think, a
watershed moment for the field. So 770 was a pioneer. It still is. We learned a lot
from that compound. It went on to become
approved as Kalydeco in 2012. It’s been out there
long enough now that we do have long term data
looking at how it affects, not only the acute lung
function and other functions, but long term function. And it does slow the
progression of the disease. We know that unequivocally now. Key question. And since then, the
drug label has also been expanded to other
mutations and younger patients. The drug was recently
approved for young children as young as one year old. So there some one-year-olds
out there now getting Kalydeco. And what we’re going
to learn from them is going to be extraordinary in
terms of progression of disease and how that’s
affected by therapy. But we still needed to reach
the other 90% of patients. So how are we going to do that? We needed the correcter. Now, the correcter
was a bit more bulky in terms of getting from
the hit to the lead. We had to make a
lot more compounds. In fact, as you
see, the hit 768, compared to the
lead compound, 809– there’s almost no
similarity chemically. Our lead chemist on
the program thinks that maybe this is a different
activity from the hit, because it’s hard
to actually explain. So nonetheless, over those 3,000
compounds we did optimize it. And what’s interesting
here is this was not like this potentiator. The initial compound,
758, was not only very weakly active
in terms of potency, it was very weakly efficacious. It was not like that
first potentiator that was pretty efficacious. This barely read out in screen. It was barely out of
the noise of the screen. So we had a very
weak correcter we had to not only make more
potent but more efficacious. And we think that’s why it
took that much more chemistry. And fortunately the binding
mode of this compound and this mechanism did
allow us to increase both the potency and the efficacy. So I’m just going to show you
some critical preclinical data that led us to select
809 and to show that it worked in combination with 770. Remember that the whole premise
was to correct and potentiate. And the two together
would have more efficacy than either one alone. This was the preclinical
data that showed that. So on the left, I’m showing
you the Western blot. This is the protein maturation
that Mike described. You see that with 770. We don’t see any
maturation of Delta-F CFTR. So that’s not a correcter. VX-809, on the other
hand, does increase the amount of mature
protein relative to the immature protein. So we showed it was a
correcter by biochemistry. And then we showed
by function that when we added 809 to
the cells, we did get some chloride transport. Now, these are those human cells
from a Delta-F Delta-F patient. And we did get into the 20%
range of normal function. When we added the
two drugs together, we got further increase to about
28% in this assay of normal. So that was around our threshold
for going into the clinic. We wanted to definitely
be above 20% of normal for our combination therapy
based on the preclinical assay. And this got us to a
little bit above that. And it was a happy day when
we nominated this compound. This is September 19, 2007. We’re all sitting in
front of Vertex there. There are about 30
people in the picture. About seven or eight of them
are in this room right now. No particular order. There’s Bob Bell, sort of happy. That’s about as
happy as you see him. [LAUGHTER] There’s Preston Campbell
from CF Foundation. Chris Penland, head of research. Melissa Ashlock,
who was our project liaison with the Cystic
Fibrosis Foundation. She’s right there. And some really important
people here on the right. This is Viji Arumugam. She’s the first
chemist to make 770. That’s Sabine Hadida. She led the team. That’s Jason McCartney. He’s the chemist
who first made 809. They’re still with us today
making more compounds. But they were there
on that day too. There’s Fred Van Gore who
was the lead biologist for the program. Really instrumental in
setting up all those assays, picking those hits that
lead to the selection of the candidates that went
further into development. And he’s leading the research
program today at Vertex. There’s probably more
people I should talk about. Jeff [INAUDIBLE],,
you heard about him. He taught us how to
culture the hBE cells. And there are many others around
here that all have a story. I wish I had time to tell you
about every single one of them. And I was impressed that the
CF Foundation had the foresight to think we might be
nominating that candidate. We might want to take a picture. So let’s bring our flag. [LAUGHTER] That is planning. So this was the
clinical results. Now I’m showing you the phase
3 results for the combination of 809 and 770. This drug it had less activity
in vitro than 770 on the G551D. And it had less activity in
vivo on the acute measures of lung function. But it was active. And based on these
data, the drug was approved as Orkambi in
July of 2015, the first drug to treat the most
common mutation in CF. And it was a combination drug. So when we think
back about this, there’s a lot one
could say about getting to this stage and this point. One is that the odds
were not in our favor that this would happen. These were the first
two compounds that entered clinical development. Any compound that enters
clinical development has a 90% chance
of failing based on things that you just
couldn’t predict pre-clinically. So each one was a 1 in
10 chance of making it. Oops, let me go back. And then to have the
two combined as a drug– so that’s 10% times 10%. That’s a one in 100 chance
that the first two that you’re going to move into the
clinic are actually going to be combinable as drugs. So I think we had
some good fortune at our back in this program
as well to beat the odds and have the first two that went
into the clinic become drugs, one as a combination agent. And so that was the start. But it was only getting
us to 50% of patients. And we knew we wanted
to get more function out of the Delta-F protein anyway. There’s another story in here
that, again, I don’t have time to tell you today. But the first two drugs,
as much as they went in together and as much
as they beat the odds, they’re not perfect. VX-809 had some drug
interaction issues that make it difficult to
combine with other agents. And it also caused a chest
tightening in some patients that we still don’t
understand the mechanism of. It’s not based on
the CFTR activity. But if you have CF and
you experience some chest tightening, you don’t
like that, and especially if you have lower lung function. Doesn’t seem to be as
much a problem when they have high lung function. Fortunately, we had
a backup compound. And I remember
many conversations with Dr. Bell
saying, Bob, I think we ought to also make a backup. And he’s like, these things
are so expensive to make. Now you want two? But we convinced him. And we did make two. And 661 was now recently
approved this year with 770 as Symdeko. We think that drug
is going to be a better basis for
future therapies, including combinations. I’m not going to show you any
data on Symdeko, though, today. But we needed to get
to this other group. The efficacy of
Orkambi and Symdeko are not sufficient to give
them to patients who only have one F508-Del mutation. So we knew we needed to
go higher in efficacy. This shows you the
natural history data of different types
of CFTR mutations as a function of chloride
transport and CFTR function. And it confirms
to us that we want to move from no function, such
as mutations with Delta-F508, to carrier levels function,
which is greater than 80% of normal. With Kalydeco we’re at 50%. With Orkambi we’re at about 20%. We want to move everyone
up to this level. So we were going to
need another drug. The two drugs that
we had, or three, weren’t going to make it. And we kind of knew that even
around 2008 and ’09 when we started to get the first
clinical data for Orkambi that we would need more efficacy. So what were the approaches? Make new correctors or what? Start from scratch? We decided to try to add a
second correct on the first. There had been some
preclinical data from us and others that suggested
that you could correct CFTR in more than one way. And so we added compounds on
top of the first generation. And what I’m showing
you here is the result of about 10 years of
research, just fast forwarding through it. Showing you here,
the competitor. This is 770 and 661. So that’s the drug
that’s approved today for the Delta-F population
in our human lung assay, human cell transport assay. And a triple combination of two
correctors and the potentiator. And you can see remarkable
additional function from that third compound. We recently reported this
year initial clinical results from a phase 2 study of
this triple in patients with this difficult to treat
genotypic pair, F508-Del on one, minimal
function on the other. And the results were– I guess you could say they
exceeded expectations. We expected to see a big amount
of efficacy, a big increase. But we saw a really
big increase. And so here I’m showing you
the sweat chloride biomarker. We saw drops in sweat
chloride of minus 40. This is in about 80 patients. And in terms of lung function,
increases in lung function of more than 10% absolute. These are big changes. That corresponds to about
350 milliliters of air that someone can
breathe in one second. Patients can feel that. So that compound
and a sister called 445 are now in phase 3 studies. They’re fully enrolled. I think there is a mother
of someone who has someone in the study here today. She came to Vertex yesterday
and just gave us all a big hug. So we are on track to– hopefully, if those
compounds come through as they did in phase
2 through phase 3, we could have something
to treat 90% of patients not in five years, not in
three years, but perhaps in one or two years. So I’ve tried to tell
you a story that spanned 20 years in about 30 minutes. It started with the
discovery of the gene. We picked it up around 1998. Started to leverage
those learnings from the basic research. And today there are patients
benefiting from the drug. Through that time, millions
of compounds were screened. We’ve moved nine compounds
into clinical development. I didn’t talk about every
compound that went in, but nine have gone in. And there are 18,000
patients being treated today. And we have visibility on
treating the vast majority in the near future. And so those are
impressive numbers. But I think as everyone knows,
every patient is special. Every person with CF
has a unique story. And this is actually
the young lady that Francis showed on
his slide who has G551D. She actually lives in San Diego. And she was one of
the first patients to take part in the
ivacaftor studies. And she’s kept in touch with us. And this was an email she sent
to me a couple of years ago. There she is! [LAUGHTER] [APPLAUSE] That’s what I’m most proud of. So that concludes my talk. Thank you for everyone in
this room for the support. This is a wonderful family. Very privileged to have
been part of this project. And I hope I did some justice
to the hundreds of people that worked on these
compounds over 20 years. As everyone else has said–
and I’m the first to say it as well– this was not a
one person effort. We all did this together. I hope everyone’s proud of what
we were able to do together. Thank you. [APPLAUSE] Wow. I actually thought
that it was going to be tough to top Francis, but– so far we’ve heard about
remarkable work leading to the identification of
the gene for this disorder and it’s mutant alleles. Then small molecules that
address the basic biology and the basic patho-mechanism. But once you have
molecules, of course, you still have to get those. Approved and doing
proper clinical trials for rare diseases poses
remarkable challenges. And so our next talk and final
talk from an award recipient is Dr. Bonnie Ramsey from
the University of Washington where she is now the vice chair
of pediatrics and director of the Center for
Clinical and Translational Research at Seattle
Children’s Research Institute. She has helped to spearhead a
lot of the clinical trials that have led to some of the
widely used medicines for cystic fibrosis today. She received her
undergraduate degree at my alma mater,
Stanford University. And then she came to my
medical school alma mater of Harvard Medical School. She stayed on here for
her pediatrics residency at Children’s Hospital
before moving to Seattle. She has dedicated more
than 30 years of her career to the care of patients
with cystic fibrosis. And her research interests
have consistently been trying to get new medicines
approved for this disease. She’s received a
number of awards, including election to the
National Academy of Medicine. And I think what’s notable
about her career is that she’s spent more than
one decade trying to get new medicines
that treat the symptoms of cystic fibrosis. And now for more
than one decade, she has shifted
towards trying to get medicines that actually
address the root cause of cystic fibrosis. [APPLAUSE] Thank you. And you’ve already
learned an awful lot about cystic fibrosis. I have learned a lot
about cystic fibrosis. Now I’m going to talk
about the last leg, which is the incredible international
effort of patients, families, clinicians, and
researchers to conduct the clinical trials that
resulted in the drug products that Paul just discussed. I have had the honor
to be an eyewitness to this journey from bench
to bedside and to understand the challenges and
the joys of being a clinical translational
researcher. And I have received
grant funding from Vertex over the last decade. And also I want to thank both
the Cystic Fibrosis Foundation and the National
Institutes of Health for supporting me
for many years. As Paul mentioned, drug
development is not for sissies. It’s a high risk,
resource intense program. As he said, over
90% of new drugs that enter into human
trials don’t make it. And you can see from this
graph that most of them are lost in the phase
1 and phase 2 trials when you first start
testing efficacy and safety. But even if you’re
lucky, and you actually get an approved drug, it’s
extraordinarily expensive. In 2014, there was
an estimate that it would be over a billion dollars
for a new drug, 10 to 15 years. And then for the pivotal
trials, you often have to enroll
thousands of patients. You can imagine
how daunting that would be in an
orphan disease where you have 30,000 patients here
and about 70,000 patients in the world. Well, now I’m going to
do my Bob Bell talk. And he really was
a vision realized. He knew in the 1990s
when we had identified both the gene and the protein in
the CFTR function that it was– that was great. But that wasn’t going
to be helping patients unless we had a drug. And so he with the
CF Foundation started a bold, unprecedented
move, which was the therapeutics
development program. It had many arms. You’ve heard about it. It had basic research
groups looking at CFTR function and structure
and the high throughput screening that Paul described. And there were multiple both
academic and industry partners for that. And then the
venture philanthropy to try and de-risk and
bring small companies, such as Aurora, into the field. But it literally was hundreds of
companies that have approached the foundation over the years. And then the last arm was
the therapeutics development network. And this was a
clinical trials network that would facilitate
trial development and conduct and identify
appropriate outcome measures. This was all in the late 1990s. And as was mentioned,
it had been a case study for the
Harvard Business School. The CF Foundation has
really paved the way for other small
disease nonprofits to take drug discovery
into their own hands. So in 1997, everyone else
talked about getting a call from Bob Bell. I got a visit. He flew to Seattle, and
arrived in my office, and said, I want you to lead the
therapeutics development network. Now, my career, I had
worked with Arnold Smith, pediatric infectious disease
at Seattle Children’s, who actually we met originally
here at Boston Children’s. And with his mentorship
we looked at treatments for the pseudomonas
infections in CF lung disease, and particularly
inhaled antibiotics, inhaled tobramycin. And through that experience
we went through the whole drug development process. I learned about
protocol development, running multi-center trials,
and leading very large teams. And for that reason,
Bob said, well, now you can do it for
hundreds of studies. And so he convinced me
to take a sabbatical– I noted that Mike
took a sabbatical– and move away from just
looking at lung infections. That was the scariest
decision of my career. And I haven’t regretted
it for a minute. So I want to thank Bob for
giving me that opportunity and the CF Foundation. So why was I willing
to take this on? It’s because I knew the
CF Foundation already had developed the
key infrastructure for a successful
clinical trials network. And I have consulted lots
of other disease groups. And if they don’t have
the infrastructure, their networks
usually don’t work. First of all, the
CF care centers that Pam mentioned meant that
you had standardized care across hundreds of sites. And that was linked to the
national data registry that’s been around for decades. And it’s phenomenal. It has provided
the data we needed for sample size calculations
or developing outcome measures. And so it was critical. And then in the
past 10 to 15 years, a quality improvement
program has been established, which, again, brings
the variability down and makes it easier to see
changes due to a new therapy. So on top of this, the
therapeutic development was started. It was initiated in 1998. And it was really made up of
academic centers dedicated to studying new therapies. They were centers of excellence
and study contact, resource centers for
measuring biomarkers, such as microbiology
or sweat chloride. And then it had a centralized
data safety monitoring board down in the
University of Arizona to oversee the safety
of this program. There was a centralized
biostatistics and data management corps in Seattle. And the last one, the
consulting service– again, this was to
de-risk drug development for small companies. They would come to us, and
we would really talk to them. What is a clinical
development plan? And so it’s not just money. It’s also personnel and
expertise to provide. The goals of the
TDN were to conduct efficient and
successful studies, to protect the safety and
rights of study participants. Now, Paul and I are both going
to talk about the efficacy. We are going to spend a
lot of time on the safety. But that is absolutely
critical, that patients not be harmed during these studies. And so there was a huge emphasis
on safety and equitable access to clinical trials. We also performed
ancillary studies in developing outcome measures
as well as biomarkers. Now, fortunately the
original eight centers that formed the TDN had
had previous experience in therapeutic development. So we understood
already the importance of powerful partnerships
with industry. We had worked with
Genentech on Pulmozyme, which thinned mucous secretions,
and with pathogenesis, which was involved in
inhaled tobramycin. From this, we understood
the regulatory pathways for approval of
chronic therapies. And we also understood that you
had to have an adequate patient base and trained
personnel so that you could run multiple development
programs at the same time. Otherwise, you’re going
to be doing sequence, and that’s just
not going to work. And finally, we understood
the utility of a biomarker. Now, our first experience was
really in anti-infectives. And we learned that
what you see in a Petri dish with an
antibiotic, if you then looked at bacterial
killing in sputum– that if you had
99% killing, that would be reflected in a
significant increase, 10% to 15%, in FEV1. So we understood that you
had to have this platform to go from the lab,
to a biomarker, to a clinical outcome. So here’s sort of the history
of the expansion of the TDN. As I said, we started
with eight centers. And you can see
that it gradually increased over time until
now it’s about 90 centers. And also the US TDN
has worked very hard with Europe, and
Canada, and Australia to set up similar clinical
trials networks there. But an important part of this
was ensuring sustainability. And that has included a
robust training program, infrastructure grants so that
you don’t keep having turnover of staff, and also an
education program– and this was really the CF
Foundation, not the TDN– to engage families in research. And the first one of these
was called I am the key. So while we were getting the TDN
established from 1998 to about 2004, Paul was telling you
about how the high throughput screening was going on. And I know nothing about
high throughput screening. So I’m not going to
go through that again. But I had the privilege of being
on the advisory board, the CF Foundation advisory
board, where we would go to Aurora and then Vertex. And I would be sitting there. And all of a sudden, I could
see that they were really moving from hits on
to lead compounds, and that some drug was actually
going to come out of this. And I kept having
this vision in my mind that we were running a race. And the clinicians would be
the last leg of the race. And so my big fear was
that we would get the drug, but we would be unable to prove
that it was effective and safe. And then it wouldn’t
get to patients. So what could go wrong? Well, we could select
the wrong patients. Remember this is
precision medicine where you have to get the right
drug to the right patient. Second of all, we might not
have the right biomarker there is going to
take us from the lab to this clinical outcome
of improved lung function. Or we could just have
under-powered studies. Or we could be unable to enroll. Now, what we were focusing
on in that 2000 to 2005 time period was really
the first two issues. So first of all, this was
an international effort. And that was, we were going to
have to genotype the global CF population. You’re going to
have to know which of those class of
mutations everybody was so you could get them
in the right studies. From the day, I guess, after
the gene was discovered, Lap-Chee Tsui developed
a database in Toronto where everybody all over the
world could submit mutations. And that’s still open today. I think Joanna Romans leads it. But that was just mutations. We didn’t know whether
they were disease causing. And then in the last
10 years, there’s a program called the clinical
and functional translation of CFTR or CFTR2 which is
an international effort to assign disease liability. So what does disease
liability mean in lay terms? It means, does this mutation
actually cause the disease CF? And almost 90,000
patients have been put in this registry
from 43 countries. You can see in the dark blue. And there are other
countries that are planning to contribute. That means that 95% of the
patients with CF that are known have had their alleles
defined, their two mutations. I mean, I don’t know how
many other diseases where you have that type of coverage. So we know the
right patients are getting in the right studies. The second thing was bridging
between the laboratory findings and the clinical
outcome measures. So here on the left,
this was a picture that Paul showed you
where with VX-770, as you increase the dose,
the chloride flow went up. And these were in cells that
were taken from patients who had the G551D defect. So that was a clear measure
that some sort of CFTR pump was being turned on. Now, we also knew
from both studies like [INAUDIBLE] or
Pulmozyme that if you’re having a good drug
effect, you would get a rise in pulmonary function. And hopefully that
would be sustainable. But the issue was,
how are you going to link what you were seeing
in these bronchial epithelial cells to what you were
seeing in the patients? And we actually
spent quite a bit of time trying to
figure out what was going to be the best measure. And I won’t go into
all the possibilities. But in the end, we chose
the sweat chloride. Now, as Pam told you,
the sweat chloride has been around since the 1950s. It’s the diagnostic test
standard for CF care. Positive being greater than
60 millimole per liter. It’s available in every
CF center in the world. And it’s very reproducible
with low variance. And the good thing
about that is that means you need a smaller sample size
than if you have something that’s highly variable, such
as nasal potential difference, which is something
else we had looked at. We also know from those
genotype phenotype studies that have been described that the
less CFTR active you have, the higher your sweat chloride. So if you look on
the left hand side, you’ll see that patients with
pancreatic insufficiency, the most severe disease,
had high, about millimole per liter. As you get milder disease,
pancreatic sufficient, it starts going down. And then below 60, where you
get carriers, it’s even lower. So we knew that the therapeutic
goal for a CFTR modulator was to bring down
the sweat chloride. So now we have come to 2006. And ivacaftor has come
out of the laboratory. And it is the first in
human attempt at mutation specific therapy. And just to remind you– this has been reviewed before– it was a potentiator. And therefore it was
going to require– oops, I’m sorry. It was going to require
a special mutation where the protein folded and
was already at the surface. And all we had to do was open
the channel or potentiate. And that naturally occurs
in the G551 defect. And therefore,
these studies were done in that population
of which there are about 1,200 people in the
US and 2,000 to 3,000 worldwide. So what was going on at
the time of the first CF study of ivacaftor,
which happened in 2007 after the normal
volunteer studies? Well, here you see– it’s a picture of the TDN again. We were now up to
about 40 sites. We’d expanded. And that little black bar is
the first ivacaftor study. And if you look at
what was going on, there were 10 other CF clinical
development programs going on. This was everything
from pancreatic enzymes to anti-infectives. Then over about a
10 year period there were 13 ivacaftor
studies that were run through the network,
plus the networks in England, Canada,
and Australia. And at the same time, there
were another 110 studies or 52 other active CF clinical
development programs going on. I want you to think about that. This is a fixed population. And yet, we were
enrolling in all of these. And that only happened
because of thousands of patients, and clinicians,
and researchers that were willing to participate. Those I forgot to mention,
the other development programs that Paul mentioned. So now we come to
the phase 2 study. And Paul talked about how
unbelievably exciting it was. And I can only reiterate that. The sweat chloride
change you see here– if you’re a first
year medical student, and you’re learning
about drug development, this is, like, an unbelievable
model of a dose response. So you start at
the 25 milligrams. And you see the
numbers are tiny. It’s only 30 patients. And there was already a 33
millimole per liter drop. And then by the
time you get to 150, which was chosen for
the phase 3 trials, you’re at a remarkable drop of
minus 46 millimole per liter. If you think about that, if you
had a sweat chloride of 100, which many patients have,
you have now dropped down below the diagnostic cutoff. I as a clinician and
anybody else here in a room wouldn’t believe that
that would be possible. But that led very
quickly to rollover into the phase 3 pivotal
trial, which, again, is remarkably small numbers. You see it’s 160 patients. And this verified– I had the honor of being one of
the investigators leading this. It verified what
we saw in phase 2 and that the sweat
chloride again dropped, almost 50 millimole per liter. But you can see it was sustained
out to 48 weeks, absolutely rock stable. And the change in FEV1
was equally remarkable. If you look at 24 weeks– so the sample size
for this study was based on being able to get
at least a 4.5% improvement in FEV1 at that middle,
at the 24 week time point. And that was felt to be– that was really risky. You can see it doubled that. It was 10.6%. So we went way over the bar. So that was extremely
encouraging. And that ultimately led to
the approval of ivacaftor, or the commercial
name of Kalydeco. There was also a second
study in six to 12-year-olds that was part of
that, of the package. So it’s all wonderful
to see these numbers. And they are very,
very impressive. But I’d like to share my
personal story with my patient, Rick [INAUDIBLE]. Now, Rick was my
very first patient when I started at the
CF clinic in 1980. He was a teenager. And he only stayed in
our clinic until he graduated as a young adult
to the adult center in 1984. And we totally lost
contact for 30 years. Now, you have to remember,
in 1984, there was no gene. So I had no idea what his
genotype was, nor did he have any idea what his genotype was. But 30 years pass. And the ivacaftor results came
out in the New England Journal. So the Seattle Times, which
is the local paper in Seattle, called me up and said, we’d like
to talk to you for two minutes. But what we really want
is a patient story. And so I contacted Rick. And the Seattle Times
decided that they would tell a story
about our meeting after 30 years for the
first time, which is exactly what this picture is. I mean, it’s a little bit like
the Dr. Phil show or something. But it was– this
is the joy of being a translational
researcher, to actually see how a drug is changing
your patients’ lives. Now, Rick had become homebound. He was on intermittent oxygen.
He could no longer physically exercise. And within the month
of starting ivacaftor, he was able to go out and
play soccer with his two kids. He went back to working
full time as a teacher. And he said it was though he
could feel those little pumps just turning on in his lung. So it was really a
joyful experience. The other thing that I think
is so important to mention besides the human impact is that
ivacaftor and these other CFTR modulators are amazing
scientific tools. And once the drug is
approved there are so many other secondary outcome
measures that we can look at. And so fortunately the
CF Foundation– and I want to think Preston
Campbell for this, for having the
insight to say, we’re only going to bring a
drug out like this once. And so we have to see what
happens before and after. And therefore, for ivacaftor
the goal study was initiated. And we use that. And there was the
goal study in the US, but there have also been
multiple studies, particularly in Ireland where there is
a high prevalence of G551D, looking before and
after initiation at clinical outcomes,
at biomarkers like inflammatory markers,
microbiology, and looking at new mucociliary clearance. Now, the primary
results of this were published by Steve Rowe and
many investigators in 2014. But I can only have time to
show you a couple of them. So if you look here
on the left, what this is is this just shows
if you go out and prescribe the drug out in the community,
what’s the effect going to be? And this is called
effectiveness. And you can see there was
about a 7% improvement. Now, it’s not as
much as the 10%. But when you get out of a
controlled clinical trial and you get in the real world,
you expect some fall off. So this is actually
quite striking. And the sweat chloride
did the same thing. It just plummeted. But as far as sub
studies are concerned, here’s just one
example where we looked at mucociliary clearance. And using gamma scintigraphy
and using radio label particles before ivacaftor
and afterwards, they looked at how much clearance
is there in 60 minutes. And if you look beforehand,
about 10% clearance. One month after, it’s 20%. So you’ve doubled the
mucociliary clearance. It’s just what Mike
Welsh was showing you. You have had some
dramatic effect in that transport mechanism. And there are hundreds of
other studies like this that have been going on. So Paul summarized
this that ivacaftor was a wonderful story. But even with expanded
access, it was only 14% of the population. So that means you
still have 86% to go. But subsequently, we have had
the approval of both lumacaftor ivacaftor or Orkambi,
and tezacaftor ivacaftor or Symdeko, the first
in 2015 and the next in 2017. So that expanded to two
copies of Delta-F508. And then we’re
currently in phase 3 studies for the triple
combination, which will– hopefully in the
very near future will provide coverage
to 93% of the patients, because it includes everyone
with either one or two copies of Delta-F508. So this is an amazing story. And let me just summarize
the lessons learned. First of all, the
establishment of partnerships across industry, foundations,
academics, federal agencies, and patients and their families. Second of all, this concept
that you do the laboratory and clinical infrastructure
development at the same time and far in advance of the first
patient you’re going to enroll. I cannot tell you how many
disease groups I’ve talked to. And they think they have a drug. And then they want
to figure out how are you’re going to go
out and find the patients. It’s not going to work. So you have to do
them in parallel. And the third is what I’ve
just been talking about, is these new drugs open
us up for further science. I think it was Pam showed that
there were 2,500 investigators that were related to ivacaftor. This is a wonderful tool
for junior investigators to further their careers. So I talked about the Goal
study there on the left. That was for Kalydeco. The Prospect study as
a similar design, pre and post starting Orkambi. Promise is out and ready to
go with approval, hopefully, in the near future of
triple combination. And then finally, the CHEC
study looks at any CFTR module. They’re looking for
changes in sweat chloride. So now I would like to
thank a lot of people. The patients and
families worldwide who made this possible. You think about
it, 70,000 people and what has been accomplished. The CF caregivers and
researchers worldwide. The TDN investigators,
who have been incredible, and the research staff. The CF Foundation. And especially to my
family and friends. Many of them are here. I thank you very much
for your support. And this is the picture
Paul just showed. I hope other disease states– excuse me, diseases– can
have this kind of partnership. Which is represented here. It’ll go a long ways. Thank you very much. [APPLAUSE] So that’s it for the talks
from the award recipients. But we have one more
special talk lined up for this afternoon’s symposium. The Warren Alpert
Prize Symposium often ends with a sort of
forward looking science talk, typically from one
of its own faculty in the Harvard community. And this year, we’re delighted
to have Dr. Jay Rajagopal, who’s a professor of
medicine at Harvard Medical School with appointments at the
Harvard Stem Cell Institute, as well as at the Center for
Regenerative Medicine at Mass General Hospital. His CV is relatively
easy, because he was an undergraduate at Harvard,
a medical student at Harvard, a resident in the Harvard
system at Mass General, a clinical fellow
at Mass General, and also a post-doctoral
fellow at Harvard. We have a term for this,
which is called Preparation H, Preparation H-er. So he’s a bonafide
Preparation H-er. He runs a research
laboratory that’s focused on the developmental
biology of the lung. And we’ll hear about some
of the exciting research from his group. [APPLAUSE] [INAUDIBLE] promised he wouldn’t
call me Preparation H. But so much for that. It’s really been inspiring
to be here this afternoon, in part just to see
a few individuals of considerable intellect make
a real difference in the world. But I would also mention
that at scientific fora, it’s very rare to see
scientists get emotional. And I think that must speak
to the passion and deep desire that they had to do
good in the world. And how hard it is to satisfy
an unquenchable curiosity. And when I see Mike
Welsh there, who’s one of my mentors
at a distance– and I often feel
like I’m talking to a kid who’s just as excited
about the things he’s doing now as he was in the past. And it’s also, I think,
inspiring in the sense that you can see science as a
process of inevitable progress. Because there will always
be young people who are interested in
applying new biologies. So we’ve seen genetics. We’ve seen electrophysiology. We’ve seen
pharmaceutical science. And then we’ve seen clinical
translational science. And I’m here just to be a
representative today of two new biologies. One is computational
biology, and the other is my specialty,
developmental biology. So we happened on CF
more by happenstance. I am a pulmonologist. And so I’m interested
in the lung. But I’m also fundamentally
interested in regeneration. And the reason we
chose the lung to study was that it is a
regenerative organ. And it’s very simple. There are basal stem
cells at the bottom of this epithelium sitting
atop a basement membrane. Those basal stem cells
make more stem cells. But they can also differentiate
into secretory cells, which when activated,
makes things like mucus. And those secretory
cells, in turn, become ciliated cells that
sweep mucus out of your lungs, as Mike showed you. So when we started our
research program some years ago we thought we’d do the
simplest thing to a stem cell system, which is to
remove one cell at a time and see how it behaved. So the first cell that
we decided to remove was the basal stem cell. I won’t tell you how we did it. But we did it genetically. And we expected
the tissue to fail. If you don’t have a
bone marrow stem cell, you would think, ultimately, you
would get bone marrow fibrosis. But that’s not
what we saw at all. We had complete regeneration
in just a matter of weeks. And what was
happening actually– whenever developmental
biologists get confused, they do something
called lineage tracing. And the strange thing
about what was happening was that secretory cells
were replicating when we removed basal stem cells. Now, that’s as though you lost
your hematopoietic stem cells and some red blood
cell progenitors started replicating. It makes no intuitive
sense, because you’re not missing any red blood cells. So what were these replicating
secretory cells doing? Well, we lineage traced
them, using two transgenes. So we made secretory
cells glow green with GFP. And then we subsequently
ablated the basal stem cells. And what happened
here on the left– the entire paper is
really in this panel. On the left, you see
secretory cells in green. And you see basal cells,
the stem cells, in red. On the right, after
you ablate stem cells, you start seeing green
basal stem cells. So this was the first example of
a fully mature vertebrate cell de-differentiating
into a stem cell. But the remarkable thing
is all you had to do was remove the stem cells. And the body knew
how to remake itself. Well, we continued
to do this work. And I told you we were
excited about this system because it was simple. But now we’ve shown this strange
process where you go backward. And in other work, we showed
that you could actually go straight to a ciliated cell,
which was also a surprise. But that only happened
in the setting of injury. Now we stepped back
and we asked ourselves, what’s going on here? Well, there’s a couple
of possibilities. A basal stem cell
might be smart. And it might make a
secretory cell sometimes. When there’s an injury, it
might make a ciliated cell. Or there actually might
be two basal cells. And one actually might be making
secretory cells all the time. And another is sleeping
until there’s an injury, and then it decides to
make a ciliated cell. So what we wanted to know was,
were there different kinds of cells? So right about this time,
an amazing technology developed, which was called
single cell sequencing. So Dr. [INAUDIBLE] talked to
you about sequencing just one genome. Now we’re talking
about taking one cell and sequencing all
the genes that it’s expressing in a given moment. So this is really– you’ll
see this technology applied universally, I think,
in not too long. But this plot is very
complicated to explain. But each cell is
represented as a dot. And you have measured all the
genes that are in that cell. And cells are
clustered according to their aggregate
gene expression. So a blob basically means
you have a kind of cell. So when we looked at those
cells, we were pleased. Because we saw the basal
cells that we anticipated would be there. We saw the secretory cells. And we saw the ciliated cells. We were also pleased
because we saw rare cells, like the
neuroendocrine cell, which I haven’t told you about,
goblet cells, which are the ones that make mucus,
tough cells, which were only recognized
histologically, but there was no meaning ascribed
to them, and then the unknown cell, which I will
tell you about in short order. And then I also want to
say a wonderful thing– and we’ve heard about
collaboration so many times. Boston people are rather
provincial as [INAUDIBLE] pointed out. So we didn’t cross the
border into Canada. I’m not even sure
that’s legal anymore. [LAUGHTER] But we did cross the river. And I worked with my colleague
at the Broad Aviv Regev, and also a friend at Novartis,
Aaron Jaffe and his student Lindsay Plasschaert,
and also a biologist who was actually instrumental
in actually making this single cell technology,
Allon Klein and his student, Rapolas Zilonis. So it’s actually been a
very positive experience, because we had a
surprising finding. And it was incredibly
reassuring to be able to share that
pre-publication. So I’ll just do a
little show and tell. So they’re tough
cells, but it turns out that you could re-cluster
those tough cells and look for more subgroups. And when you do that, you
see mature tough cells. You also see tough
cell flavor one. You see tough cell flavor two. And then you actually see
another set of tough cells that actually also seem to
express the stem cell markers, suggesting that maybe tough
cells come from stem cells. And Dan worked very hard
to find antibodies that labeled each of these cells. And they did indeed seem to
be unique cell populations. But here was one
interesting thing. If you looked at these
two kinds of tough cells and compared their differential
gene expression in one, you had all the genes for
leukotriene synthesis, which as you know, is a
target in asthma therapy. And in the other tough cell,
you had a remarkable number of genes that were associate
with taste reception. Now, it doesn’t take
too long to figure out how these types of
cells might be involved in a disease like asthma. You have one cell, it’s
chemo sensing, perhaps the atmosphere. And you have another cell that’s
making inflammatory mediators. For now, it will
remain supposition. But similarly, goblet
cells seem to– the mucus producing cells
seem to come in two flavors. And for reasons that we
completely don’t understand, some of them are making lipases. So really very
mysterious things. And we also found that not
only could you go from a basal to a ciliated cell, but you
could take weird transits through different cells to go
from a basal to a club cell. So we were able to
computationally compare cells and say that there
were cell A and cell B and you could fill
in the intermediates with all these cells that had
mixed gene expression as they were changing, presumably,
from one cell type to another. And that allowed
the identification of this keratin positive cell. And amazingly, those
cells are actually in novel structures
in the airway. We called these hillocks
because the name Peyer’s patches was taken by someone
else for the intestine. But it’s quite remarkable. You can actually
find a new structure. So I told you we started
studying this epithelium because it was simple. And unfortunately,
it’s not so simple. It looks like it’s
rather complex, in fact, with multiple cell types
that we don’t understand. But where do all these
cell types come from? Some of the cell types,
like neuroendocrine cells, were thought to
self-renew themselves to be separate from the
entire rest of the epithelium. Well, I’ve hinted at it already. Being a developmental biologist,
I didn’t believe Allon and Aviv when they said cells
came from this point and moved to that point. We did what I showed
you we always do, which is lineage tracing. But we married
these two biologies. And what we did was
we lineage traced basal cells and single
cell sequenced them. Then we waited, and we
single cells sequenced mice with that lineage trace again
at 30 days and then at 60 days. And the result is that you could
see that lineage trace started in basal cells. And then that
lineage trace spread to all the other clusters,
essentially proving, for the most part, that all
of these new lineages that now look more like hematopoiesis
are arising from the basal stem cell. And I think that will
become important. Now I’ll tell you about the
unknown cell, which many of you have probably heard about. We’ve chosen to call
it the ionocyte. And the strange thing
about this cell type is when we superimpose CFTR
expression on these cluster maps we found that they were
only very rare cells that express the CFTR gene, which was
a surprise, because we thought it was supposed to
be on ciliated cells. That would be the
textbook version. Those cells were also associated
with a specific transcription factor Foxi1 And Dan, again,
worked incredibly hard to make sure that his CFTR
antibodies were correct, and found one with
no background, and was able to show that a
Foxi1 positive cell had this little nubbin of
CFTR at its apex, and that the other cells
surrounding it didn’t. And those same
Foxi1 positive cells also had subunits of
a vacuole or ATPase, making it at least seem more
possible that this could indeed be the case. And then if you
looked at other places in the respiratory tract
epithelium in the nose, you saw the same kinds of
cells with CFTR expression. We still didn’t believe
this, because it just seemed contrary to the
prevailing hypothesis. So we separated cells into
ciliated cells and ionocytes genetically, and then
did more deep sequencing, and also used RT PCR,
which is more quantitative, to look for the
cystic fibrosis gene. But we could only find
the cystic fibrosis gene in ionocytes. We just simply couldn’t
detect it in ciliated cells. It doesn’t mean it’s not there. But if it’s there, it’s
there in tiny amounts. And the ionocytes are enriched
at least 100-fold for CFTR. And then Dan did
an experiment where he got a Foxi1 knockout mouse. And what you could see is that
the CFTR protein decreases substantially when you knock
out the ionocyte transcription factor, again lending credence
to this surprising finding. And then this was one of
those fun experiments. One day Dan called me into
the tissue culture room after he had already called
all the other post-doctoral fellows. I’m always the last to know. But he showed me
this pile of plates. And there was wells
and tons of wells. And he said, which ones
of these looked different? And so I said, number
eight, number 32, number 19. And it turned out
that he whipped out his Excel spreadsheet. And every single one of
these funny looking cells– I’m not sure if you can tell,
but these wells have a sheen on their surface
that these don’t. If you were looking at them
yourself, they’re binary. It’s very easy to
pick out these cells. And those are all the
Foxi1 knockout cells, the ionocyte gene knockout. And then if you do optical
coherence tomography, as Steve Rowe did, you see an
increased airway surface fluid layer with increased
reflectance suggestive of increased amounts of
mucus, which is perhaps what’s explaining that gene. Again, lending
credence to the idea that this rare cell
type is important. Then we looked in the human. And since the– there are
over 250 CFTR antibodies. So it’s very scary
to use them actually. And many of them generate
a whole ton of background. So we just decided
to go straight to RNA in situ
hybridization in human with an extremely
sensitive technique that in certain settings can
pick up single molecules. And what we found was that
the gene for Foxi1 and CFTR co-localized in rare
cells, approximately 1% of those cells. That seems very hard
to explain by accident. Now, you’ll note– and we
looked very hard– sometimes there is an occasional
dot that looks like there could be a little
CFTR out there somewhere. But you see those dots
on the control also. So I suspect that they’re
background, but can’t be sure. Then we looked at single
cell sequencing of humans. And what we found was
that these ionocytes seemed to occur as well. And Allon and Lindsay
and Aaron also did single cell sequencing
of air liquid interface and found the same
finding, that there appeared to be human ionocytes. And then they did this. Allon and Aaron kindly lent
me some of their slides to show they did this very
nice experiment where they put the Foxi1 gene into air
liquid interface cultures and then single
cell sequenced them. And it looked like they
started acquiring ionocyte gene profiles. So it’s really
elegant demonstration. Then Allon and Aaron
and Lindsay took a bunch of different patients
and grew out their air liquid interface cultures. And it turns out that
different patients had different numbers of ionocytes. And when you looked at
the short circuit current, the number of ionocytes
correlated to the short circuit current. This is just
correlation, not proof. But it’s intriguing. And the other thing
that they noted was that, if you look at
the number of ciliated cells in those air liquid
interface cultures, there was a relatively
poor correlation. Again, lending
credence to the idea that the locus of CFTR
action is in the ionocytes. And then why did we
call it the ionocyte? Think that’s actually important. And the reason that we
called it the ionocyte was not because we
invented this name, but because there was a
lot of biology out there. That, first of all,
in the frog skin, there was also a
slimy structure. There was a cell called a
mitochondrion rich cell. In fish gills, there was a
cell called a Foxi1 ionocyte. Foxi1 led us down that path. There’s things called
chloride secretory cells that sound similar, light
stained cells, flask cells. And then Foxi1 controls what are
called proton-secreting cells in the distal collecting tubule
and in the endolymphatic sac. So it was starting to sound
difficult not to believe this. Then it turns out that after
talking to all my colleagues in the CF field, including
Mike, people in the CF field had been wondering about
these things called hot cells or jackpots cells
that express tons of CFTR. And it looks like, indeed,
maybe those jackpot cells were ionocytes. And I just do want
to mention this, because scientists
are really amazing. And the amount of work that’s
been done in fish ionocytes is extraordinary. I spent almost an
entire day trying to find all the
species of fish whose ionocytes have been described. This one is from– this is actually the
nostrils of a sea skate. And this one is a catfish
that occurs in an Iranian dam. But there’s something
interesting about some of these. Salmon is anadromous. That is to say, it
lives in the ocean and then swims into
freshwater streams to spawn. And you can imagine that
diffusive water movements that that encumbers. There are other fish,
like eels, that actually live in freshwater and then go
to places like the Sargasso Sea to spawn. And they’re called catadromous. And then there are
other organisms, like the milkfish,
that have to contend with constant changes
in ionic environment, because they live in
estuarine conditions. In all these cases– oh, and actually
a really fun one that I like a lot, which
is the mud skipper. The mud skipper can actually
come out of the water onto land. And it probably uses its
ionocytes to protonate ammonia, so it could excrete that way. But it’s just too
much not to start believing that this cell
could do something important. And then I found this paper
after a day of researching, which is ionocytes have been
looked at in some species, in this case the alewife, which
has a salmon-like behavior and anadromous. And they’ve looked at
ionocytes when they’re in freshwater or saltwater. And those ionocytes
seem to change. They didn’t do lineage
tracing, so maybe they’re different ionocytes. But the other possibility
is that they change. And look at the gene product
that those ionocytes express when they’re in saltwater. Now, again, I don’t know if
their antibody is right, yada, yada, yada. But it does seem like there
must be some conservation across species. So there’s bench to bedside,
and then maybe there’s streamside biology that
we’ll also be able to do. And then just to tell
you a little bit more about how powerful this
is beyond cystic fibrosis, what about all the
other lung diseases that we’re interested in? Dan decided to take
cells from the proximal and the distal trachea and
single cell sequence them. And he found that just
the secretory cells were different
proximally and distally. But they were different
in interesting ways. Like the distribution of genes
that made you sensitive to asthma were different, for
example, the IL-13 receptor. And IL-13 has been
targeted in asthma. So look what happens if you
take proximal airway epithelia. There aren’t that
many mucus cells. But distal airway epithelium
has quite a few mucus cells. If you then add
IL-13 proximally, you get some mucus cells. But look what happens when
you take the distal epithelium and you add IL-13. Essentially your entire
airway becomes mucous-like. So there’s also heterogeneity
along the proximal distal axis of the lung. And then he did a
very nice experiment. He took GWAS genes for asthma
and superimposed that gene expression on the t-SNE map. And look at this gene, the
receptor for rhinovirus-C that’s a GWAS hit for
asthma is expressed only in conciliated cells. Converting a
potential genetic hit into a genetic abnormality
localized within a particular kind of cell type. So the predisposition
to asthma from this gene is probably because rhinovirus
infects ciliated cells. So it’s now easier to
mechanistically go after this. And then I’ll say that we’ve
now for quite a while now been able to make patients’ skin
into respiratory epithelium. And we’ve also been able
to grow up stem cells just from CF patients’ cough. So we should have abundant
amount– they’re not perfect. I don’t know what their
ionocytes are doing. Their CFTR currents
aren’t great. But it becomes
tractable problem. I did just want to say once
again that it was absolutely terrific to work with Aaron and
Lindsay and Allon and Rapolas. Because as you might discern,
some of these findings were very, very surprising. And we really didn’t believe
them for a long time. And it was incredibly reassuring
to have rigorous friends who found nearly the same results. And also it was
nice to have Steve Rowe of the University
of Alabama help us, along with his
post-doc Susan Birket. And Martin and Hermann from CFF. So we had real
experts helping us with the physiology part
of these experiments. And then another collaborator
was John Engelhardt. And this is the amazing thing. There is nothing
new under the sun. John wrote a review
article in 1998, where he put this
MR on this cell. That MR stands for
mitochondria-rich cell. So he had hypothesized that
these hot or jackpot cells would exist, and that they would
have some analogy to Xenopus mitochondrial-rich cells. So in some sense,
all we did was found what John knew would be there
or guessed would be there. And then I’ll just
say a few more things. Even if ionocytes
exist, that doesn’t mean that loss or
gain of ionocytes will give you see
CF-like disease. In fact, there are
patients with mutations of Foxi1 who have distal
renal tubular acidosis and deafness based on
their endolymphatic ducts. But there is no homozygous
human deletion of Foxi1, maybe because it’s lethal. But I don’t want us to conflate
that this is the CF cell. It seems to be the
cell that’s expressing the abundance of CFTR in
the airway epithelium, and so very unlikely not
to be very relevant for CF. And then what are the open
questions that this generates? I think the biggest
one is in gene therapy. Do we have to make sure
ionocyte numbers are restored? Will over expressing CFTR in
ciliated cells be beneficial? It might. Just because it’s
in the cell type where it doesn’t naturally
occur might be fine. Which cell type must be
target for a durable cure? That lineage tracing
test I showed you does imply that we have to get gene
therapy into the basal stem cells, because otherwise the
cells will just turn over. So that I feel relatively
confident in saying. And then how do
we make ionocytes? What’s the rare cell code? How do you make all
these different cells? And can we exploit
that phenomenon I showed you originally
to convert one cell type into another? And then just the
pathophysiologic questions, like what is the role
of the ciliated cell? It could actually
have a huge role even though it’s
not expressing CFTR. And what is the rare
ionocyte actually doing? Is it actually moving that much
ions with all its mitochondria? Or is it acting as a sensor? Or is it talking to
its ciliated cell colleagues to make
them do something? I think these are
just open questions. But from my being completely
ignorant about anything about CF, it seems
to me like there’s something fundamental
that we don’t understand, very fundamental. And then I’ll just return
to that theme that I had, establishing a cellular
narrative for CF. I think one of the
things we need to do is understand cell partitioning
of these gene products. It’s unclear where CFTR
is definitively now. But all the vacuole
or ATPases, they also might be distributed
in different cells. We also have the capacity now
to make cell type specific gene modulation. So we can make mice that
target just the ionocytes or the ciliated cell and
remove gene products from them one at a time. And I think we should
do that in animal models and in human systems. And then we really should
go look at patients again, CF patients, and
look at their cells. What are their ionocytes doing? Maybe they’re doing something
very interesting in an attempt to compensate. Maybe we can help them. Maybe we can make more of them. All of these become
open questions. And then I’d just like to end
by thanking my own lab, who have influenced all of our thinking. And Aviv, who has been a
wonderful collaborator. This was really equal. I don’t think a
developmental biologist or a computational
biologist could have done this alone at all. And then the real
privilege is working with absolutely
wonderful young people. And Dan Montoro is
my graduate student. Adam Haber and Moshe Biton are
Aviv’s post-doctoral fellows. And they really carried
us through this. All they got from
me was disbelief. But they slowly just
provided the experiments that ultimately
convinced us that this had to be a real cell. And so I’ll end with that
and just say what a privilege it’s been to be
able to talk to you. [APPLAUSE] Wow. OK. Well, I think this afternoon
has been extraordinary. We’ve been informed. We’ve been entertained. We’ve been inspired. And we’ve been moved. I think it has been a
wonderful celebration of the great possibilities
of modern biomedicine. I hope it won’t be our last. I want to thank
[INAUDIBLE] for doing such a great job in introducing
and moderating the speakers. I want to thank Dr. Davis, Dr.
Rajagopal for their keynote. And then finally, before
we end this evening, please join me in applauding
once more the tremendous work of our honorees. Thank you very much. [APPLAUSE]

1 comment

Leave a Reply

(*) Required, Your email will not be published