Skin Stem Cells: Their Biology and Promise for Medicine

Transcript of Part 2: Tapping the Potential of Adult Stem Cells

00:00:14.25	I'm Elaine Fuchs.
00:00:16.10	I'm a professor at The Rockefeller University in New York City.
00:00:21.04	I'm also an investigator of the Howard Hughes Medical Institute.
00:00:25.20	And today I'd like to tell you a little bit about tapping the potential of adult stem cells.
00:00:32.12	Every tissue of our body has to replenish dying cells, dead cells,
00:00:39.01	that through wear and tear are lost from our tissues.
00:00:43.18	50-70 billion cells die every day in our body, and they have to be replaced.
00:00:50.28	How does that happen?
00:00:52.09	Our body must repair its tissues, as well, when damaged.
00:00:56.26	How does that happen?
00:01:00.13	Every tissue of our body has reservoirs of stem cells for homeostasis and wound repair.
00:01:07.05	So, it's the stem cells that replace these 50-70 billion di... billion cells that die
00:01:13.13	every day in our body, and that also repair our wounds.
00:01:20.11	Some tissues turnover rapidly and have many stem cells.
00:01:24.24	Those are... such as hematopoietic stem cells, which regenerate the red blood cells and immune cells.
00:01:30.28	Or skin stem cells, which regenerate the epidermis and the hair follicles and our sweat glands.
00:01:37.02	Or intestinal stem cells, that rejuvenate the food adsorptive cells of our body.
00:01:44.00	Other tissues have very little regenerative capacity.
00:01:47.25	Our central and peripheral nervous system, for instance, pancreatic tissue, or cardiac tissue.
00:01:55.24	So, what are the characteristics of tissue stem cells that exist in adult tissues?
00:02:02.04	These cells have the capacity to divide, to produce more stem cells
00:02:06.21	-- that's a process known as self-renewal --
00:02:10.02	and they produce a differentiated tissue, such as the epidermis or the hair.
00:02:15.04	And they do so long-term.
00:02:19.09	These cells are relatively immature compared to their progeny, compared to the tissue cells.
00:02:27.19	They're typically multipotent; they're able to produce a subset of cell lineages.
00:02:33.14	Examples: hematopoietic stem cells can make the blood and immune cells;
00:02:38.13	skin stem cells can make the epidermis and hair.
00:02:43.23	It's only the fertilized egg that has unlimited potential, that can make all the cells of our body.
00:02:51.15	Tissue stem cells tend to be more restricted in what they can do.
00:02:55.28	And to understand why, we need to study the biology of adult tissue stem cells.
00:03:02.14	So, my laboratory studies the skin and skin stem cells.
00:03:06.25	And I've always thought that nature has clearly had a lot more fun and fancy in creating body surfaces
00:03:11.15	than she has in creating any of the ugly organs that most scientists work on.
00:03:18.13	My group works on the stem cells of the skin.
00:03:23.16	So, where do adult skin stem cells come from?
00:03:27.14	They began in early embryonic development, shortly after gastrulation, when the embryo
00:03:37.20	effectively turns inside-out and begins to develop into the animal.
00:03:44.10	The skin begins as a single layer of cells that are on the surface of the embryo.
00:03:51.03	Those are the cells that are going to give rise to the stratified epidermis that you see here.
00:03:58.15	They'll also generate the hair follicles.
00:04:01.04	They'll also generate the corneal surface of our eye, the sweat glands, and the mammary glands.
00:04:07.15	All of these cells come from this single layer of multipotent progenitor cells,
00:04:15.10	that are the origins of the stem cells of our adult skin.
00:04:20.07	So, skin stem cells provide the nearly endless supply of cells that replenish the epidermis
00:04:28.04	and the hairs of our body.
00:04:30.21	Can we isolate them?
00:04:32.02	Can we coax them to become hair as well as epidermis?
00:04:37.21	What value do they have in regenerative medicine?
00:04:42.12	Back when I was a graduate student, I was working on bacteria.
00:04:45.26	And I heard a lecture by Howard Green, who at the time was at MIT as a research scientist.
00:04:53.20	And Howard had developed a method where he could take a piece of human skin,
00:04:58.11	and put the cells into culture, and be able to propagate them endlessly, without losing their ability
00:05:04.20	to make skin.
00:05:06.08	I was just fascinated by his lecture.
00:05:09.18	And I knew then that I wanted to become a stem cell biologist.
00:05:18.09	So, here are examples, then, of the stem cells that are growing in culture.
00:05:24.20	And you can see by the dark chromosomes that these cells are actively dividing.
00:05:30.15	And they can be propagated, as I mentioned, endlessly.
00:05:34.20	Below, what you see is a three-dimensional, so-called organoid culture, or a tissue,
00:05:43.06	where we've reconstituted the epidermis starting from scratch using these cultured
00:05:48.05	human epidermal stem cells.
00:05:50.11	It looks pretty good compared to normal human skin epidermis.
00:05:57.04	So, Howard Green went on to develop this technology for the treatment of burn patients,
00:06:03.04	where he could take a small piece of the patient's good skin, propagate those cells in culture,
00:06:09.00	and make sheets -- hundreds of sheets of epidermal cells -- that then could be grafted
00:06:15.06	in the patient's bad skin.
00:06:16.27	So, the epidermal stem cells could be grown for months in a dish, and forever when grafted
00:06:23.16	onto a patient.
00:06:25.00	And this shows you an example of some of the grafted skin from one of these patients.
00:06:30.24	The patients that Howard Green treated successfully with these cultured human epidermal stem cells
00:06:37.05	from the patient covered up to 95% of the patient's burned skin,
00:06:43.16	with only 5% of the patient having unburned skin.
00:06:47.07	And from that stem cells could be cultured.
00:06:50.08	A remarkable advance.
00:06:53.17	After a year, the skin epidermis still looked like the normal skin epidermis would.
00:07:00.27	And it didn't show signs of cancer or deleterious effects that might give us an indication that
00:07:07.06	all of that culturing is a bad thing for these human epidermal stem cells.
00:07:13.04	Michele De Luca and Graziella Pellegrini, who were then studying in Howard Green's laboratory
00:07:22.06	and then started up their own labs in Italy, adapted the pioneering culturing procedures
00:07:28.23	that Howard Green had performed, in this case, to culture cornea or corneal epithelial cells
00:07:36.22	from a patient's stem cells of the cornea.
00:07:40.16	So, in this case, the patient had an industrial accident, where one of the eyes
00:07:46.12	-- the eye that you see here -- was blinded.
00:07:50.02	By taking stem cells from the healthy eye, culturing those cells, and producing a sheet
00:07:57.00	of corneal epithelial cells, it was possible, then, to graft it onto the blind eye.
00:08:02.25	And here is the result of the stem cell therapy: a restored corneal epithelium that effectively
00:08:11.00	restored the vision in this eye to this patient.
00:08:14.22	So, let's take a look at ten years later, after those experiments with the cornea were performed.
00:08:24.22	Just in the last month, there was a very successful report of whole body regenerative medicine
00:08:34.02	from disease-corrected epidermal stem cells.
00:08:38.01	And in this case, the patient, a seven-year-old boy, suffered from a genetic disorder known
00:08:45.15	as junctional epidermolysis bullosa.
00:08:49.05	This genetic disorder is due to recessive genetic mutations that can be either in the
00:08:55.18	laminin5, an extracellular matrix protein that's important for the epidermis
00:09:01.02	to adhere to the underlying dermis, or mutations in integrins alpha-6 or beta-4,
00:09:08.05	which are the integrins in the epidermal stem cells that are necessary to adhere the epidermis
00:09:14.16	to the underlying dermis.
00:09:16.15	And this particular patient had a disorder, or genetic lesion, in the laminin5 chain.
00:09:22.13	And as a consequence, the epidermis did not adhere, and effectively sluffed off
00:09:30.06	from the patient's body with the slightest of mechanical trauma.
00:09:34.10	The patient was near death at the point at which Michele De Luca, my colleague in Rome,
00:09:42.15	took on this task of working with German clinicians in order to be able to take a small sample
00:09:51.01	of the patient's cells, junctional epidermolysis bullosa-defective keratinocytes;
00:09:58.25	used gene therapy to be able to put back the normal laminin chain and to be able to excise
00:10:05.00	the genetic mutation of the patient; and effectively, at that point, then, generated healthy
00:10:12.20	human epidermal stem cells that could be maintained and propagated for generating sheets of epidermis
00:10:19.23	that were then grafted onto this patient.
00:10:23.11	So, this gene therapy of epidermal stem cells required only a very few number of stem cells
00:10:34.09	in order to be able to replenish, or repopulate, the entire epidermis of this patient
00:10:40.13	using whole-body regenerative medicine from disease-corrected epidermal stem cells.
00:10:45.18	And here is the boy today, basically able to go out and play in a way that he was
00:10:54.23	barely able to survive at the time of the therapy.
00:11:00.08	So, even a year ago, scientists did not think, or were concerned, that it might not be possible
00:11:10.15	to be able to produce normal, healthy epidermal stem cells from gene editing of the mutation
00:11:21.11	out of the stem cells.
00:11:24.02	And it also was doubtful that whole-body regenerative medicine would be possible to correct
00:11:31.04	the genetic basis of this disease, or to correct this genetic disease.
00:11:35.20	The patient's other cells still harbor that gene defect, but what the important aspect
00:11:42.26	of this is is that the patient -- the patient's skin epidermis -- is now corrected.
00:11:50.05	And this boy may be able to at least live a relatively normal life.
00:12:03.01	So, what are, then, the various different associated therapies that have followed
00:12:14.11	the culturing of epidermal stem cells that were first performed in 1975?
00:12:22.10	And now we have the correction of third-degree burns with epidermal stem cells, and the correction
00:12:29.24	of junctional epidermolysis bullosa with epidermal stem cells that have been gene edited.
00:12:36.04	In 1985, we had corneal epithelial stem cells to correct blindness from burns and from chemicals.
00:12:45.03	In 1992, Peterson and Bissell, scientists at Lawrence Berkeley Laboratory, and also
00:12:53.10	in Denmark, were able to produce milk-producing glands from human mammary epithelial stem cells
00:13:02.21	cultured in the laboratory.
00:13:05.01	And these have been enormously helpful, not only to study mammalian mammary biology,
00:13:11.11	but also to study breast cancer.
00:13:14.08	In 2013, Sato and Clevers made the impressive advance to be able to produce organoid cultures
00:13:25.11	from simple epithelia, from intestinal epithelial, lung epithelial, stomach, and liver stem cells.
00:13:32.09	These were, again, advances that took years for scientists to make with, again, the realization
00:13:39.20	that what's important is to understand enough about the biology of the stem cell to be able
00:13:45.18	to know about their environment and basically recreate the environment that that stem cell
00:13:51.22	normally has in its normal tissue, and recreate that in a culture dish.
00:13:56.27	It was determined early on, by Howard Green and others, for epidermal stem cells.
00:14:02.16	Many years later, it's now become possible to culture many different types of adult stem cells.
00:14:09.06	And so, here, by the use of the ability to culture, for instance, lung stem cells,
00:14:15.08	it was also possible in... again, in just the last year in the Netherlands, to be able to
00:14:21.02	treat a patient with cystic fibrosis by gene editing mini-lung organoid cultures grown
00:14:29.22	from the patient, and... and basically allow the patient to... to again be able to breathe properly,
00:14:41.05	and... and promise for the future.
00:14:47.17	So, where are we in 2018 with our knowledge about the biology of tissue stem cells?
00:14:56.13	Because it's the biology that is going to drive future advances in the clinics,
00:15:02.24	and also in the pharmaceutical and biotechnology industry.
00:15:06.11	So, what we know is that stem cells of adult tissues typically reside in niches, in defined
00:15:14.02	microenvironments.
00:15:15.07	That's true for the hair follicle, for the intestine, and for the hematopoietic system,
00:15:22.00	the three stem cells for which we know most about the stem cells and their niches.
00:15:30.23	When does a tissue stem cell decide, am I gonna make tissue, or am I going to be dormant,
00:15:38.22	or quiescent?
00:15:40.06	It's the interactions between the stem cell and its microenvironment, or the niche stem-cell interactions,
00:15:46.04	that make that decision.
00:15:48.27	When a stem cell is not making tissue and exists, for instance, in the skin, it basically
00:15:54.24	is receiving inhibitory signals from the microenvironment,
00:15:59.01	telling it to stay cool, be quiet, be dormant, and not make tissue.
00:16:05.21	But if a stem cell needs to make tissue, now there have to be interacting cues that are
00:16:12.26	positive-acting cues that are now going to override those inhibitory cues.
00:16:18.22	And at that point, those activated stem cells then go on to make short-lived progenitors,
00:16:24.11	which then go on to produce the bulk of the tissue.
00:16:27.13	And in the case of the skin, it would be, for instance, the epidermis or the hair follicle.
00:16:33.07	How many and how often stem cells are active depends upon the needs of the particular tissue.
00:16:40.22	So, we know a lot about the skin stem cells from my laboratory and other laboratories' research
00:16:47.01	over the last several decades now.
00:16:51.22	And we know that the tissue stem cells reside at the interface between the epidermis
00:16:57.09	and the dermis.
00:17:00.03	But they're defined in their task and also their program of gene expression
00:17:05.24	by their local niches.
00:17:06.25	So, the epidermal stem cells that reside in the innermost layer of the epidermis are in
00:17:12.14	a niche microenvironment that's very different from that of the hair follicle stem cells
00:17:17.23	that are located at the very base of the non-cycling portion of the hair follicle.
00:17:25.04	And that stem cell-niche interactions are what defines the stem cells in each of these
00:17:33.07	niches... niches, and tells the stem cells of the epidermis to make epidermis
00:17:38.24	-- produce our barrier -- whereas stem cells from the hair follicle to make hair.
00:17:45.23	So, my laboratory has studied the hair follicle for a number of years now.
00:17:51.13	It's really the perfect system to understand a basic biological problem.
00:17:56.07	And that is how stem cells transition between quiescence and tissue regeneration.
00:18:02.02	This is a problem that's faced by all the stem cells of our adult tissues.
00:18:06.26	And we've learned a lot about the various different interactions that are necessary.
00:18:11.23	We know that these red arrows are basically inter... inter-niche cells that are
00:18:17.15	sending out a signal called BMP that tells these stem cells, in purple, to be quiescent.
00:18:25.12	We know that at the base of the stem cell niche, there are signals -- new signals, activating signals --
00:18:30.14	called Wnts, which are produced by the hair follicle stem cells themselves,
00:18:36.25	and BMP inhibitors produced by the underlying green cell,
00:18:41.13	basically a pocket of specialized mesenchymal cells called dermal papilla cells.
00:18:45.27	And these are the pro-activating signals that are necessary to take the stem cells
00:18:51.24	out of their quiescent state and coax them to make hair.
00:18:55.18	So, when the threshold levels of activating signals overcome the inhibitory BMP signal,
00:19:02.26	then, the stem cells are basically activated.
00:19:06.25	They start to make those short-lived progeny.
00:19:09.16	And now the short-lived progeny may get a new signal, this one called sonic hedgehog.
00:19:15.02	Sonic hedgehog then acts in two ways, which we learned from conditional knockout
00:19:23.09	of both the receptor and the ligand.
00:19:26.02	What we learned is that sonic hedgehog first acts to stimulate the specialized mesenchymal cells,
00:19:32.17	the dermal papilla cells, to produce more BMP inhibitors and also pro-activating FGFs.
00:19:40.08	Then, several days later, sonic hedgehog acts on the quiescent stem cells, those purple cells,
00:19:48.23	to be able to restock the stem cells utilized in the course of tissue regeneration,
00:19:55.05	but also to produce the shaft that then grows downward, pushing the signaling center
00:20:02.21	away from the stem cell niche.
00:20:04.18	And now the stem cells return to quiescence, but the short-lived progeny produce hair
00:20:11.23	and the channel that the hair follicle sits in.
00:20:15.06	And so, in this way, we understand very well about how the hair cycle works because
00:20:23.28	once the hair cycle is begun and gets going, it's essentially the short-lived progenitors
00:20:30.26	that then do the bulk of making tissue, in this case, the hair in its channel.
00:20:35.19	So, the short-lived progenitors are short-lived.
00:20:38.10	Eventually, they lose their proliferative capacity, and undergo apoptosis and differentiation.
00:20:46.02	And now the system returns back to normal again, to reset the hair cycle.
00:20:52.15	So, what does it take to make a hair follicle?
00:20:56.24	I talked about the importance of the Wnt signal, but many years ago my laboratory activated
00:21:04.17	the Wnt signal... at the time we didn't even know it was a Wnt signal...
00:21:10.06	we activated its interacting partner known as beta-catenin.
00:21:15.04	And when we activated beta-catenin in the mouse skin, what we got is this super furry mouse,
00:21:20.11	relative to its brother next to it.
00:21:24.03	These mice ended up producing far more hair follicles than normal.
00:21:29.13	Initially, we were very excited by that.
00:21:32.19	But the downside of this is is that those mice develop tumors over time.
00:21:38.02	Obviously, not yet ready for primetime.
00:21:41.24	But it told us a very important thing, and that is that Wnt signaling is important in
00:21:47.06	dictating these unspecified progenitors to make a hair follicle, not to make epidermis.
00:21:54.14	So, when taken from their niche and cultured, the green stem cells that you see here,
00:22:01.05	of the hair follicle, turn out to acquire plasticity.
00:22:05.00	We can culture these stem cells much the same way that Howard Green cultured epidermal stem cells
00:22:11.03	many years ago.
00:22:13.01	And when we culture those cells, we can take a colony, or a clone, derived from one of
00:22:19.09	the single stem cells, and we can engraft it onto a mouse that doesn't have hair,
00:22:25.14	and those mice now can produce hair.
00:22:28.06	They also produce epidermis and sebaceous glands.
00:22:31.25	And that's something that they normally don't do.
00:22:35.01	Normally, they just make hair.
00:22:38.14	They will, however, make epidermis and sebaceous glands when the skin is wounded,
00:22:44.25	and when the epidermis and the sebaceous glands are missing.
00:22:47.16	So, the stem cells receive signals, in these case... in this case, wound signals,
00:22:54.04	that stimulate those stem cells to be able to make hair, but also to be able to make epidermis
00:22:59.21	and sebaceous glands, to replenish the tissues that were missing
00:23:04.13	as a consequence of the wound.
00:23:07.05	And when you take the stem cells out of context, out of the hair follicle niche,
00:23:11.23	and you put them into culture, now the stem cells kind of forget what they were supposed to do
00:23:18.13	when they're grafted back onto the skin.
00:23:20.20	And now they basically act like they're in a wound state, and effectively will also make
00:23:26.21	epidermis and sebaceous glands.
00:23:29.00	We call this phenomenon stem cell plasticity, and it's something that we see in
00:23:35.10	a variety of different stem cell... stem cells from tissues.
00:23:40.24	It's a phenomenon that basically allows a tissue to receive, essentially, a danger call
00:23:49.18	in a wound that tells the stem cells nearby to be able to repair and replenish the tissue
00:23:58.14	at all costs, and as fast as possible.
00:24:00.25	So, it's always the closest stem cells, then, to the wound that can repair it.
00:24:06.03	Unfortunately, these stem cells will not automatically make neurons or make other tissues
00:24:13.05	of our body without some manipulation.
00:24:15.27	But these stem cells will, however, naturally make not only hair follicles but also epidermis
00:24:24.16	and sebaceous glands.
00:24:26.26	So, we've studied these stem cells over the course of many years, now.
00:24:33.16	And we know that these stem cells are defined by their transcriptional profile, effectively,
00:24:41.16	the gene expression pattern that these stem cells make.
00:24:44.28	And we know a good deal about this.
00:24:46.20	We know that there are the transcription factors in purple, a variety of different transcription factors
00:24:52.19	that are made by these hair follicle stem cells and not by other stem cells,
00:24:59.03	at least as a cohort.
00:25:00.17	We know that there are some factors, like LGR5, which are made by a number of different stem cells:
00:25:05.27	intestinal epithelial stem cells, hair follicle stem cells, and many other
00:25:11.15	types of stem cells.
00:25:12.28	We know that integrins are a characteristic of many different types of stem cells,
00:25:18.22	and they're expressed at very high levels in the hair follicle stem cells.
00:25:23.11	And then for a quiescent stem cell, the niche signals basically produce a gene expression profile
00:25:31.09	by the stem cells that reflect reduced levels of Wnt signaling and reduced levels
00:25:37.04	of sonic hedgehog signaling.
00:25:39.00	Those are the pro-activating cues for the stem cells.
00:25:42.08	Whereas BMP signals and FGF18 signals, the inhibitory cues that tell those stem cells
00:25:48.19	to remain in quiescence, those signals are upregulated.
00:25:53.13	And then differentiation, whether it be differentiation for the epidermis or the sebaceous gland
00:25:58.27	or the hair follicle, all of those programs are suppressed by the stem cells.
00:26:04.13	So, their transcriptional program reflects the behavior of these stem cells.
00:26:11.10	We characterized the transcriptional program by taking the stem cells directly out of the tissue,
00:26:17.28	using fluorescence activated cell sorting to purify those cells,
00:26:21.26	and basically taking those RNAs and directly profiling them.
00:26:26.24	We also used a technology known as conditional knockout technology to be able to excise
00:26:34.19	the transcription factors from the skin of the animal.
00:26:39.06	When we excise TCF3 and TCF4 -- these are DNA binding proteins which can also bind beta-catenin,
00:26:46.15	which is a downstream Wnt effector -- we found that the loss of TCF3 and 4 basically no longer...
00:26:53.19	the stem cells can no longer maintain the quies... their... their activity.
00:26:59.20	We learned that the loss of Lhx2 is required for maintaining the stem cells.
00:27:05.11	And similarly, the loss of Sox9 is required for maintaining the stem cells.
00:27:11.06	Surprisingly, when we knocked out beta-catenin, what we found is no hair follicles made.
00:27:17.15	So, with no Wnt signaling there were no hair follicles made.
00:27:21.28	Interestingly, when we knocked out TCF3 and 4, even though it was necessary
00:27:27.20	to make a hair follicle, what we found to maintain the hair follicles stem cells... what we found
00:27:32.23	is a burst of hair growth.
00:27:35.02	So in this way, we learned that TCF3 and 4 antagonize what Wnt signaling can do,
00:27:41.10	and that keeps the stem cells in quiescence, because Wnt is typically a proliferation-associated factor.
00:27:49.04	Lhx2, in the absence, basically produces, where the stem cell niche was,
00:27:56.01	now a niche full of sebaceous gland.
00:27:58.05	So, Lhx2 is necessary to repress the differentiation program of the sebaceous gland while
00:28:06.12	the stem cells are stem cells in their native niche.
00:28:09.12	Sox9 represses the epidermal stem cell differentiation program.
00:28:14.26	And without Sox9, the hair follicle stem cell niche turns into an epidermal cyst.
00:28:19.28	So, we understand how these different factors are working and how these different transcription factors...
00:28:25.27	what they're necessary for.
00:28:28.13	So, let's put this all together now.
00:28:31.14	And I'll give you one example, and that is from BMP signaling, which I told you is important
00:28:37.12	to control stem cell quiescence.
00:28:40.23	So, we've learned, over the course of our studies, that the BMP interacting with
00:28:47.09	the BMP receptor on the stem cells is sufficient to activate the canonical transcription factor
00:28:55.18	called phospho-SMAD1, along with phospho... along with SMAD4, which then control, downstream,
00:29:02.03	a group of transcription factors that I've shown you here.
00:29:06.17	Together, these factors are necessary to control the quiescence of the stem cells.
00:29:13.11	So, we then looked at what happens in an aging mouse.
00:29:18.11	And what we found was that in the first hair cycle of the mouse, that happens relatively quickly.
00:29:26.08	Only a couple days separates the time at which one hair follicle has completed its cycle
00:29:32.11	and the next hair cycle begins.
00:29:34.05	But now, as the animals get older, there's longer and longer times between when our hairs grow
00:29:40.20	-- or when the animal's hair grows -- and when the hair, basically... when the stem cells
00:29:46.20	sit in quiescence.
00:29:48.05	They spend far longer sitting in quiescence than they do in the younger mice.
00:29:54.23	And when we traced this, what we found is that the dermis and the old fat,
00:30:01.26	or the old adipose tissue, of these older mice turned out to produce very high levels of BMP.
00:30:08.15	And that's what maintained these stem cells in their quiescent state.
00:30:12.15	We could treat the old mice with a drug called VIVIT which is an inhibitor of Nfatc1,
00:30:21.14	one of the transcription factors downstream of the BMP signal.
00:30:25.23	And we could regrow... reactivate those dormant hair follicle stem cells in order to make
00:30:32.05	a new round of hair growth.
00:30:34.07	And so, at this point in time in our studies, what we learned is that actually the stem cell numbers
00:30:39.16	still remain high in the older animal, but it's their activity that wanes with age.
00:30:46.10	So, hair grain is also a phenomenon that... that happens with age.
00:30:53.26	And in this study, what others, May...
00:30:57.11	Mayumi Ito and her coworkers, learned was that the melanocyte stem cells and
00:31:04.21	the hair follicle stem cells reside in the same niche, and they respond to similar signals.
00:31:09.24	And this was also the work of Emi Nishimura and her coworkers in Japan.
00:31:17.19	And so hair follicle and melanocytes stem cells are hanging out in the same niche.
00:31:22.18	They respond to the same signals.
00:31:25.02	And the importance of that is that the hair follicle stem cells and melanocytes stem cells
00:31:32.18	will get activated at the same time.
00:31:35.08	They'll differentiate so that the melanocyte cells will now begin producing pigment,
00:31:42.06	and the hair follicle cells will begin producing hair.
00:31:45.27	And at that time, the differentiated melanocytes can provide that pigment to the differentiating
00:31:53.00	hair cells.
00:31:54.04	And that's what gives our hair its color.
00:31:57.03	And so, when the stem cells of the hair follicle are unable to be able to produce hair,
00:32:04.09	so too, typically, the melanocytes stem cells end up being unable to be activated,
00:32:12.04	because they reside in the same niche.
00:32:14.05	So, one of my graduate students, Kenneth Lay, had the remarkable brilliant idea that perhaps
00:32:20.28	if he simply deleted these transcription factors that are important in controlling quiescence
00:32:29.01	downstream of the BMP pathway, that he might be able to accelerate the differences, then,
00:32:34.20	that exist between the younger mice and the older mice
00:32:38.06	with regards to the length of their hair cycle.
00:32:41.21	Could he reduce the length of time that stem cells spend in quiescence?
00:32:47.23	And the answer to that experiment was quite remarkable.
00:32:51.09	And in fact, now, the mice just kept on cranking out hair.
00:32:54.21	Their stem cells were almost constantly activated.
00:32:58.11	A new hair would grow, the process would go through its cycle, and then the cycle would
00:33:03.22	reinitiate again, over and over and over again.
00:33:07.13	So, this was pretty exciting for us.
00:33:12.20	And we began to wonder whether BMP inhibition could be the Fountain of Youth
00:33:17.26	for endless hair growth.
00:33:19.16	Well, unfortunately in science, things aren't always so simple.
00:33:24.14	And what happened was that the animals precociously became grey, sooner than did their counterparts.
00:33:32.14	And it turned out that when the transcription factor was mutated, and lacking,
00:33:39.05	while the hairs kept on being cranked out eventually the number of stem cells declined.
00:33:45.03	And that then led to a precocious graying and a precocious thinning of the hair.
00:33:52.07	So, we're not yet there.
00:33:54.06	But we have some predictions.
00:33:56.15	And the predictions might be that the key to the Fountain of Youth with regard to hair growth
00:34:00.17	is a well-balanced BMP.
00:34:03.01	Clearly, by cranking out and activating those stem cells continuously.... was deleterious
00:34:09.12	for the mouse's ability to produce hair endlessly.
00:34:14.21	But perhaps with a balance tempering those numbers of hair cycles, one might be able
00:34:21.14	to achieve the goal of being able to extend the ability of... time that individuals
00:34:27.27	could grow their hair.
00:34:29.17	So, BMP also plays an important role in development and development of the skin.
00:34:37.08	And here we were fascinated by the fact that the dermis instructs the fate of what
00:34:44.02	the epidermal bud would do.
00:34:46.06	So, in embryonic development, the skin begins as that single layer of epithelium and epidermal buds
00:34:54.09	begin to appear.
00:34:56.06	And those buds can become a mammary gland, they can become a hair bud, they can become
00:35:02.02	a sweat gland bud.
00:35:04.14	It's depending upon what the dermis is instructing them to do.
00:35:09.14	And so regionally, we end up getting these different appendages of the epithelium,
00:35:17.26	generated as a consequence of the different types of mesenchyme or the derm... dermis that we have.
00:35:23.10	It turns out that in most mammals there's a BMP signal called BMP5 which is
00:35:29.27	regionally spiking only in the paw skin of the animals.
00:35:33.00	And so what happens is that most of the body region of the animals has hair, and there's
00:35:37.20	very few sweat glands.
00:35:39.02	There's only sweat glands, or eccrine sweat glands, on their paws.
00:35:43.04	And eccrine sweat glands are what allows us, what allows our body, to thermoregulate.
00:35:48.20	It's what allows us to go out there and run a marathon as we sweat three liters of sweat
00:35:54.18	in so doing.
00:35:56.04	But most mammals can't do that, and they can't do it because they have fur instead of
00:36:02.17	sweat glands over their body surface.
00:36:05.13	And so we began to study this process.
00:36:09.07	And as we discovered, the BMP5 regionally spikes in the paw dermis.
00:36:15.22	It doesn't spike in the body dermis, and therefore the animals get hair exclusively.
00:36:21.11	But then when we looked at human embryonic development, what we discovered is that BMP5
00:36:28.22	seems to be deregulated in a different way in humans than it is even in higher primates.
00:36:35.15	And in this case, BMP5 spikes broadly and temporally in the body dermis, so that
00:36:41.14	the very first few waves of epidermal bud development are forming hair in humans,
00:36:50.08	but then that last wave, when BMP5 spikes over the body, basically we end up getting hair...
00:36:57.04	getting sweat glands over our body.
00:36:59.17	And that's what gives us those wonderful properties to be able to run marathons, and to be able
00:37:07.03	to sweat and survive in extreme climates relative to other... other animals.
00:37:15.02	So, what else can we do with these adult skin cells?
00:37:19.26	Well, in part one of my discussion, I already told you about the ability to change the behavior
00:37:29.05	of an adult skin cell, and turn it into an induced pluripotent stem cell, by the activity
00:37:37.28	of four transcription factors that are produced by, or expressed by, embryonic stem cells.
00:37:43.08	And that was a remarkable change and a remarkable advance because it allowed us, now,
00:37:48.24	to be able to culture neurons, pancreatic beta cells, heart cells, etc. in vitro, starting with
00:37:56.20	skin cells.
00:37:58.23	So, can we reprogram in smaller steps?
00:38:02.17	Well, our ability to be able to culture, for instance, hair follicle stem cells...
00:38:07.22	we now know that if we simply switch off LHX2, we can make sebocytes in culture.
00:38:13.25	If we switch off the transcription factor SOX9, we can make epidermis rather than hair.
00:38:21.10	And if we switch off the regulatory factors for NFATc1 or Foxc1, we can basically halt tissue...
00:38:28.02	we can... if we switch them on, we can halt tissue regeneration.
00:38:33.24	If we switch them off, we can favor hair follicle tissue regeneration and hair regeneration.
00:38:40.04	And finally, an interesting aspect is that if we simply give the hair follicle stem cell
00:38:47.04	a transcriptional regulatory factor called PAX, a particular type of the PAX transcription factor,
00:38:52.07	then we can instruct hair follicle stem cells to make corn...
00:38:56.18	to become corneal stem cells.
00:38:59.02	And when you think about cases where patients have both eyes that are blinded by some industrial burn,
00:39:05.05	then this type of potential reprogramming in smaller steps might offer the potential
00:39:12.02	for future therapy.
00:39:14.08	So, I'd like to switch, now, to a problem of, what do I mean by switching on genes and
00:39:21.14	switching off genes?
00:39:23.18	And two of my graduate students, Rene Adam and Hanseul Yang had the interest of being able
00:39:31.04	to examine the chromatin landscape, the gene expression landscape, of the... of the
00:39:40.03	hair follicle stem cells.
00:39:41.26	And what they learned from that is that there are about 350 genes that are expressed by
00:39:48.04	the hair follicle stem cells, that are regulated by very large enhancers
00:39:53.00	-- Rick Young has coined them "super enhancers" --
00:39:57.22	that bind all of the different transcription factors,
00:40:00.20	as we showed, that are expressed by the hair follicle stem cells.
00:40:04.28	So, Tcf3, Tcf4.
00:40:07.21	We talked about Sox9.
00:40:09.06	We talked about Lhx2 and Nfatc1.
00:40:12.20	And they all bind to relatively short stretches of DNA that are contained within these large
00:40:18.27	"super enhancers".
00:40:21.09	So, those short... short stretches of DNA seem to have the... not only the binding of
00:40:29.22	all these transcription factors, but they also have the sequence motifs to which
00:40:35.06	those transcription factors bind.
00:40:38.00	So, these genes turn out to be also the genes that are most important in maintaining
00:40:45.26	the hair follicle stem cells in their quiescent, undifferentiated state.
00:40:50.01	If we take a look at a handful of the 350 genes, then, that are these special genes
00:40:55.27	regulated by super enhancers, what we find is that they include the Sox9 gene, the Lhx2 gene,
00:41:03.13	the Nfatc1 gene, and DNA... the BMP downstream gene... transcription factor activators.
00:41:12.11	They also include the BMP inhibitor and the BMP receptor.
00:41:16.24	They include a Wnt inhibitor, Dkk3, and the Wnt receptor, or Frizzled.
00:41:22.17	They also include the integrin genes that I emphasized as being important for maintaining
00:41:28.22	hair follicle stem cells.
00:41:30.23	Effectively, this short list of super enhancer-regulated genes contains all of the critical information
00:41:38.14	that is necessary for these genes to be able to essentially be a hair follicle stem cell.
00:41:45.21	I also point out that many of the genes on this list have already been associated
00:41:51.00	with human genetic disorders of the hair and the skin.
00:41:56.15	So, what I've told you is that there's a master group of transcription factors for the skin
00:42:03.21	stem cells, but it also turns out that there's a master group of transcription factors
00:42:08.04	for every stem cell type.
00:42:10.11	And they're different for different stem cell types.
00:42:13.09	So, the first studies that were done were on embryonic stem cells.
00:42:18.15	And Rick Young and his coworkers showed that the super enhancers are regulated by
00:42:24.09	the transcription factors that are necessary to maintain embryonic stem cells.
00:42:30.13	What I've just showed you is that in an adult tissue stem cell, the hair follicle stem cell,
00:42:36.02	that there's a different cohort of transcription factors that are necessary to maintain
00:42:41.00	these stem cells in being hair follicle stem cells.
00:42:44.11	So, we're starting to dig deeper into understanding, what are the differences in embryonic versus
00:42:49.26	adult stem cells?
00:42:51.26	Another interesting aspect is that all of these transcription factors are only expressed
00:42:58.03	as a cohort in hair follicle stem cells.
00:43:01.04	So, you might see Tcf3, for instance, in brain stem cells.
00:43:05.17	You'll see Tcf4 in intestinal stem cells, Lhx2 in hematopoietic stem cells,
00:43:10.20	Sox9 in bone stem cells.
00:43:14.08	But the only place you see them all together at one place in time are
00:43:17.25	the hair follicle stem cells.
00:43:19.21	And it turns out that we can take advantage of that.
00:43:22.15	We can take one of these elements that contain the binding sites for all of these
00:43:29.10	different transcription factors, and we can use that to drive the expression of a fluorescent green protein
00:43:34.20	in mice.
00:43:36.22	And now, the fluorescent green protein is only expressed by the hair follicle stem cells
00:43:42.05	-- far more specificity than we've ever achieved simply by taking a chunk of chromatin or
00:43:49.24	a promoter region, an enhancer region, and using that to drive expression in mice.
00:43:56.09	So, this type of information is really important for researchers to be able to dig deeper into
00:44:02.23	the biology of stem cells, in this case the hair follicle stem cells.
00:44:08.05	We also are learning that stem cells change their gene expression program outside of their niche.
00:44:14.05	I already showed you that a stem cell outside of its niche, when grafted back onto the...
00:44:19.18	to an animal, can generate hair follicles, epidermis, and sebaceous glands.
00:44:24.25	But now, if we take a wound induced enhancer, if we characterize the chromatin state of
00:44:31.21	the stem cell while it's in the act of repairing a wound, and now identify those regulatory elements,
00:44:37.25	and use those to drive eGFP in a mice... in a mouse, now what we find is that
00:44:45.00	in the unwounded skin there's no activity.
00:44:48.14	And now we wound the skin, and the activity turns on.
00:44:52.00	So, these regulatory elements contain the information necessary to not only
00:44:58.27	maintain this stem cell, but in this case to tell the stem cell what to do, and to change
00:45:04.16	its program of gene expression if it's going to repair a wound versus if it's just going to replace
00:45:10.27	a cell that is dying in the tissue.
00:45:14.00	So, we're also looking at how skin inflammation elicits changes in gene expression in epidermal stem cells.
00:45:22.19	So, when stem cells receive an inflammatory assault, how do they respond?
00:45:29.05	And what we learned from our studies is that they respond, and they activate new genes.
00:45:36.11	And by taking the regulatory elements from those genes, we can use those to drive
00:45:42.01	the expression of, in this case, a reporter GFP gene.
00:45:45.13	So, here we've generated mice that express these particular reporter elements,
00:45:52.21	but they only do so when the skin has been subjected to inflammation.
00:45:57.24	So, in the naive skin, we see no expression of the eGFP protein.
00:46:05.03	In the presence of inflammation -- if we give the skin an inflammatory assault --
00:46:09.26	now we see the activation of these enhancer elements.
00:46:14.27	So we call these elements inflammation sensors.
00:46:18.03	And what we've found is that there are about 2,000 inflammation-sensing chromatin domains
00:46:23.17	that remain open in skin stem cells long after the inflammation has resolved.
00:46:30.20	These sensors rapidly, then, become activated again with... and activate their associated genes
00:46:37.20	when the skin sees inflammation again.
00:46:40.19	And that's a remarkable property, because, basically, this confers to the epidermal stem cells
00:46:47.03	a memory, an inflammatory memory, that it's seen the inflammation before.
00:46:52.28	And even though the gene expression program is normal in... after the information has resolved,
00:47:00.01	there are these hidden memory open chromatin domains that basically allow
00:47:05.24	the genes to be poised so that the genes can now be activated again, and activated quicker,
00:47:12.21	the next time inflammation occurs.
00:47:16.24	And why do we think that that's important or relevant?
00:47:20.17	Skin inflammation elicits marks on epidermal stem cell chromatin.
00:47:25.09	Remember that the epidermal stem cells are there for the long haul.
00:47:28.05	They see the inflammation; they remain in their active state, as stem cells,
00:47:34.16	long after the inflammation has resolved.
00:47:37.10	But they have that memory that they've seen the inflammation before.
00:47:41.24	That memory resides in the virtue of epigenetic marks.
00:47:47.26	And here are some examples.
00:47:49.08	There are columns, now, of the epigenetic marks that are shown, in this case... in control chromatin,
00:47:56.12	shown in gray, there's no epigenetic marks.
00:48:00.22	Now, there's inflammation, the skin experiences inflammation, and we see two epigenetic marks
00:48:08.16	appear: here in this gene on the right.
00:48:11.25	And now, 30 days later, what we find is those epigenetic marks are still there.
00:48:19.02	And even 6 months later, there's still traces of those epigenetic marks residing within
00:48:26.15	these... this chromatin.
00:48:29.16	So, the epidermal stem cells have a memory, harbored in chromatin, that they've seen
00:48:37.07	the inflammation before.
00:48:39.10	And what's the importance of that?
00:48:40.21	Well, let's think about psoriasis or atopic dermatitis or contact dermatitis infections.
00:48:47.24	The infections typically come and go.
00:48:53.26	And when they come back again, they typically come back to the same places.
00:48:58.26	We're now starting to understand why that is, because, effectively, the epidermal stem cells
00:49:04.14	have a memory.
00:49:06.05	They've seen the inflammation before so that, if they see it again, they respond faster
00:49:11.14	the next time.
00:49:12.20	So, what I've told you is that adult stem cells replenish dying cells of a tissue,
00:49:19.10	and that they repair wounds.
00:49:20.27	I've told you that they reside in niches whose crosstalk dictates their behavior.
00:49:26.15	I've told you that key stem cell genes are regulated by super enhancers that bind stem cell
00:49:32.13	transcription factors, as well as transcription factors that are induced by the niche microenvironment.
00:49:38.13	For infla... inflammation, for instance, it might be a phosphorylated,
00:49:43.11	active version of STAT3, along with the epidermal stem cell factors.
00:49:48.24	I've also told you that trained immunity, the memory of inflammation, is not a phenomenon
00:49:58.14	that's restricted to the immune system, but occurs in epidermal, and we think likely
00:50:03.26	other epithelial, stem cells that bear the brunt of inflammation.
00:50:09.05	Inflammatory bowel disease, for in... for instance, or certain types of lung inflammation.
00:50:14.12	Somehow the tissue remembers that it's seen the inflammation before.
00:50:18.11	So, for immune cells, trained immunity enhances the ability of the immune cells
00:50:26.04	to eliminate the infection if they see it again.
00:50:29.13	For epidermal stem cells, trained immunity enhances their ability to restore
00:50:34.22	the skin barrier when breached and to aid in recruiting the immune system.
00:50:39.22	And remember that the most important thing that the skin epithelium does is
00:50:45.13	basically to produce that barrier that keeps harmful microbes out, and essential bodily fluids in.
00:50:51.24	And whether we're wounding the skin, or whether we end up with a hyper...
00:50:56.02	hyperinflammatory response, the barrier is at risk.
00:51:01.04	And the epithelium, the epidermal stem cells, respond by trying to repair that barrier
00:51:06.16	as fast as they can.
00:51:08.06	Sometimes they're not very effective at doing it, and that then can lead to chronic wound-healing
00:51:13.28	or chronic inflammation.
00:51:15.25	And those are obviously areas that we'd like to be able to adapt our understanding for
00:51:21.26	to make progress in treating those diseases in the future.