Session 6: T Cells: Development and Differentiation

00:00:08.05 So, my name is Diane Mathis.
00:00:09.25 I'm a professor of immunology
00:00:11.21 at Harvard Medical School.
00:00:13.21 My two talks have to do with
00:00:17.00 immunological tolerance,
00:00:18.25 in particular T cell tolerance.
00:00:22.05 In the first talk,
00:00:23.29 I'll relate some general concepts
00:00:26.17 and then go on to
00:00:29.23 define and describe the major mechanisms of tolerance.
00:00:33.17 And then in the second talk,
00:00:35.17 I'll focus on one particular mechanism
00:00:39.07 of enforcing tolerance.
00:00:42.19 So, some general concepts...
00:00:46.17 the major function of the immune system
00:00:49.15 is to fight off microbial challenges.
00:00:52.10 It does this by mobilizing its two major arms:
00:00:56.12 the innate response and the adaptive response.
00:01:00.05 The innate response comes in early
00:01:02.26 and it's quite stereotypic.
00:01:04.23 It's a hardwired system
00:01:08.09 that includes cells like macrophages and neutrophils.
00:01:13.01 The adaptive system... response
00:01:18.01 comes in a little bit later
00:01:20.00 and it's a more nuanced response.
00:01:22.10 The major players in the adaptive response
00:01:26.21 are B and T lymphocytes.
00:01:29.22 Now, these cells have, displayed on their surfaces,
00:01:33.26 receptors that are specific for particular antigens.
00:01:39.19 The B cell receptor, or BCR,
00:01:42.21 for B cells,
00:01:44.20 which is also called immunoglobulin,
00:01:47.05 and the T cell receptor, or TCR, for T cells.
00:01:50.23 Now, in general,
00:01:52.29 each B or T cell
00:01:55.29 has multiple copies of a receptor
00:01:59.03 that sees one particular antigen
00:02:01.21 or a group of structurally related antigens,
00:02:06.13 so you can understand that
00:02:08.10 in order to fight off the myriad of microbial challenges,
00:02:11.27 that the repertoire must be both big and diverse.
00:02:17.02 And this happens by
00:02:19.14 random rearrangement of gene segments
00:02:24.05 encoding variable regions of the T and B cell receptors
00:02:27.27 as these cells are differentiating,
00:02:32.00 in the thymus for T cells
00:02:33.24 and in the bone marrow for B cells.
00:02:38.16 Now, since it is a random process,
00:02:41.28 just by chance,
00:02:44.11 sometimes specificities will be generated
00:02:46.27 that are able to recognize
00:02:49.08 the body's own constituents.
00:02:51.05 For example, a T cell may see insulin
00:02:55.07 or it may see the acetylcholine receptor,
00:02:59.05 or it may see myelin basic protein,
00:03:01.23 and if these T cells are let loose
00:03:05.05 they would cause autoimmune attack on the pancreas,
00:03:08.20 or myasthenia gravis,
00:03:10.24 or multiple sclerosis,
00:03:13.20 which is an autoimmune attack in the central nervous system.
00:03:17.21 Now, since this is
00:03:20.08 a very dangerous problem for the individual,
00:03:22.29 through evolution,
00:03:25.09 multiple levels of immunological tolerance
00:03:29.23 have come into play.
00:03:32.24 Now, these are generally
00:03:35.28 divided into central and peripheral mechanisms.
00:03:39.12 Central tolerance has to do with
00:03:42.16 the primary lymphoid organs
00:03:44.15 where T cells or B cells are generated,
00:03:46.08 so, for T cells, in the thymus.
00:03:48.25 So, the antigens that would be dealt with
00:03:51.17 in central tolerance
00:03:53.07 would be antigens that are expressed in the thymus
00:03:56.19 or are expressed in all cells,
00:03:59.03 or are expressed in cells which are
00:04:02.16 trafficking through the thymus,
00:04:04.00 like blood cells, for example.
00:04:06.08 Peripheral tolerance has to do with those cells,
00:04:11.05 once they've emerged from the thymus
00:04:13.15 and are now in the periphery,
00:04:15.16 and can see potential self-antigens
00:04:19.04 in the liver or in the heart or in various tissues.
00:04:25.19 Now, the major mechanisms of central tolerance are:
00:04:30.05 first, clonal deletion,
00:04:31.28 which is physical removal of that particular T cell from the repertoire;
00:04:38.12 clonal inactivation,
00:04:42.09 where the T cell is there,
00:04:44.21 but once it gets into the periphery
00:04:47.07 it's no longer able to make a response;
00:04:51.17 and then clonal diversion,
00:04:53.27 the T cell is there and it can respond,
00:04:55.21 but during its differentiation
00:04:58.21 it gave up its identity as an effector T cell,
00:05:03.04 which is the kind of cell
00:05:05.26 which actually does the damage,
00:05:08.06 and has taken on the mantle
00:05:10.23 of another kind of cell, a regulatory cell.
00:05:14.19 The mechanisms of peripheral tolerance
00:05:16.29 are somewhat similar.
00:05:19.07 So, the first three are the same
00:05:21.08 -- deletion, inactivation, and diversion --
00:05:22.23 but onto this are added three more.
00:05:25.21 So, clonal ignorance just refers to the fact that
00:05:29.19 some T cells can escape,
00:05:31.07 but they're quite fine,
00:05:33.02 because the antigen that they see
00:05:35.00 is hidden from the lymphocyte circulation.
00:05:39.18 So, for example, it might be an antigen
00:05:42.02 which is in a cell
00:05:43.27 that's behind the blood-brain barrier
00:05:45.09 or it might be an antigen
00:05:47.06 which is in the eye,
00:05:48.18 which is thought to be "immune privileged",
00:05:52.10 or an antigen which is in the testis.
00:05:54.00 So, normally, these are not seen
00:05:55.26 unless there's some kind of damage
00:05:58.16 to these different tissues,
00:06:00.16 which releases the antigen into the circulation.
00:06:03.22 Helplessness refers to the fact that
00:06:06.07 many B cell and cytotoxic T cell responses
00:06:12.01 must have help from CD4+ helper cells
00:06:17.16 in order to respond effectively.
00:06:19.20 So, as you can imagine,
00:06:21.25 you might not need to directly
00:06:24.15 tolerize that B cell or that cytotoxic T cell,
00:06:26.13 but rather the CD4+ T cells
00:06:29.09 that are helping them.
00:06:31.21 And then, finally, suppression,
00:06:36.22 which is a major mechanism,
00:06:39.02 refers to the fact...
00:06:39.17 refers to cells which keep in check the activity
00:06:42.10 of effector T cells.
00:06:46.19 So, looking at that net of mechanisms
00:06:51.25 which I showed you,
00:06:53.00 you might be tempted to think that
00:06:55.27 this is very comprehensive
00:06:57.27 and autoimmunity should be rare.
00:07:01.10 But, breaks in tolerance do occur,
00:07:02.24 leading to autoimmunity
00:07:04.12 and, in the most extreme form,
00:07:06.22 to autoimmune disease,
00:07:08.13 and they occur quite often
00:07:12.10 -- 5-8% of the population in Western countries
00:07:16.22 have some form of an autoimmune disease --
00:07:21.03 and they occur by different means.
00:07:23.22 So, at the last calculation,
00:07:26.04 there are more than 80 different types
00:07:28.29 of autoimmune diseases,
00:07:30.25 and probably double that
00:07:34.17 because sometimes we lump together
00:07:36.19 a particular autoimmune disease
00:07:39.00 just because they have the same manifestations,
00:07:41.10 for example, arthritis,
00:07:43.15 but the way to get there might be,
00:07:45.09 actually, quite different.
00:07:48.12 So, let's look a little bit more closely
00:07:51.19 into central tolerance.
00:07:53.22 And I'd like to focus on the first mechanism that I mentioned
00:07:59.09 -- clonal deletion.
00:08:01.02 Now, as I said,
00:08:02.19 clonal deletion is physical removal
00:08:04.09 of the self-reactive T cells from the repertoire,
00:08:07.24 and this is a quite important mechanism
00:08:12.14 because fully two-thirds of the T cells
00:08:16.08 that reach full maturity in the thymus
00:08:19.00 actually undergo clonal deletion,
00:08:22.04 because they have some self-reactivity.
00:08:24.27 And it's also the most definitive mechanism
00:08:26.18 because it ends in death,
00:08:28.04 and you can't be more definitive than that.
00:08:32.11 So, I'd like to tell you
00:08:37.22 more about this mechanism of clonal deletion,
00:08:40.22 but before I do that
00:08:43.04 I have to explain to you
00:08:44.28 how T cells see antigen
00:08:46.26 and how they differentiate in the thymus.
00:08:49.20 So, a T cell actually sees its antigen quite differently
00:08:52.25 from how a B cell sees its antigen.
00:08:56.04 B cells see antigens through their B cell receptor,
00:09:00.29 or immunoglobulins,
00:09:02.10 very directly.
00:09:03.20 They actually recognize
00:09:06.11 a piece of the three-dimensional structure
00:09:09.02 and then make a response.
00:09:11.02 For T cells, it's quite different.
00:09:13.20 They must recognize
00:09:16.09 a short peptide sequence,
00:09:18.12 a primary sequence,
00:09:20.23 presented by major histocompatibility complex molecules,
00:09:27.02 which occur at the surface of most cell types.
00:09:32.01 Now, this presentation of antigens to T cells
00:09:36.06 is a quite complicated
00:09:38.22 and very elegant mechanism,
00:09:40.04 which, if you're interested in,
00:09:43.02 you can refer to the talk by Dr. Mellman,
00:09:46.09 which goes into much more detail.
00:09:49.01 But, suffice it to say, in this context,
00:09:51.14 an antigen would be either
00:09:54.09 synthesized by a cell or taken up by a cell,
00:09:57.12 and then undergoes proteolytic degradation
00:10:01.00 to make peptides.
00:10:02.14 And according to where that peptide
00:10:04.26 was actually made,
00:10:06.06 and what enzymes were involved,
00:10:07.23 as well as the primary sequence of the peptide,
00:10:10.23 it can bind either to an MHC class II molecule
00:10:15.16 or an MHC class I molecule.
00:10:17.14 And then these molecules
00:10:21.26 shuttle the peptides to the surface.
00:10:23.26 Now, some T cell receptors
00:10:26.04 can see a peptide
00:10:29.29 in the context of an MHC class II molecule,
00:10:34.20 and it does this with the help of the CD4 co-receptor,
00:10:37.11 and these T cells turn out to be
00:10:40.10 CD4+ helper T cells.
00:10:43.07 Other T cells have receptors
00:10:45.11 that are capable of recognizing
00:10:47.16 MHC Class I molecules
00:10:49.17 and peptides within them,
00:10:51.16 and they're helped along with the co-receptor CD8,
00:10:55.26 and these cells turn into
00:10:57.29 CD8+ cytotoxic T cells.
00:11:01.26 Now, as I mentioned,
00:11:03.23 T cells undergo differentiation in the thymus.
00:11:07.05 The precursor, which gives rise to a T cell,
00:11:09.28 comes from the fetal liver
00:11:12.25 or the adult bone marrow,
00:11:14.24 and these precursors are really very ignorant.
00:11:19.27 So, they must go through a number of processes in the thymus
00:11:22.27 to educate them,
00:11:26.07 they must learn their antigen specificity
00:11:29.08 -- and they do this by rearrangement of the T cell receptor genes --
00:11:35.00 they must have an enrichment for T cells
00:11:37.13 which are able to see the MHC molecule
00:11:39.23 that the individual is expressing,
00:11:43.02 but not see it so well
00:11:47.06 that it might cause autoimmunity.
00:11:49.15 So, these are positive selection
00:11:51.02 and negative selection events.
00:11:53.14 And then, finally,
00:11:55.24 phenotypic specialization takes place.
00:11:59.08 So, I mentioned that there are CD8+ cytotoxic T cells
00:12:02.19 and CD4+ helper T cells,
00:12:04.27 and these are what we call effector cells,
00:12:07.06 which get the job done in an immune response.
00:12:11.02 There are also CD4+ regulatory T cells,
00:12:14.01 which control the activities of these effector cells.
00:12:19.14 Now, this all takes place in the thymus
00:12:22.00 by an orchestrated series of events.
00:12:27.17 The precursors enter through the blood vessels
00:12:32.23 at the corticomedullary junction,
00:12:35.25 and then they percolate through the cortex
00:12:38.25 and the medulla,
00:12:40.20 coming into contact with various stromal cell types
00:12:43.29 that express MHC molecules
00:12:48.05 with peptide at their surface,
00:12:49.15 and also produce different cytokines
00:12:53.02 and have different other types of receptors on their surface.
00:12:56.29 And through this series of contacts,
00:13:00.24 these different processes,
00:13:03.09 which are so important for the life of the T cell,
00:13:06.24 take place.
00:13:08.15 Now, it's possible to monitor this
00:13:10.28 using flow cytometry,
00:13:12.08 using the CD4 and CD8 co-receptors
00:13:16.12 as markers of the differentiation pathway.
00:13:20.19 So, when the precursor enters from the blood,
00:13:24.19 it doesn't express either CD4 or CD8,
00:13:29.16 and during this early time period
00:13:32.08 it's basically undergoing cell division
00:13:36.13 and is starting gene rearrangement.
00:13:40.06 Now, all of this takes place inside the cortex.
00:13:45.25 Now, eventually, CD4 and CD8 are both turned on,
00:13:50.13 to have the double-positive stage,
00:13:52.23 which also takes place in the cortex.
00:13:55.18 Now, immunoglobulin...
00:13:57.23 sorry, T cell receptor gene rearrangement
00:14:00.15 is completed
00:14:03.01 during this double-positive stage,
00:14:05.12 and these cells are then
00:14:07.17 ready to be positively and negatively selected.
00:14:12.17 If they're positively selected
00:14:15.12 and deemed worthy of final maturation
00:14:19.07 they become either CD4 single-positive,
00:14:22.02 if they saw an MHC class II molecule with peptide,
00:14:24.18 or CD8+,
00:14:26.27 if they saw an MHC class I molecule with peptide.
00:14:29.02 And this final step of differentiation
00:14:33.04 happens as the cells are entering the medulla.
00:14:36.11 So, the fate of a differentiating T cell
00:14:42.28 is very much focused on these molecules
00:14:45.15 that I've portrayed here
00:14:47.11 -- the interaction between the T cell receptor and co-receptors,
00:14:52.00 and the MHC molecules.
00:14:55.03 And it has to do with
00:14:58.06 the affinity of this interaction
00:15:00.12 -- so, how strongly the T cell receptor can see this complex --
00:15:04.21 and the avidity
00:15:07.00 -- how many T cell receptor and MHC complexes
00:15:09.14 actually become engaged on a particular set of cells?
00:15:15.26 So, with the least amount of interaction,
00:15:22.18 so, no interaction,
00:15:24.15 or with the strongest interaction,
00:15:26.13 death is the outcome.
00:15:30.00 So, with little interaction, the cell with die because of neglect
00:15:32.20 -- it doesn't get any signaling --
00:15:34.23 within three days of becoming a double-positive cell.
00:15:39.19 Now, with very strong interaction,
00:15:43.10 clonal deletion will take place,
00:15:45.11 and that cell is removed from the repertoire.
00:15:48.24 Now, with a signal that's a little bit stronger,
00:15:50.22 that's quite weak,
00:15:52.28 but stronger than nothing,
00:15:54.28 naive T cells will
00:15:58.23 continue maturation and eventually leave the thymus.
00:16:01.21 And these will be
00:16:04.02 both CD4 and CD8-positive T cells,
00:16:06.03 depending on whether the MHC molecule
00:16:08.19 was class I or class II.
00:16:11.05 And then somewhere between
00:16:13.22 the naive T cell and clonal deletion,
00:16:16.06 there's clonal diversion,
00:16:18.02 and that signal leads to
00:16:21.04 the changing of the effector T cell phenotype
00:16:24.09 to a regulatory T cell,
00:16:26.15 as I've described before,
00:16:27.26 or perhaps a cell which gets
00:16:30.18 shunted off to the intestine,
00:16:32.02 where it's rather innocuous.
00:16:37.11 So, I'd like to show you
00:16:39.22 some examples of clonal deletion,
00:16:42.24 and the most striking ones come
00:16:45.27 from T cell receptor transgenic mice.
00:16:47.22 So, unfortunately, any particular T cell specificity
00:16:51.06 is very rare in an individual
00:16:54.03 -- it's usually only 1 in 10^4
00:16:56.25 to 1 in 10^6 T cells
00:16:59.21 are of one particular specificity.
00:17:03.23 That means that that T cell receptor sees a particular antigen.
00:17:07.09 And that's very difficult to study and follow these cells,
00:17:09.28 so the trick that immunologists play
00:17:12.15 is to take a T cell clone,
00:17:15.20 which they know the specificity for
00:17:17.25 and they know where it came from,
00:17:19.11 and they take the already rearranged T cell receptor genes
00:17:24.01 from that T cell clone,
00:17:25.13 and use those to make transgenic mice with.
00:17:28.13 And since those are already rearranged,
00:17:30.07 they shut down endogenous rearrangement,
00:17:32.14 and you end up with a transgenic mouse
00:17:35.07 where the repertoire is highly skewed
00:17:37.22 for that transgene-encoded T cell receptor.
00:17:42.19 So, in one case, umm...
00:17:45.11 investigators started with a T cell clone
00:17:48.15 that was CD8+
00:17:53.20 and recognized an antigen which is male-specific,
00:17:56.04 and that's called the H-Y antigen
00:17:58.16 because it's encoded on the Y chromosome.
00:18:01.17 And they took those T cell receptor genes
00:18:03.14 and made TCR transgenics,
00:18:05.23 and they looked at differentiation in the thymus.
00:18:09.23 And what they saw, at steady state,
00:18:11.13 was that, yes, you get double-negatives,
00:18:14.15 as expected,
00:18:15.24 you get double-positives,
00:18:17.15 and then the single-positives are primarily CD8+,
00:18:21.08 because the clone that you started with was a CD8+ T cell.
00:18:25.07 Now, this is in females,
00:18:26.16 where the H-Y antigen does not exist.
00:18:29.27 If instead, you look in males
00:18:32.09 that are TCR transgenic,
00:18:33.23 what you find is you get double-negatives and you don't see any other T cells after that
00:18:40.02 -- they've been clonally deleted.
00:18:42.03 Now, another example
00:18:46.01 comes from a T cell clone that is CD4+
00:18:47.23 and it sees an antigen
00:18:50.17 which is found in the blood,
00:18:52.29 and it's called the C5 antigen.
00:18:54.26 It's a complement protein.
00:18:57.16 And when people make these T cell receptor transgenics,
00:19:00.29 in a line of mice
00:19:03.19 that naturally does not have C5,
00:19:05.27 they find that there are double-negatives,
00:19:08.09 double-positives,
00:19:09.14 and primarily CD4 single-positive cells.
00:19:12.12 However, if instead, they looked on a line
00:19:16.04 where C5 does exist,
00:19:18.18 they find double-negatives and double-positives,
00:19:22.02 but not CD4 single-positives.
00:19:25.03 Now, from this one slide,
00:19:26.26 you can already learn several things
00:19:28.19 about clonal deletion.
00:19:30.13 First of all, clonal deletion
00:19:34.02 happens to both CD4+ and CD8+ T cells.
00:19:38.06 And then, secondly, it can occur in different places
00:19:41.24 at different stages
00:19:43.17 during T cell differentiation.
00:19:45.06 So, in the top case,
00:19:47.24 we had an antigen which is male-specific,
00:19:50.16 and it's a ubiquitous antigen found on all cell types.
00:19:54.03 So, as soon as T cells... as double-positives turned...
00:19:59.04 have finished the rearrangement of their T cell receptors,
00:20:01.16 they get deleted,
00:20:02.26 and there's no T cells available for continuing with maturation.
00:20:08.09 It's a different case with this C5 antigen,
00:20:10.26 which is found in the blood,
00:20:12.14 because blood circulates through the thymus,
00:20:15.10 in the medullary region
00:20:17.20 and not in the cortical region,
00:20:18.23 and so double-positives never see this antigen
00:20:21.28 and it's only when the T cells are fully mature
00:20:25.15 and move into the medulla
00:20:26.28 that clonal deletion occurs.
00:20:28.27 In fact, clonal deletion can even occur
00:20:31.28 in the periphery under certain conditions.
00:20:38.14 So, the mechanism of clonal deletion is apoptosis,
00:20:43.25 and I think this is nicely shown
00:20:47.20 by the experiment illustrated on this slide.
00:20:49.06 So, here, people are
00:20:52.06 dealing with a T cell whose receptor
00:20:56.00 sees a quite common antigen
00:20:58.07 and sees it in the medulla.
00:21:01.26 And what they find...
00:21:03.19 and the way they look for clonal deletion
00:21:05.28 is to use what's called the TUNEL assay,
00:21:08.23 which is an assay which radioactively labels
00:21:11.04 free DNA ends,
00:21:13.09 which are generated during the process of apoptosis.
00:21:18.29 And so you can see, at the top,
00:21:21.07 when the antigen is present,
00:21:23.00 that there's a lot of apoptosis occurring in the medulla...
00:21:27.21 a little bit in the cortex, but mostly in the medulla...
00:21:29.08 however, when the antigen is absent,
00:21:31.12 you don't see these apoptotic structures.
00:21:39.15 As an example of how important
00:21:43.03 clonal deletion can be,
00:21:44.16 one should recognize the fact that
00:21:48.18 there are human autoimmune diseases
00:21:50.27 which reflect deficits in clonal deletion.
00:21:56.06 So, in one example,
00:21:58.05 there's a mutation in a transcription factor
00:22:01.04 which is important as a general
00:22:07.03 effector of immunological tolerance,
00:22:11.00 and so what happens is that
00:22:13.14 these individuals get a multi-organ autoimmune disease
00:22:16.07 called autoimmune polyglandular syndrome type-1.
00:22:20.02 In the second example,
00:22:22.03 it's a quite specific deficit
00:22:24.20 in central tolerance,
00:22:26.14 having to do with expression of the insulin gene
00:22:29.21 in the thymus,
00:22:31.24 and these individuals
00:22:34.26 specifically develop type-1 diabetes.
00:22:36.29 Now, the first example
00:22:39.11 will form the basis of my second talk,
00:22:42.06 so I won't go any further on that now,
00:22:45.11 but I will give you a little bit of information,
00:22:48.11 more information about the second example.
00:22:52.29 So, people have done a number of genetic studies
00:22:55.21 and identified several genes
00:23:00.11 which predispose to the development of type-1 diabetes in humans,
00:23:04.18 or protect from the development of type-1 diabetes.
00:23:09.05 Now, the most important one of these genes
00:23:11.15 is the HLA locus,
00:23:14.11 which is the equivalent of the MHC gene,
00:23:15.29 which I introduced you to.
00:23:17.29 But the second most important
00:23:20.00 is actually the insulin gene itself.
00:23:22.03 Now, the variation in
00:23:25.15 the diabetes-predisposing
00:23:28.00 versus the diabetes-protective insulin gene
00:23:31.23 is not in the coding region of the insulin gene,
00:23:35.15 but rather in the promoter region,
00:23:38.08 where there's a class of sequences
00:23:41.15 called variable number of tandem repeats
00:23:45.14 -- just a bunch of repeated sequences --
00:23:48.07 and depending on how many of these sequences
00:23:51.28 there are in the promoter region,
00:23:53.10 the allele is either
00:23:56.29 diabetes-promoting or diabetes-protective.
00:23:59.19 So, when there are just a few of these repeats,
00:24:01.17 it's diabetes-promoting,
00:24:03.01 and when there are more
00:24:05.22 it's diabetes-protective.
00:24:07.12 So, if we look through the population at these alleles
00:24:11.08 that have either protective or promoting VNTRs,
00:24:15.03 what we find is that those people who have the protective...
00:24:22.15 the promoting allele, the predisposing allele,
00:24:25.12 develop diabetes much more commonly
00:24:28.08 than the population in general.
00:24:32.00 And those are the examples on your left.
00:24:36.10 Now, the ones on your right
00:24:38.03 are those individuals who have
00:24:40.04 two copies of the protective allele,
00:24:42.14 and you can see that they develop diabetes
00:24:44.26 much less frequently than the general population.
00:24:49.04 And then, if you look to see
00:24:53.17 what these variations in the promoter regions
00:24:56.28 do to insulin gene expression,
00:24:58.17 what you find is that
00:25:01.06 there's not much difference in the expression of insulin
00:25:04.14 in the pancreas.
00:25:05.25 Where there is a big difference
00:25:07.11 is actually in expression of insulin
00:25:10.06 in the thymus.
00:25:11.18 And people have made the speculation
00:25:18.02 that those cases where there's higher expression,
00:25:22.00 for example, with the class III alleles,
00:25:24.27 these people have more expression of insulin,
00:25:28.26 greater clonal deletion in the thymus,
00:25:31.18 and then less development of autoimmunity.
00:25:34.11 And this has actually been modeled in mice
00:25:37.19 by making transgenic mice
00:25:40.06 which are expressing the different types of human alleles,
00:25:44.09 and the results fit very well this idea.
00:25:49.20 So, I hope I've convinced you by now that
00:25:53.22 central tolerance, and in particular clonal deletion,
00:25:55.11 is an important mechanism of immunological tolerance.
00:25:59.05 However, clonal deletion is never complete.
00:26:03.26 There are antigens which are not expressed in the thymus
00:26:07.02 or they're not expressed...
00:26:11.15 there are antigens which T cells don't see
00:26:13.28 with a high enough affinity,
00:26:15.15 or there are antigens whose concentration in the thymus
00:26:19.15 is not high enough to permit clonal deletion.
00:26:23.12 And one might actually make the statement
00:26:28.00 that it would be devastating
00:26:29.25 if clonal deletion was actually complete,
00:26:33.15 because, as you can imagine,
00:26:35.04 if you deleted all the T cells that saw any antigen
00:26:39.17 with any affinity
00:26:43.02 that there would be a very small repertoire
00:26:45.02 that would emerge into the periphery,
00:26:47.08 and not be able to fight off
00:26:49.09 the millions and millions of microbes
00:26:51.25 that the individual is going to encounter.
00:26:56.03 So, that leaves open
00:26:59.26 the space for peripheral T cell tolerance.
00:27:05.14 And I'm going to focus on one particular and important mechanism
00:27:09.23 of peripheral tolerance,
00:27:11.16 and that's suppression.
00:27:13.28 And, as I mentioned, suppression
00:27:17.28 is regulation of the behavior of self-reactive T cells
00:27:20.28 by other T cells... by other cells.
00:27:24.04 This is, again, an important mechanism,
00:27:27.01 and that can be seen very clearly
00:27:30.19 by the phenotype of either humans of mice
00:27:36.04 who are lacking a particular kind of regulatory T cell,
00:27:39.21 a suppressor T cell,
00:27:43.04 and these people get a very severe
00:27:45.13 autoinflammatory disease.
00:27:47.14 And, in fact, there are
00:27:49.14 more than just this type of suppressor cell
00:27:53.14 that keep effector T cells in check.
00:27:58.17 There are actually several types:
00:28:00.10 T cells;
00:28:02.19 some people think some B cells suppressor cells;
00:28:04.15 and also some macrophage suppressor cells.
00:28:08.15 But I'm going to focus my comments
00:28:11.06 on one very famous regulatory T cell,
00:28:14.16 the most famous T cell in immunology, actually,
00:28:18.11 and these are regulatory T cells
00:28:21.13 which express the transcription factor Foxp3,
00:28:24.11 and we call them, affectionately,
00:28:27.04 Treg cells.
00:28:28.23 So, these T cells express
00:28:31.24 the αβ T cell receptor,
00:28:33.17 the CD4 co-receptor,
00:28:34.19 and they express high levels
00:28:37.12 of the high-affinity receptor for IL-2,
00:28:42.10 and this molecule is called CD25,
00:28:45.07 and that was used for many years
00:28:47.18 as a means to distinguish them from other types of T cells.
00:28:50.21 But later, it was found that
00:28:53.16 these T cells are actually a particular lineage of T cells
00:28:56.18 and what defines that lineage
00:28:58.27 is the transcription factor Foxp3.
00:29:01.13 Now, normally circulating through the body
00:29:07.12 of just a standard person or mouse,
00:29:10.02 around 5-15% of the CD4+ T cells
00:29:14.23 are these Foxp3+ regulatory T cells.
00:29:18.19 And they're importance became very clear
00:29:21.24 when it was understood that
00:29:24.20 humans that have the IPEX disease,
00:29:27.08 and mice that have the Scurfy disease,
00:29:30.12 have a very severe autoinflammatory disease
00:29:35.01 affecting many organs,
00:29:36.13 because they are missing Foxp3
00:29:39.24 -- they have a mutation in it --
00:29:41.14 and sub...
00:29:43.20 consequently they are missing this population of regulatory T cells.
00:29:48.12 Now, as I mentioned,
00:29:49.26 these T cells are generated...
00:29:54.17 most of them are generated in the thymus
00:29:56.21 by the process of clonal diversion.
00:30:00.12 In other words, their T cell receptors
00:30:03.03 see a self-antigen expressed in the thymus
00:30:07.13 at an intermediate affinity
00:30:10.12 between what a naive T cell sees
00:30:13.10 and allows it to undergo maturation,
00:30:16.20 and a clonally-deleted T cell
00:30:22.27 that undergoes negative selection sees.
00:30:25.21 By now, using gain-of-function and loss-of-function experiments,
00:30:29.24 it's been determined that
00:30:32.23 Tregs actually control almost all types of immune responses.
00:30:36.26 They control autoimmune diseases
00:30:39.21 like inflammatory bowel disease or type-1 diabetes.
00:30:43.07 They control inflammatory diseases like rheumatoid arthritis,
00:30:47.14 allergic diseases such as asthma.
00:30:50.07 They also keep in check graft rejection.
00:30:53.28 They promote tumor escape,
00:30:58.12 and they inhibit the response
00:31:03.13 to various types of infections.
00:31:06.18 So, from graft rejection up,
00:31:10.20 the regulatory T cells are the good guys
00:31:13.25 -- the more you have, the less disease you have.
00:31:16.24 From tumors down,
00:31:19.18 the regulatory T cells are the bad guys,
00:31:22.04 so the more Tregs you have,
00:31:23.28 the less chance the individual has
00:31:27.09 of mounting an immune response against the tumor,
00:31:30.12 and the same thing for infections.
00:31:36.00 It was found that regulatory T cells,
00:31:41.17 as I've said,
00:31:44.05 are critical for controlling all types of inflammatory responses in the body,
00:31:49.25 and actually that this happens throughout life,
00:31:52.06 and that was determined by looking at mice
00:31:57.08 where it was possible to ablate regulatory T cells
00:32:03.04 by deleting the Foxp3 gene.
00:32:05.22 And so, if this was done from birth on,
00:32:09.25 it was very clear that the mice were very sick.
00:32:13.26 They died before 25 days of age
00:32:17.05 and they had a massive expansion
00:32:24.18 of immune cell populations
00:32:26.22 in the spleen and the lymph node,
00:32:28.22 and you can see in the bottom picture that they're actually quite small
00:32:31.25 and wasting away.
00:32:33.27 And in fact, you can wait
00:32:36.03 until the mouse becomes an adult,
00:32:38.04 and so it's had Tregs during this whole time,
00:32:40.28 and then just deplete Tregs
00:32:43.23 in an adult mouse,
00:32:45.09 and within a week or so
00:32:48.18 these mice will also develop
00:32:51.13 a very severe and fatal autoinflammatory disease.
00:32:57.08 More specifically,
00:32:59.07 Tregs will control particular types of autoimmunity,
00:33:02.00 and one nice example is the
00:33:05.19 development of type-1 diabetes in a mouse model
00:33:08.08 called the NOD mouse.
00:33:10.10 Now, when Tregs are there,
00:33:13.08 these Foxp3+ regulatory T cells,
00:33:16.14 autoimmunity does not develop
00:33:20.12 or it develops very slowly.
00:33:22.17 Whereas, if you get rid of Tregs,
00:33:26.16 it will develop much, much more quickly.
00:33:30.11 You can add in Tregs
00:33:34.09 and what you find is that these animals
00:33:36.20 that got diabetes really, really quickly
00:33:40.18 would now develop it much, much later,
00:33:45.18 off-scale on the timeframe that I'm showing you.
00:33:51.15 So... and then I'd just like to finish
00:33:54.21 by talking a little bit about the different effector mechanisms
00:33:57.13 that regulatory T cells use.
00:34:00.05 There are several.
00:34:02.19 One is by the production of anti-inflammatory cytokines
00:34:06.21 such as IL-10 or IL-35.
00:34:11.04 The second is by actually killing the effector T cell.
00:34:17.19 Third, they disrupt the metabolism of the cells,
00:34:23.13 for example,
00:34:25.19 I mentioned that regulatory T cells have
00:34:28.10 high levels of the high-affinity IL-2 receptor,
00:34:30.25 so they can act as a sink
00:34:34.12 for sucking up all this IL-2, which is actually required for the health
00:34:39.09 of the effector T cells.
00:34:41.24 And then, lastly, Tregs can
00:34:44.22 affect other types of cells,
00:34:46.23 antigen-presenting cells in the region,
00:34:49.09 and stop them from
00:34:54.22 triggering an immune response,
00:34:56.28 either by killing the cell or
00:34:59.08 changing what type of cytokines these cells make.
00:35:02.07 So, this is a quite complex mechanistic scenario
00:35:11.16 and people in the field have been asking,
00:35:14.04 is it the case that every single Treg cell
00:35:17.10 is able to do these different types of inhibition
00:35:22.22 and, if so,
00:35:26.16 when does one come into play and the other replace it?
00:35:29.00 Or it is that in any kind of Treg response,
00:35:33.04 there is a heterogeneous set of cells there,
00:35:37.22 some of which are specialized in doing mechanism 1
00:35:40.23 and others in doing mechanism 4
00:35:43.06 and others in mechanism 3?
00:35:45.05 And that's something which the field is very interested in
00:35:49.29 determining at the moment.
00:35:52.03 So, I'll just finish by saying that regulatory T cells
00:35:54.11 are a very exciting field at the moment in immunology,
00:35:59.06 that there's a lot of interest in
00:36:02.17 actually using them to control autoimmune diseases
00:36:07.00 or other inflammatory diseases
00:36:09.25 by taking them out, expanding them,
00:36:13.11 and reintroducing them into individuals with various diseases.
00:36:19.01 Thank you.

00:00:15.01 My name is Dan Littman.
00:00:16.04 I'm a professor of Molecular Immunology at the Skirball Institute,
00:00:19.10 which is part of New York University School of Medicine.
00:00:22.11 And I'm also an Investigator of the Howard Hughes Medical Institute.
00:00:26.04 What I'm going to tell you about is how the microbiota, as well as other components
00:00:33.17 of the environment, influence the immune system at barrier surfaces.
00:00:37.13 I'm going to focus mostly on the intestine, which is the area that's been best studied
00:00:43.02 by many groups around the world during in the last decade.
00:00:46.11 And I'll tell you, in my first part of the talk, about the different kinds of cells
00:00:51.11 that are involved, as well as some of the signals that are involved, particularly in the functions
00:00:56.23 of lymphocytes, in lymphocytes that are both within the innate immune system
00:01:01.16 and the adaptive immune system.
00:01:03.21 The adaptive immune system consists of B lymphocytes and T lymphocytes that differentiate
00:01:09.10 from common lymphoid progenitors.
00:01:11.09 And these are adaptive because they have rearranging genes that give rise to T cell receptors
00:01:17.20 or antibody receptors on the surface of these cells,
00:01:21.05 so that each cell has a clonally restricted type of receptor.
00:01:24.12 The innate cells... lymphoid cells also differentiate from a common lymphoid progenitor, but they
00:01:30.27 have fixed receptors, and they are basically hardwired to respond to various cues that
00:01:37.02 are presented to them, be they cytokines or be they some type of danger,
00:01:44.13 antigens that are presented to them.
00:01:45.26 So, I'll start out talking about the T cells, and in the second part of this presentation
00:01:51.00 I will tell you a little bit about some of the innate lymphoid cells.
00:01:54.25 T lymphocytes develop in the thymus from common progenitors that... that enter the thymus
00:02:02.07 as so-called double negative cells.
00:02:04.11 And the double negative cells are so called because they don't express on their cell surface
00:02:08.19 the molecules CD4 and CD8.
00:02:11.00 As they undergo development, these cells can take one of two lineages.
00:02:17.13 One lineage is to become gamma delta T cells, meaning they have receptors encoded by
00:02:22.16 the gamma and delta genes.
00:02:24.07 Or they become alpha beta T cells.
00:02:26.08 The alpha beta T cells typically express both CD4 and CD8 on the surface as
00:02:32.28 they become double positive cells.
00:02:34.17 And these cells develop, if they have appropriately rearranged receptor genes that give rise to
00:02:40.10 the protein on the surface, a heterodimer of the alpha and beta chains.
00:02:44.14 The vast majority of these cells undergo cell death, because for a cell to develop in the thymus,
00:02:49.11 it needs to have a receptor that interacts with self MHC proteins,
00:02:53.24 major histocompatibility complex proteins,
00:02:57.01 that have peptides presented to the T cell receptor.
00:03:00.10 So, since most of the cells do not have an appropriate receptor, they undergo cell death.
00:03:05.00 A few of the cells also have receptors that interact with very high affinity with self-antigen.
00:03:10.13 And these are potentially damaging cells that can lead to autoimmunity, and those are
00:03:14.23 also eliminated through a process called negative selection.
00:03:17.20 And then the few cells that make it through this gauntlet undergo positive selection,
00:03:23.04 and most of them become either CD4-positive cells or CD8-positive cells.
00:03:28.13 The CD4-positive cells are those that are selected on MHC class II molecules,
00:03:33.18 and they are typically helper cells.
00:03:35.26 And I'll talk a great deal about this throughout the rest of my presentation.
00:03:39.11 The CD8-positive cells are selected on MHC class I, and they mostly become
00:03:44.28 cytotoxic or killer T cells.
00:03:46.13 There's another type of CD4 cell that's also selected on MHC class II.
00:03:51.00 Typically, these are cells that have a higher affinity receptor.
00:03:54.20 And these cells up regulate the transcription factor Foxp3 and become regulatory T cells.
00:04:00.02 So, these are thymically derived regulatory T cells, which are essential to maintain tolerance
00:04:05.15 in the periphery and prevent autoimmune activation of the other types of cells of the immune system.
00:04:12.06 I'll tell you a little bit about another type of regulatory T cell that arises in the periphery,
00:04:18.01 the so-called induced Treg cells, in... a little bit in this presentation
00:04:24.07 and a lot more in the second part of my talk.
00:04:27.28 Once these cells develop the thymus, they are exported into the periphery.
00:04:32.16 And most of the classical alpha beta T cells, the CD4 cells and the CD8 cells
00:04:38.16 go to secondary lymphoid organs, where they are naive T cells awaiting to be activated by immune signals,
00:04:47.03 potentially from invading microorganisms, and following activation of innate immunity.
00:04:53.21 But there are also cells, such as the gamma delta T cells, as well as some subsets
00:04:58.24 of alpha beta T cells, that go directly to peripheral organs: into skin, the epithelium in the intestine,
00:05:05.09 as well as the lamina propria in the intestine, as well as the female reproductive tract, or the lung.
00:05:11.19 And these... these are cells that are very much like innate lymphoid cells, in that
00:05:18.02 they can often be activated very quickly to deliver the cytokine load that they have.
00:05:24.18 But the more conventional cells from the lymphoid organs, once they are activated by antigen
00:05:30.06 in those organs, can migrate out to the very same sites in the periphery, into these different...
00:05:36.18 different tissues.
00:05:37.18 And once these cells migrate to the different tissues, they can stay there and continue
00:05:41.28 to replenish themselves.
00:05:43.21 And these become tissue-resident memory T cells.
00:05:46.13 And they consist not only of the T lymphocytes that I just mentioned, but also of
00:05:50.26 innate lymphoid cells that I'll tell you more about, and also macrophages.
00:05:54.13 In particular, there are different populations of macrophages, some of which arise very early
00:05:59.17 during fetal development, that establish themselves into tissues and then reside in those tissues
00:06:04.19 for the life of the organism, continuing to replenish themselves.
00:06:08.02 So, these are to be distinguished from the other types of cells that are circulating
00:06:12.05 between the blood and the lymph and the secondary lymphoid organs.
00:06:16.02 And that is a distinction that one should keep in mind.
00:06:20.08 So, these are many different tissues that now harbor these tissue-resident cells.
00:06:25.08 An example is shown here from work of Daniel Mucida, in which he described T lymphocytes
00:06:31.28 that... that established themselves in the epithelium of the intestine, in this case,
00:06:37.19 the small intestine.
00:06:38.28 And you can see the tracings of these cells as they traffic through the epithelium.
00:06:44.17 They basically undergo a flossing-like movement in which they are detecting potentially harmful pathogens
00:06:51.04 as well as any kind of damage to the tissue, and then make the cytokines
00:06:56.22 and growth factors to repair the tissues, or rid the organism of the potential pathogen.
00:07:03.28 I'm going to focus mostly on the CD4-positive T cells, and most of these cells are
00:07:11.05 helper T cells, because they help the cytotoxic T cells to undergo their functions.
00:07:16.18 They also help B lymphocytes to make antibodies.
00:07:20.14 But it's... about 30 years ago, Tim Mosmann and Bob Coffman, at the DNAX Institute at that time,
00:07:26.02 first described different properties of CD4-positive T cells in that
00:07:31.16 they could make different types of cytokines.
00:07:34.02 What they found was one subset of cells made the cytokine interferon gamma.
00:07:40.10 And they and others then found that these cells are critical for killing a variety of
00:07:44.17 intracellular microbes -- bacteria, viruses, protozoa -- are controlled by these T helper 1 cells.
00:07:53.24 These cells express the transcription factor T-bet, which is required for their differentiation.
00:07:58.25 The other cell type that they identified secreted interleukin-4 and interleukin-5,
00:08:06.08 and later was shown that they also make IL-13.
00:08:08.28 And these are critical for controlling infection with helminths, or parasitic worms.
00:08:14.03 And on the other hand, these cells are also very important in allergy and in asthma,
00:08:19.25 and need to be controlled.
00:08:21.26 The Th1 cells were initially thought to be the key cells involved in autoimmune disease.
00:08:27.02 But about a decade ago, a third type of differentiated T cell was described, called the Th17 cell.
00:08:33.28 And these cells are so-called because they make the cytokines interleukin-17A and interleukin-17F.
00:08:38.23 They also make interleukin-22.
00:08:42.07 And it was found that these are actually the cells that are most often involved in autoimmune inflammation.
00:08:50.02 And these are cells that are normally needed to kill extracellular bacteria and fungi
00:08:55.26 at mucosal surfaces.
00:08:57.18 They're very important for repairing damage to mucosal tissues.
00:09:01.19 And in a number of different models for autoimmunity, they have been found to be the critical cells.
00:09:06.15 But more important, they have been found to be critical cells in autoimmunity in human.
00:09:13.13 I'm going to tell you a lot about these.
00:09:15.20 First of all, just as a way of background, the IL-17 cytokines are very important
00:09:20.26 for inducing chemoattractant cytokines, chemokines that attract neutrophils to the site of the secretion.
00:09:28.01 They're also involved in tissue remodeling.
00:09:31.17 IL-17 induces matrix metalloproteinases, as well as VEGF, which leads to angiogenesis
00:09:42.00 in a variety of tissues.
00:09:43.27 Interleukin-22, on the other hand, is more of a cytokine that leads to
00:09:49.00 proliferation of epithelial cells, and protects the barriers from damage.
00:09:55.21 There's another, fourth cell type that I want to describe here, and I mentioned it briefly already.
00:10:00.07 That is the induced regulatory T cell, which, like the one that's made in the thymus,
00:10:04.12 also expresses FOXP3.
00:10:05.12 But in the periphery, these are cells that are induced by combinations of cytokines,
00:10:12.10 particularly TGF-beta and retinoic acid, as well as interleukin-2.
00:10:16.26 And you'll note that the Th17 cells can also rely on TGF-beta for their differentiation.
00:10:24.03 And I'm going to concentrate on telling you a little bit about the requirements
00:10:28.20 for the differentiation of these cells into... in these different directions.
00:10:32.21 When a T lymphocyte is activated, it requires two signals through... in order to proliferate
00:10:40.00 and produce their cytokines.
00:10:43.06 One of the signals of course comes from the signaling pathway linked to the T cell antigen receptor,
00:10:48.06 which interacts with MHC and... either class I or class II MHC and peptide.
00:10:54.08 But the second signal is mediated through CD28, which is called a costimulatory molecule.
00:10:59.28 It interacts with ligands on antigen-presenting cells, on specialized antigen-presenting cells,
00:11:05.12 particularly dendritic cells.
00:11:06.25 And only when these two signals are integrated will the cell, now, become activated and proliferate.
00:11:12.23 But a third signal is needed for the differentiation.
00:11:15.19 And that is a signal provided by cytokines that are made, also, by these antigen-presenting cells
00:11:20.09 most of the time.
00:11:22.13 And these cytokines signal through a variety of different receptor subsets in order to
00:11:29.12 provide the cell with the... with the function that it's going to adopt.
00:11:34.20 An example of this is shown here, in which I show the critical cytokines involved in
00:11:40.17 the differentiation of Th1 cells and of Th17 cells.
00:11:45.17 And these are the cytokines interleukin-12 and interleukin-23.
00:11:49.04 IL-12 and IL-23 share a subunit, the p40 subunit shown here.
00:11:54.18 And they also share a receptor, the IL-12 receptor beta-1, you can see is present
00:12:00.05 as a receptor for both cytokines.
00:12:01.21 But then they also have unique subunits -- p35 for IL-12 and p19 for interleukin-23 --
00:12:09.27 as well as unique receptor polypeptides, which link them up to the signaling pathways downstream.
00:12:18.13 Much of the early confusion about Th1 cells was because of these shared components,
00:12:24.08 of p40 and the IL-12 receptor beta-1, and some of the functions for IL-23
00:12:30.12 were ascribed to IL-12 at the time.
00:12:33.03 But there's been a lot more clarity in the last few years, with the discovery of IL-23.
00:12:38.17 And what we now know is that these receptors both signal through the JAK-STAT pathway
00:12:45.05 of signaling... of signaling molecules.
00:12:47.23 The JAKs are cytoplasmic tyrosine kinases that are engaged by the different receptors.
00:12:53.15 And they transphosphorylate to become activated, and then phosphorylate different types of
00:12:58.19 STAT proteins, which are transcription factors, which, when they are tyrosine phosphorylated,
00:13:03.19 form dimers that then translocate to the nucleus and activate a variety of sets of genes.
00:13:09.04 In the case of IL-12, it activates the STAT4 transcription factor.
00:13:14.26 In the case of IL-23, it activates the STAT3 transcription factor.
00:13:19.14 And many of the other cytokines that I'll tell you about, such as interleukin-6,
00:13:23.28 also activate STAT3, but each of these receptors that utilizes STAT3 also has distinct signaling components
00:13:29.11 to target particular genes.
00:13:33.10 STAT4 can also, under some circumstances, be activated by IL-23.
00:13:38.20 And that is when IL-23 is engaged in pathogenic processes in vivo.
00:13:43.21 And I'll tell you a little bit about the IL-23 function in the next... in the next few slides.
00:13:50.16 The key experiment that introduced the concept of Th17 cells and showed the importance
00:13:55.10 of interleukin-23 was from the laboratory of Dan Cua at DNAX in 2003.
00:14:01.20 What they did was to use a model that's widely used for multiple sclerosis,
00:14:08.16 called experimental autoimmune encephalomyelitis.
00:14:11.11 And this is a model that I'll be talking about throughout my presentation, because it's
00:14:16.23 often used because it's fairly rapid and it's fairly robust.
00:14:20.23 In this kind of a model, a myelin protein is injected into mice along with adjuvant.
00:14:26.22 And typically, within about two weeks, the animals began to develop paralysis.
00:14:31.03 And as you can see over here, in wild type mice the paralysis develops as expected.
00:14:36.17 And in mice that are deficient for p40, which, as you recall, is absent in... in...
00:14:44.21 leads to an absence of both interleukin-12 and interleukin-23, you can see these animals are protected.
00:14:50.24 But the surprise at the time was that mice lacking just IL-12, that were deficient for p35,
00:14:56.04 not only developed disease, but they actually had even more severe disease than the wild type.
00:15:02.03 Whereas mice deficient for p19 -- that... what was then the newly discovered component
00:15:07.15 of interleukin-23 -- were completely protected.
00:15:10.20 So, that then led to the important concept of the target of p19 as being the Th17 cell.
00:15:18.11 And these cells are critical for barrier defenses against a variety of different bacteria
00:15:24.14 and fungi, fungi including Candida albicans.
00:15:28.26 But the flip side of this is that these cells can also be highly pathogenic through
00:15:33.18 the production of their variety of cytokines.
00:15:37.19 And under inflammatory conditions, they can now produce not only IL-17 and IL-22,
00:15:44.01 but they can also make interferon gamma.
00:15:46.28 And these cells have been validated to be very important in many human diseases, particularly psoriasis.
00:15:52.23 In psoriasis, antibodies against interleukin-17A are very effective in therapy.
00:15:58.01 Also, a variety of different arthritides, psoriatic arthritis and ankylosing spondylitis
00:16:03.08 in particular, can be treated by blockade of interleukin-17 and interleukin-23.
00:16:10.05 And then there are also some other... there are some other suggestions that multiple sclerosis,
00:16:17.04 as well as a variety of inflammatory bowel diseases, are Th17-mediated diseases.
00:16:23.05 For IBD in particular, it is known that polymorphisms in the Th17 signaling pathway can contribute
00:16:31.02 to the disease, again adding further argument for a role for these kinds of cells in the disease.
00:16:39.01 There's also some evidence from animal models that Th17 cells in the mother can influence
00:16:45.22 the development of the fetal brain, if they are expressed in very high levels
00:16:50.08 and can cross the placenta.
00:16:51.19 This is only an animal model that has been looked at, but I will discuss this
00:16:58.15 in the second part of my presentation.
00:17:01.20 At the center of the Th17 differentiation process is a transcription factor, ROR gamma t.
00:17:08.03 And ROR gamma t is encoded by this locus, Rorc, in which two different isoforms
00:17:13.23 can be ex... can be transcribed, depending on the promoter that's used.
00:17:18.08 And the longer form is expressed quite broadly, and it's expressed in a circadian manner.
00:17:24.26 But the shorter form of ROR gamma t is expressed exclusively in lymphoid lineage cells.
00:17:32.16 And these include the cells that develop in the thymus, the double positive thymocytes
00:17:39.18 that require ROR gamma t for their survival, as well as lymph nodes and Peyer's patches,
00:17:46.19 secondary lymphoid organs that develop in the fetus and that require the lymphoid tissue inducer cells
00:17:51.17 that are dependent on ROR gamma t.
00:17:54.27 In this slide, I show a crystal structure of ROR gamma t, of the ligand binding domain,
00:18:00.03 which regulates its function.
00:18:02.19 So, this is a nuclear receptor, very similar to estrogen receptor and glucocorticoid receptor.
00:18:08.21 And it is thought that it's regulated by ligand, but the precise ligand has yet to be defined.
00:18:14.07 What we know is that molecules in the cholesterol biosynthetic pathway are very effective
00:18:22.16 at regulating ROR gamma t function.
00:18:25.15 But we don't yet have strong genetic data to tell us which of these intermediates
00:18:32.11 are important in vivo, and in particular, in the differentiation of the different cell types,
00:18:37.07 whether there may be different ligands that are involved.
00:18:40.08 Now, in... beyond the function and development, the transcription factor, of course,
00:18:46.15 is required for Th17 cell differentiation.
00:18:50.03 It's also plays... playing a role in the differentiation of the induced regulatory T cells
00:18:58.07 found in the intestine, in particular in response to microbiota.
00:19:02.22 And I'll talk more about that in the second part of my presentation.
00:19:07.02 Also, there are gamma delta T cells that are specialized to make IL-17, and that's
00:19:12.12 also dependent on ROR gamma t.
00:19:14.11 And then there are the innate lymphoid cells, and these include the lymphoid tissue inducer cells,
00:19:18.15 both those that are involved early in the fetus, in lymphoid development,
00:19:23.25 but also those that appear... that develop postnatally and that are involved in some of
00:19:32.19 the tertiary lymphoid tissues.
00:19:34.07 I'll talk a bit more about the type III innate lymphoid cells a little bit later.
00:19:39.00 But I want to come back to the Th17 cells.
00:19:41.17 And this experiment that we did a bit more than a decade ago showed the importance of
00:19:48.08 ROR gamma t in the EAE model.
00:19:51.15 So, you can see here that animals deficient for expression of ROR gamma and ROR gamma t
00:19:57.08 are highly protected from EAE.
00:20:02.01 And you can see that there are very few IL-17 producing cells, shown by the FACS analysis,
00:20:07.14 over here.
00:20:08.24 These are cells in the central nervous system.
00:20:11.27 On the other hand, in the wild type mice, you see not only that there are IL-17-producing cells
00:20:17.03 but also cells that make both interferon gamma and interleukin-17.
00:20:21.12 And it turns out these are very important cells, because these are those cells
00:20:25.12 that are found in pathological situations, in which there is tissue destruction and autoimmunity.
00:20:33.21 I will discuss two pieces of evidence that suggest a critical role for interleukin-23
00:20:39.05 in the generation of these cells that produce both interferon gamma and interleukin-17.
00:20:46.00 One of the experiments is from an in vitro model for EAE, in which it is possible to
00:20:51.08 differentiate cells into Th17 in vitro, and if these cells are specific for a myelin protein,
00:20:59.14 then they can be injected into mice.
00:21:01.21 And within two weeks the animals get EAE.
00:21:03.28 So, people typically use transgenic mice in which the transgene for the T cell receptor
00:21:11.06 is for a T cell receptor that recognizes a myelin protein.
00:21:16.02 And typically, most people use, in vitro, interleukin-6 and TGF-beta to differentiate cells
00:21:22.13 into Th17 cells that make interleukin-17 and interleukin-22.
00:21:29.27 But it turned out that under these conditions these cells did not induce EAE.
00:21:35.04 On the other hand, when interleukin-23 was included, it was possible, now, to get EAE.
00:21:40.13 And it was even possible to do so in the absence of TGF-beta, just by including interleukin-1 beta
00:21:46.04 instead of TGF-beta.
00:21:47.15 In that case, again, EAE could be induced.
00:21:51.10 And so the interleukin-23 molecule is critical in this process.
00:21:55.18 And this is a very elegant experiment that was done by Brigitta Stockinger's laboratory in England,
00:22:01.21 in which they showed the importance of IL-23 in the generation of these Th17 cells
00:22:07.24 that make interferon gamma.
00:22:09.23 What she did was a fate mapping experiment, in which the yellow fluorescent protein is
00:22:15.13 knocked into a ubiquitous locus, but it's only expressed when a transcriptional stop signal
00:22:20.18 is excised by the action of Cre recombinase.
00:22:23.28 So, she bred these mice to animals in which the Cre recombinase was knocked into the IL-17a locus,
00:22:32.10 so only those cells that make IL-17a will have the capacity to then express YFP.
00:22:38.00 So, upon expression of Cre, the stop signal is excised.
00:22:41.27 YFP is expressed for the life of that cell, even after the cell stops making IL-17.
00:22:48.14 So, what the group then did was to gate on just those YFP-positive CD4 cells in the model of EAE,
00:22:56.13 looking now at what happens.
00:22:59.02 And of course, in the absence of interleukin-23, there is no EAE.
00:23:03.10 But they were able to now look at these YFP-positive cells and see what their phenotype was.
00:23:08.19 And you can see here in wild type mice, many of these cells, now, express interferon gamma,
00:23:14.10 or a combination of interferon gamma and interleukin-17A.
00:23:17.15 But these cells are absent in the absence of interleukin-23.
00:23:22.26 So, that really provided the critical piece of evidence that these double-producing cells
00:23:32.06 are important in the pathogenesis in this model.
00:23:35.12 But there's evidence that it's important in pathogenesis in other models as well.
00:23:39.12 So then, what is the role of TGF-beta?
00:23:42.02 Because, as I showed you from the in vitro experiment, it seemed like TGF-beta was
00:23:46.10 not necessarily needed to be able to induce EAE.
00:23:50.23 But on the other hand, if... when people looked in animals in which the TGF-beta receptor
00:23:57.22 was ablated, they saw that there was no EAE, compared to the wild type mice, here.
00:24:02.19 So, how could one explain this?
00:24:04.17 Well, a recent paper by Zhang et al has begun to shed some light on what TGF-beta is doing
00:24:11.15 during differentiation of Th17 cells.
00:24:15.18 And typically, when interleukin-6 signals by itself, through phospho-STAT3, there is
00:24:21.24 no expression of ROR gamma t and no IL-17 that is made.
00:24:26.05 But that appears to be because the... one of the targets of the TGF-beta signaling pathway, SMAD4,
00:24:32.22 recruits a co-repressor complex that includes the SKI molecule, which itself brings
00:24:39.11 in the histone... histone deacetylases.
00:24:41.24 And the histone deacetylases basically shut down chromatin at the ROR gamma t locus.
00:24:47.26 But when TGF-beta is applied, that now leads to a degradation of SKI, of SKI, and basically
00:24:57.22 the relief of the histone deacetylase function, and activation of transcription of ROR gamma t.
00:25:03.09 And now interleukin-17 and other cytokines can be produced.
00:25:09.27 We don't really know yet whether some of the other targets of the TGF-beta signaling pathway,
00:25:14.26 like phosphorylation of SMAD2 and 3 have some positive effects here, but what's clear is that
00:25:20.14 the TGF-beta relieves this negative function.
00:25:23.26 And indeed, what you can see here, in this very nice experiment...
00:25:29.04 if, in the TGF-beta receptor knockout mouse, one introduces also a mutation,
00:25:34.02 a deletion of the SMAD4 locus, now EAE can be restored
00:25:39.09 because Th17 cells can now differentiate in the absence of TGF-beta.
00:25:44.03 So it appears, then, that, in least in vivo, TGF-beta is also important in autoimmunity,
00:25:50.09 although in... even though it doesn't appear to be necessary in the in vitro models.
00:25:58.14 So, what I've just shown you suggests, then, that there are two different types of Th17 cells.
00:26:05.17 Those that can differentiate in the absence of interleukin-23, and that can make interleukin-17A and F,
00:26:12.03 and IL-22, and these we can call homeostatic or non-pathogenic Th17 cells.
00:26:18.03 They typically are induced by microbiota, and they're found at barrier surfaces.
00:26:24.00 On the other hand, in various cases of inflammation, which can be found in different types of tissues,
00:26:29.21 interleukin-23, along with IL-1-beta, but at least in some cases probably aided by TGF-beta,
00:26:35.24 leads to the differentiation of these cells that can make not only the Th17 cytokines
00:26:41.10 but also Th1-like cytokines, like interferon gamma, and that can contribute, now, to disease.
00:26:47.23 And these kinds of pathogenic Th17 cells have also called... been called Th1* cells in human.
00:26:55.26 And I'll summarize for you what we currently know about these Th17 cells, be they non-pathogenic
00:27:01.22 or pathogenic.
00:27:02.22 So, the homeostatic cells are typically induced by the commensal microbiota and protect
00:27:08.18 the mucosal barriers.
00:27:10.02 And they produce not only IL-17a and f and IL-22, but also interleukin-10,
00:27:14.24 which is an anti-inflammatory cytokine.
00:27:18.00 But the pathogenic Th17 cells are induced by selected microbial pathogens,
00:27:24.13 or what we call pathobionts, which can live in our bodies under normal circumstances
00:27:29.09 without causing disease, but then can be stimulated to cause disease in some circumstances.
00:27:35.02 They participate in the autoimmune diseases, dependent on interleukin-23.
00:27:39.10 And in addition to the Th17 cytokines, they make these other cytokines, particularly interferon gamma.
00:27:45.15 And the human equivalent was described by Federica Sallusto.
00:27:49.24 These are cells in the circulation that produce interferon gamma, but they also express target genes
00:27:54.08 for both Th1 and Th17 transcription factors, such as CXCR3 and CCR6,
00:28:02.06 which are chemokine receptors there are targets of Tbet and ROR gamma t, respectively,
00:28:08.16 the Th1 and Th17-specifying transcription factors.
00:28:14.11 So, there are many mutations that have been now described in humans that give us
00:28:20.16 some clues about the Th17 pathways.
00:28:23.04 There are individuals who have chronic cutaneous candidiasis, and oftentimes there are mutations
00:28:31.24 in interleukin-17, interleukin-17 receptor, as well as the signaling components
00:28:37.06 that have been identified.
00:28:39.02 And then there are a few patients who have been described who have disseminated mycobacterial infections
00:28:44.16 after immunization with BCG, which is used as a vaccine for tuberculosis.
00:28:50.07 And it was found that ROR gamma t mutations can account for this, in which there is
00:28:55.24 loss of both the Th1* cells and Th17 cells, even though there's no effect on Th1 cells
00:29:02.24 that can serve an anti-viral function.
00:29:05.22 And some of these mutations are depicted here in this slide.
00:29:09.25 The STAT3 mutations, both gain-of-function and loss-of-function, have been described.
00:29:15.24 And also a gain-of-function mutation in STAT1, which blocks the differentiation of Th17 cells
00:29:22.27 while promoting Th1 cell differentiation.
00:29:26.18 Those kinds of mutations can also account for some of the cases of candidiasis.
00:29:32.20 Mutations in IL-17...
00:29:34.23 IL-17F have been described, as well in the IL-17 receptor A and an adaptor molecule,
00:29:41.24 ACT1.
00:29:42.24 And polymorphisms in the IL-23 receptor are some of the most commonly associated polymorphisms
00:29:49.18 in inflammatory bowel disease, again providing evidence of the importance of this pathway
00:29:55.20 in autoimmune diseases.
00:29:58.13 There are many different therapeutics that are currently being brought to the market
00:30:02.09 to target this pathway.
00:30:04.06 There are therapeutics at target IL-12/IL-23, that is, p40, or just IL-23 alone, p19.
00:30:12.22 And these are very effective in psoriasis and some forms of arthritis.
00:30:16.27 Then, there are antibodies that target interleukin-17 alone, or a combination of IL-17A and IL-17F.
00:30:25.00 And again, these have been very effective for psoriasis.
00:30:28.13 And finally, there are antibodies that target IL-17 receptor A.
00:30:33.07 Now, these are a little bit more complicated, because IL-17RA is shared by the IL-17A and F cytokines,
00:30:41.02 with other cytokines, IL-17E... which is also called IL-25, which is made by... not only by type II...
00:30:49.04 which acts on type II innate lymphoid cells that I'll tell you about...
00:30:54.07 and also IL-17C acts on the IL-17RE receptor, which is found primarily on epithelial cells.
00:31:00.26 So, one needs a word of caution here, that we don't yet know that much about
00:31:06.15 the biology of some of these other cytokines and receptors.
00:31:09.25 And this particular molecule targets all of these.
00:31:12.28 ROR gamma t is also being targeted for therapeutic purposes, for the obvious reason that
00:31:18.12 it's upstream of these different cytokines, but that has not yet been tested clinically.
00:31:23.26 So, ROR gamma t, if... when it is targeted, is going to affect not only the Th17 pathway,
00:31:30.20 but also innate lymphoid cells.
00:31:32.24 And I'm going to tell you just a little bit about these innate lymphoid cells that
00:31:36.01 have been described only during the past decade.
00:31:38.18 And we believe that these may have been early evolutionary precursors of the
00:31:43.09 differentiated types of T helper cells.
00:31:45.16 So, you can see here that there are innate lymphoid cells -- ILC1, 2, and 3 --
00:31:51.19 that mirror in their transcription factors those transcription factors found on the Th1, Th2, and Th17 cells.
00:31:58.02 In addition, there are innate lymphoid cells that make both ROR gamma t and Tbet.
00:32:04.10 And these are cells that have an NK, natural killer, cell surface marker, NKp46.
00:32:10.18 And these seem to resemble very closely the pathogenic Th17 cells, the Th1* cells
00:32:18.16 that I was mentioning.
00:32:20.12 And these cells also express cytokines that are very similar to those expressed by the...
00:32:26.13 by the T helper cells.
00:32:28.15 And I won't dwell on this, but keep in mind that the T helper cells and the innate lymphoid cells
00:32:36.12 can share many different functions.
00:32:39.11 I'll show you one example here, for type II innate lymphoid cells, in which these are cells
00:32:44.18 that have a very important function in protection from parasitic worms, or helminths.
00:32:50.16 And what was found was that a very highly specialized cell in the intestinal epithelium,
00:32:55.28 in the tuft cell, responds to some product of helminths and also of protozoa by
00:33:03.13 producing interleukin-25 or IL-17E.
00:33:07.10 And that acts on a receptor on type II innate lymphoid cells.
00:33:10.25 And these cells will now make interleukin-13, which acts back on the intestinal stem cells,
00:33:16.05 leading to production of more tuft cells as well as goblet cells, secretory goblet cells,
00:33:21.12 which are very important for expulsion of the... of the parasites.
00:33:28.18 In addition, the type II innate lymphoid cells have another receptor that selectively expressed
00:33:33.19 on these cells.
00:33:34.24 It's a receptor for a neuropeptide, neuromedin-U.
00:33:38.13 And that also is important in this type of regulation.
00:33:42.02 An example of the expansion of these tuft cells is shown here, from work in Richard Locksley's lab,
00:33:48.01 in which they infected mice with a parasite, Nippostrongylus brasiliensis.
00:33:52.08 They generated mice in which the red fluorescent protein was knocked into the IL-25 locus.
00:33:58.02 And you can see, here, red fluorescence in rare tuft cells present, here, in uninfected mice,
00:34:03.22 but a great expansion of the number of cells making RFP following infection.
00:34:09.20 And this can also be shown by using a tuft cell-specific marker, showing expansion over here.
00:34:15.24 But this expansion of tuft cells is probably not only dependent on their making IL-25,
00:34:21.18 but there is also an effect on the type II innate lymphoid cells by neurons in the intestine,
00:34:26.19 by the enteric nervous system.
00:34:28.00 The enteric nervous system consists of many different cell types that are distributed
00:34:33.18 throughout different plexi in the layers of the intestine, and they can communicate locally
00:34:37.28 as well as extrinsically with the central nervous system through the vagus nerve.
00:34:44.16 You can see here, with a pan-neuronal marker, that these neurons can extend into the villi
00:34:51.23 in the small intestine.
00:34:54.07 And it was found that there's a class of neurons that respond to helminthic infection,
00:35:00.14 through detection of patterns from these... from these organisms, through the innate immune responses.
00:35:08.19 And they then produce the neuropeptide neuromedin-U.
00:35:13.21 These are cholinergic neurons that also make acetylcholine, but neuromedin-U acts on this receptor
00:35:18.04 on type II innate lymphoid cells, leading to production of interleukin-13.
00:35:23.14 And again, interleukin-13 leads to expansion of goblet cells and tuft cells.
00:35:27.25 And that leads, now, to more rapid expulsion of the parasitic worms, which are found in
00:35:35.23 both the intestine and also in the lung.
00:35:39.16 There's another example, in type III innate lymphoid cells from work that Jhimmy Talbot
00:35:44.06 in our laboratory has been doing.
00:35:46.06 And what he noted was that neurons in the small intestine, here stained in red,
00:35:54.00 are in very intimate contact with type III innate lymphoid cells found within structures called
00:35:59.09 cryptopatches.
00:36:00.09 These are sentinel posts just beneath the mucosal layer, beneath the epithelium in the intestine,
00:36:07.14 and these respond to commensal microorganisms.
00:36:11.13 And you can see that there are very close contacts made by neurons and these
00:36:16.08 type III innate lymphoid cells, which are marked here with a knockin of the green fluorescent protein
00:36:21.03 at the ROR gamma t locus.
00:36:23.07 And what we now can appreciate is that these cells are specialized for making the neuropeptide
00:36:29.24 vasoactive intestinal peptide, VIP.
00:36:33.23 And what VIP does is to act on receptors that are selectively expressed on
00:36:40.27 type III innate lymphoid cells, inhibiting their production of cytokines, particularly of interleukin-22.
00:36:47.21 So that... we think that these are neurons that respond in a way to maintain homeostasis
00:36:53.10 and prevent too much activation of this pathway, which can lead to hyperproliferation of the epithelium,
00:36:58.21 and potentially can lead to damaging consequences.
00:37:02.19 So, what I hope that I've shown you here is that the homeostasis at the mucosal barrier
00:37:10.01 involves not only a couple of lymphoid cells, but also epithelial cells,
00:37:15.13 various types of myeloid cells that make a variety of different cytokines, as well as enteric neurons,
00:37:21.17 whose relationship to each of these different cell types is only now beginning to be elucidated.
00:37:27.03 So, we are now in a very exciting period in which the various relationships between
00:37:32.26 the different cell types can begin to be explored in much greater detail.
00:37:37.25 So, I will stop there and acknowledge those in our laboratory who contributed to what
00:37:42.09 I showed you here.
00:37:43.15 Teruyuki Sano did much of the work on the Th17 cell induction in the... in the intestine.
00:37:50.02 Jhimmy Talbot worked on the... on the VIP-positive neurons in the gut.
00:37:55.24 And Ivaylo Ivanov, who is now at Columbia University, did the very important work, early on,
00:38:00.18 showing the relationship between ROR gamma t and Th17 cells.