Session 3: Bridging Innate & Adaptive Immunity

Transcript of Part 3: The Role of Toll-Like Receptors in the Control of Adaptive Immunity

00:00:14;28	Hello.
00:00:15;28	My name is Ruslan Medzhitov.
00:00:16;28	I'm a professor at Yale University School of Medicine and an Investigator of the
00:00:21;10	Howard Hughes Medical Institute.
00:00:23;11	And I will discuss, today, our work on characterizing toll-like receptors and their role in control
00:00:32;21	of adaptive immunity.
00:00:36;26	My story started in this building.
00:00:40;05	This is a Library for Natural Sciences of the Russian Academy of Sciences in Moscow
00:00:47;18	when I was a graduate student there from 1990 to '93.
00:00:53;05	At that time, the Soviet Union just collapsed and there was a profound economic crisis,
00:00:59;05	so there were no reagents to... to be able to do experimental work, and there were
00:01:04;14	not even periodicals available to read.
00:01:07;23	And this was the only library that still had a subscription to Nature, Science, Cell,
00:01:14;03	and other journals.
00:01:15;22	And so I would go there every morning and spend all day in this library on the second floor, and...
00:01:22;09	just trying to read the literature, since I wasn't really able to do much in the lab
00:01:28;05	at the time.
00:01:30;21	And during one of these trips to the library, I completely randomly ran into a paper written
00:01:37;22	by the late Charles Janeway that changed the direction of my career and life.
00:01:46;22	And this was a paper in... that appeared in the Proceedings of the Cold Spring Harbor Laboratory
00:01:53;20	from 1989.
00:01:57;02	And this is the paper.
00:01:59;02	It was entitled "Approaching the Asymptote? Evolution and Revolution in Immunology".
00:02:03;24	And this is Charlie Janeway, who unfortunately passed away in 2003.
00:02:09;22	But what he described in that paper was a new perspective, a new theory to think about
00:02:16;11	the immune system that, when I read it, I thought that this is... makes so much sense
00:02:23;20	and this is going to completely revolutionize our understanding of immunity.
00:02:29;20	And I got completely enchanted with this idea and this logical framework that Charlie proposed there.
00:02:39;19	And that was the time when I became interested in immunology.
00:02:46;14	And to put this into historical context, what the idea was about is how the immune system
00:02:54;21	knows when to be activated and when to respond to a challenge.
00:03:02;25	And what was thought at the time is that the immune system -- T cells and B cells --
00:03:11;16	only react to foreign antigens but not to self-antigens, for example,
00:03:16;01	antigens that come from our own tissues.
00:03:19;02	And so on the top panel, you'll see here a situation where self-antigens that may be
00:03:26;27	presented by antigen-presenting cells -- so, APCs -- to T cells, but there would be
00:03:32;24	no response.
00:03:34;08	Even though T cells can see the antigen, they should not be responding.
00:03:40;13	And at the lower part, you see the situation where there is an infection.
00:03:44;18	And when APCs detect pathogens, then there should be response.
00:03:48;24	As we all know, we respond to pathogens.
00:03:52;20	But it wasn't clear what makes the difference.
00:03:54;18	And again, one idea was that the immune system distinguished between self-antigens and non-self-antigens.
00:04:01;13	So, pathogens would be non-self, the immune system will react; self-antigens are self,
00:04:07;26	and the immune system would not react.
00:04:10;22	But it... there was also known at the time that for T cells to become activated
00:04:18;17	they require two signals.
00:04:20;05	And one signal is the antigen itself, which is presented by MHC molecules,
00:04:26;18	major histocompatibility complex molecules.
00:04:29;15	So, it's a complex between MHC and antigen that is seen by T cells.
00:04:35;19	And what became known at the time, from the work Ron Schwarz and Mark Jenkins at the NIH,
00:04:42;22	is that when T cells see only signal 1, there is no response that is generated,
00:04:49;27	but when they see two signals -- the second signal also coming from antigen-presenting cells --
00:04:54;21	then there would be a response.
00:04:57;11	And the identity of that second signal was subsequently characterized, and it's
00:05:03;02	known as molecules called B7-1/B7-2, or CD80/CD86.
00:05:09;27	So, that was the state of knowledge, and it was... many different views about what controls
00:05:18;12	activation of the adaptive immune response, of T cells and B cells.
00:05:22;25	And there were different sorts of schools of thought and... and it was still very much
00:05:28;10	unclear, for somebody who's coming from the outside of the field, what was going on and
00:05:34;22	how the system worked.
00:05:36;26	And that's what was changed with this article by Charlie Janeway, where he proposed that,
00:05:43;02	in fact, what might be happening is that in addition to T cell receptors and immunoglobulin receptors
00:05:48;19	there is a whole different set of receptors, that are shown here in blue,
00:05:56;01	that Charlie hypothesized would be involved in direct detection of pathogens, and that
00:06:04;08	these receptors would then control expression of the signal 2.
00:06:09;09	And in this manner, when antigen-presenting cells encounter self-antigens or any other
00:06:15;02	antigens that are not infectious, there would be just signal 1 and there will be no response.
00:06:21;04	But when they encounter pathogens, then there would be this detection, by these hypothetical receptors,
00:06:26;08	of common microbial structures and that would lead to the induction of signal 2,
00:06:33;13	and that would result in the immune response.
00:06:36;00	And that was a very profound insight and it was a very beautiful idea that made so much sense.
00:06:42;17	And... and Charlie further proposed that these receptors, that he called pattern-recognition receptors,
00:06:48;18	because they detect conserved patterns found in pathogens... things like lipopolysaccharides,
00:06:54;16	peptidoglycans, and so forth... he proposed that these receptors are evolutionary ancient
00:07:00;09	and they are involved in immunity in invertebrates as well as in vertebrates, where they control
00:07:08;04	the innate arm of the immune system.
00:07:11;05	But in addition, in vertebrates they acquire a second function where they control activation
00:07:15;15	of the adaptive immune system.
00:07:17;03	And that was a really revolutionary proposition.
00:07:22;15	And the only problem was that the receptors that Charlie hypothesized should exist were
00:07:29;19	not known at the time.
00:07:31;14	And... and so, after reading Charlie's paper, I started communicating with him,
00:07:36;23	initially through email, and eventually I was fortunate enough to be able to join his lab as a...
00:07:42;22	as a postdoc in 1994.
00:07:45;27	And then my focus of my research was on trying to identify such receptors that can
00:07:52;17	control expression of signal 2 to induce... to control activation of adaptive immunity.
00:07:58;28	And at that time, there was really little known about how microbes can be recognized directly.
00:08:05;15	There were a few proteins that were known to bind to microbial cell walls.
00:08:08;25	One of them was a protein called mannan-binding lectin from the complement system.
00:08:14;02	Another is CD14 molecule, which is a GPI-anchored protein on macrophages.
00:08:19;20	And there were a few proteins known in invertebrates that were involved in recognition of LPS.
00:08:27;21	But other than that, it was really not clear what these receptors are supposed to
00:08:31;17	look like and how they... how to identify them.
00:08:37;06	And a couple of things that changed, that played a role in the development of the story,
00:08:43;21	was that... one was that in 1994, the year when I joined Charlie's lab, David Baltimore
00:08:48;20	came to Yale to give a talk in our department, and at that time he just... his lab just started
00:08:55;20	characterizing the first knockouts for NF-kappaB, which is a critical transcription factor
00:09:00;26	involved in inflammation.
00:09:03;12	And there he talked about the phenotype of the first NF-kappaB knockouts, which had
00:09:08;16	both defects in innate and in adaptive immune response.
00:09:12;26	And I remember, after David's talk, Charlie and I were talking in the hallway,
00:09:19;15	and we both thought that whatever it is that we're looking for, it's... whatever these receptors are,
00:09:25;10	they probably work by activating NF-kappaB.
00:09:30;22	And then we thought, okay, so one criteria we should use is that they... they are likely
00:09:34;15	to be NF-kappaB activators.
00:09:37;07	And at that time, the only two receptor families known to activate NF-kappaB where the TNF family
00:09:43;22	and IL-1 receptor family.
00:09:47;01	And IL-1 receptor is the one that we got particularly interested in because of this remarkable conservation
00:09:58;05	in IL-1 receptor signaling portion with a receptor that at the time was already known
00:10:06;08	to exist in Drosophila.
00:10:09;08	And in Drosophila that receptor is called Toll.
00:10:11;20	That receptor was identified and cloned in Kathryn Anderson's lab by Carl Hashimoto.
00:10:19;00	At the time, they were at Berkeley.
00:10:21;11	And the signaling pathway was elucidated by several investigators, including Steve Wasserman
00:10:28;03	and Mike Levine, Kathryn Anderson and others.
00:10:32;10	And it was known that this pathway operates in fly development.
00:10:37;15	But what was interesting for us is that it worked through NF-kappaB in flies, and the
00:10:42;15	related receptor, IL-1 receptor, worked through NF-kappaB in mammals.
00:10:47;17	And what they shared is the cytoplasmic portion, shown here in blue, whereas their ectodomains,
00:10:54;13	involved in ligand recognition, were unrelated.
00:10:59;06	And because of this conservation with flies, and because IL-1 receptor is involved in inflammation,
00:11:09;10	we thought that perhaps the receptor that we are looking for would be... would have
00:11:15;03	this C-terminal signaling domain, and the ectodomain perhaps would be something that
00:11:21;06	can recognize microbes.
00:11:22;26	And the only structures that were known at the time to recognize microbial components
00:11:28;20	were C-type lectins.
00:11:30;22	So I thought that maybe there is another version of this receptor that has cytoplasmic signaling domain
00:11:35;15	from Toll and IL-1 receptor, and the ectodomain would be from C-type lectin.
00:11:40;08	And... and that was my strategy to try to identify it.
00:11:44;28	And... that... what I took advantage of is that, after many, many failed attempts
00:11:52;01	to do it through various types of approaches... degenerate screening or by hybridization or
00:12:00;22	through PCR... took advantage of the fact that, at the time, new types of genomic sequences
00:12:07;19	started to be become available -- these so-called expressed sequence tags -- that were
00:12:13;10	just short sequences that were generated randomly in multiple tissues.
00:12:17;22	And using bioinformatics approaches and using conserved consensus sequences for the cytoplasmic domain
00:12:23;26	for Toll and IL-1 receptor, as well as for C-type lectins, I started searching
00:12:29;04	these databases and found several clones that corresponded either to C-type lectins or to
00:12:35;12	this cytoplasmic domain.
00:12:37;21	And then I pursued both of them.
00:12:40;02	And the one that corresponded to C-type lectin domain, I found that it was most similar to
00:12:46;15	Drosophila Toll.
00:12:48;22	And this was in some... starting in January, February in 1996.
00:12:55;27	And the... the gel shown on the left here was one of the first clones that I got,
00:13:06;24	which was... later would turn out to be a human homologue of Toll.
00:13:11;15	And the gel on the right in this bottom band...
00:13:15;17	I still remember that band because this was one of those happy moments in the lab
00:13:21;05	when I got this portion of the receptor, which was particularly difficult to clone.
00:13:27;09	And then by doing standard library screening through hybridization, we pulled out a full-length
00:13:34;24	of the cloning... coding Toll receptor followed by 5' RACE technique, which was at the time
00:13:44;21	popular in gene cloning, and eventually got the receptor.
00:13:50;21	And... by about May of '96.
00:13:54;18	And then another event that happened that... in the... in the summer of '96, we had a meeting
00:14:00;21	in Charlie's summer house in Annisquam, near Boston, where several other investigators
00:14:06;09	participated that shared Human Frontier Grant, including Alan Ezekowitz, Fotis Kafatos,
00:14:13;00	Jules Hoffmann, and Charlie.
00:14:15;07	And... and Jules Hoffmann then presented work of Bruno Lemaitre from his lab about
00:14:20;13	genetic work on Drosophila Toll pathway, where they found it was involved in immunity.
00:14:25;01	So, that was... it was further confirmation for our expectation that this receptor
00:14:34;04	might be involved in immunity.
00:14:36;08	And so, then I characterized this receptor in terms of its ability to control NF-kappaB signaling,
00:14:46;22	shown here on the right.
00:14:49;09	And more importantly for us at the time, for its ability to induce inflammatory cytokines.
00:14:54;20	And the most important for us, the holy grail for us, whether this receptor can induce
00:14:59;02	the second signal, or costimulatory molecule, a channel known as B7.
00:15:04;11	And that was the... the probably... one of those lucky moments in the lab when I saw
00:15:10;01	that Toll receptor, which I rendered constitutively active because the ligand was not yet known
00:15:16;24	at the time, and transfected in this monocyte cell line.
00:15:22;04	And when I saw that it can induce expression of this gene, that was the moment when
00:15:27;09	we thought, okay, this... this whole hypothesis, now, we can connect from the beginning to
00:15:32;03	the end.
00:15:33;11	And I even called Charlie... this was late in the evening, I even called Charlie and
00:15:37;03	told him that that human Toll can induce B7 expression.
00:15:40;15	That was... we probably would be the only two people in the world who would be
00:15:44;14	excited about that, but nevertheless others also appreciated it.
00:15:50;03	And so we published this paper in '97, where... which is an extremely simple paper showing
00:15:55;08	just cloning, and showing these couple of functional assays.
00:15:59;28	And... and then, subsequently, in the same year, Ben Lewin, who was the editor of Cell at the time,
00:16:07;26	asked Charlie to write a mini-review about innate immunity.
00:16:12;14	And so in that mini-review we discussed questions about evolution of innate and adaptive immunity,
00:16:23;05	and put this schematic together to... to show the parallels and differences between
00:16:28;24	Drosophila Toll and mammalian Toll pathways, where we suggested that recognition of pathogen-associated
00:16:37;13	molecular patterns -- things like LPS and teichoic acids and so on -- is detected by Tolls,
00:16:44;02	either directly or through some intermediary, as is the case in Drosophila.
00:16:49;05	And that leads to induction of NF-kappaB and various innate immune response and genes
00:16:54;07	genes controlling adaptive immunity.
00:16:57;04	And subsequent work indeed demonstrated that the Toll that I worked on is part of a bigger family.
00:17:06;23	There are about a dozen different Toll receptors in mammalian species.
00:17:13;04	And it's now known as TLR4 -- Toll-like receptor 4 -- and it's a receptor for LPS, and the
00:17:18;06	receptor part that detects LPS... actually, a protein called MG2 that complexes with Toll.
00:17:25;09	And subsequent work with... by many investigators elucidated specificities of different
00:17:32;24	Toll receptors.
00:17:33;24	And Shizuo Akira from Osaka University, in particular, played a major role in this
00:17:39;11	set of investigations.
00:17:42;02	And what we know now is that... if we... our main interest was not so much in microbial
00:17:50;15	specificity of receptors but in this idea that microbial receptors can control activation
00:17:56;02	of adaptive immunity.
00:17:58;08	And this is what is schematically illustrated here.
00:18:00;21	So, we have a dendritic cell, which is a cell type that normally activates T cells.
00:18:07;13	And when dendritic cell encounters pathogens, what's shown on the left side, here,
00:18:11;24	there is an endocytic receptor that will take up pathogen, take it into lysosomes, and then
00:18:16;17	proteins will be cut into peptides and loaded into MHC molecules and presented to
00:18:23;19	T cell receptors.
00:18:25;22	But that information by itself is not sufficient for T cells to know whether they should become
00:18:30;12	activated or not, because this peptide can come from pathogens or it can come from innocuous
00:18:37;26	food antigens or it could be even a self-antigen.
00:18:40;22	So, T cells have no way...
00:18:42;20	no way of knowing what the origin of the antigen is.
00:18:46;22	And the reason for that is because T cell receptors are generated at random.
00:18:50;13	And each T cell has one single specificity, but they are random, so it doesn't know
00:18:56;12	what it's specific for.
00:18:58;16	So therefore, there is a need for another signal, so-called signal 2, which will provide
00:19:04;09	information about the origin of the antigen that T cell is specific for.
00:19:09;15	And that is provided by this detection of microbial structures by Toll receptors
00:19:14;25	-- and now we know, several other families of pattern-recognition receptors --
00:19:18;25	that detect those structures and induce expression of costimulatory molecules.
00:19:23;15	And now a T cell that has specificity for peptide derived from pathogen will also sec...
00:19:28;28	have a confirmation signal from a costimulatory molecule -- CD80/CD86, also known as B7-1/B7-2 --
00:19:38;04	and then the T cell will become activated.
00:19:40;16	And in addition, Toll receptors and other pattern-recognition receptors will
00:19:45;03	induce production of cytokines such as IL-12 that will tell T cells what kind of effector response
00:19:51;11	to generate.
00:19:52;21	And that is now, of course, a well-accepted view of how immune response is activated.
00:20:03;20	And... and it's a... it's very satisfying to see this confirmation of the proposal
00:20:11;22	by Charlie Janeway, which was at the time largely ignored.
00:20:16;03	And that now it's become... it's a... it's a textbook knowledge in the... of the immune system,
00:20:22;25	that that's how activation of the adaptive immune system is controlled.
00:20:27;15	And Toll receptor was one of the first receptors to be demonstrated to... to be involved
00:20:34;17	in this process.
00:20:35;17	So, that's the story of this discovery.
00:20:38;24	And thank you for listening.