Session 5: B Cells: Development, Selection, and Function

Transcript of Part 1: Early B Cell Development: A Look at the Defining Questions in Immunology

00:00:07;06	I'm gonna be talking today about some of the defining questions in immunology.
00:00:11;15	And I'm gonna take you back to the 19th century.
00:00:17;07	So, the Franco-Prussian war ended in 1872 with the siege of Paris.
00:00:25;13	And this rivalry between the French and the Germans carried on into the field of science
00:00:31;28	for the next few decades.
00:00:33;20	Two of the most important investigators of the time were Robert Koch in Berlin and
00:00:41;00	Louis Pasteur in Paris.
00:00:42;25	And their laboratories had helped establish the germ theory of disease, showing us
00:00:47;24	for the first time that disease was actually caused by pathogens, by microbes, and not by some
00:00:53;28	unknown imbalance of bodily humors.
00:00:57;19	The pace of discovery was truly remarkable.
00:01:00;24	To just cite one example, in 1884, Loeffler in Berlin described the Bacillus that causes diphtheria.
00:01:09;15	In 1887, Roux and Yersin in Paris showed that these bacteria secreted a toxin
00:01:15;21	into culture supernatants, and this was a lethal toxin.
00:01:19;06	In 1890, Brieger and Frankel, in Koch's laboratory in Berlin, showed that they could
00:01:25;16	inject the stocks in... into small animals and showed some form of immunity.
00:01:31;00	Later that same year, also in Koch's laboratory, Shibasaburo Kitasato and Emil von Behring
00:01:38;24	chemically inactivated the diphtheria toxin, injected it into large animals in increasing doses,
00:01:44;10	and showed the presence of a principle in the serum of these animals which they called
00:01:50;15	an antitoxin, which could neutralize the diphtheria toxin.
00:01:55;22	Within a year of the... of this discovery, children who were dying -- gasping for breath
00:02:01;11	from diphtheria -- were being miraculously saved by the injection of this antitoxin.
00:02:09;19	Serum therapy caught on all across the world.
00:02:12;02	This was a remarkable new form of therapy, which was the first scientific approach
00:02:18;10	to actually treating a disease.
00:02:19;28	A young man called Paul Ehrlich returned to the laboratory from Egypt, came back to Koch's laboratory.
00:02:27;08	He saw all this excitement about serum therapy and antitoxins.
00:02:31;28	And he realized, while his colleagues had made a remarkable discovery, a life-saving discovery,
00:02:36;22	what they had missed was an underlying principle.
00:02:41;28	And the underlying principle was that when a foreign substance is injected into a vertebrate
00:02:49;11	it creates complementary molecules.
00:02:53;27	Ehrlich would name these molecules that were injected into animals as antigens.
00:03:00;17	And the complementary molecules, which were then produced by the recipient animal,
00:03:05;24	he called antibodies or antikorpers.
00:03:09;17	Ehrlich was a remarkable scientist.
00:03:12;03	He would make many discoveries in his lifetime.
00:03:14;07	As a medical student, he described the existence of mast cells.
00:03:18;25	He had devised stains, and he looked and found a new cell type which he then described,
00:03:23;27	which is an important cell in immunity.
00:03:26;06	He would go on to describe antigens and antibodies.
00:03:29;12	He would also describe a phenomenon that he called horror autotoxicus, or autoimmunity.
00:03:35;04	He predicted that there might be immune phenomena where we attacked ourselves.
00:03:39;25	He's the first person to think about the immune surveillance of cancer.
00:03:44;12	He was also the first person to start the field of chemotherapy.
00:03:48;22	The entire field of cancer chemotherapy and antibody therapy owes a debt to Ehrlich,
00:03:54;15	who came up with the first chemotherapeutic.
00:03:57;22	But his most remarkable contribution was to think about a model for how the immune system
00:04:04;24	worked.
00:04:06;03	And this he called the side chain hypothesis.
00:04:09;15	And in a sense, you have to think about a time 60 years before the fluid mosaic model
00:04:16;01	of the membrane is known, more than 65 years before we have the first polypeptide
00:04:22;12	hormone receptor described, and Ehrlich comes out with the model where he imagines
00:04:26;26	an immune cell has antibodies on the surface as receptors.
00:04:31;25	And then he visualizes antigens coming, triggering the immune cell -- so, triggering a receptor --
00:04:37;04	and then the cell induces more of the same antibody.
00:04:40;21	So, this remarkable hypothesis, which is described in this slide, is to suggest, as Ehrlich did,
00:04:49;02	that we already have all these different antibodies on the cell surface and that an antigen
00:04:54;28	comes by, interacts with one of these an... anti... antibodies, and triggers the release of
00:05:00;28	the same antibody into the blood.
00:05:03;13	Now, at this time, when he came out with this model, Ehrlich had also done experiments
00:05:09;15	to show that he could modify a chemical at a single atom and he could create a new antibody.
00:05:17;17	Karl Landsteiner, who was one of Ehrlich's greatest opponents, intellectually, had done
00:05:22;10	similar experiments.
00:05:24;04	He had taken chemicals, also modified them at a single atom, and he had made new antibodies.
00:05:29;12	So, there were antibodies which are highly specific and could recognize
00:05:33;26	a single atom difference between two molecules.
00:05:36;27	It was clear that there was an immense number of antibodies.
00:05:42;04	The number of antibodies within us was probably infinite.
00:05:46;26	How could you possibly conceive of the fact that we already have all these antibodies
00:05:52;01	within us?
00:05:53;03	That was what Ehrlich was suggesting.
00:05:55;00	So to reframe this, the clear picture that emerged at that time was, how can we
00:06:02;04	create complementary structures?
00:06:04;03	Are these pre-existing?
00:06:06;10	Or have they to be induced?
00:06:09;01	To give you an analogy, imagine you have a hundred people showing up for a job at UCSF,
00:06:14;11	or they are going to start at UCSF today, or maybe at Google or wherever you want.
00:06:19;25	Each one of these people goes to an office and gets an ID card made.
00:06:24;24	And the ID card requires that he has a photograph taken, and then it gets laminated,
00:06:30;01	and then you have an ID.
00:06:32;02	Now, this is a model where you show up, you get your photograph taken
00:06:36;26	-- so, this is an induced model --
00:06:39;15	and then you get your ID.
00:06:41;27	Imagine on the other hand, if you showed up at Google and they already had your picture
00:06:46;22	on file.
00:06:47;22	In fact, they had your great-grandchildren's pictures on file, and your grandfather's pictures
00:06:52;06	on file.
00:06:53;06	They had pictures on file for everybody who is around in the world today, or who will
00:06:58;05	ever be around in the universe.
00:06:59;19	And that was the model Ehrlich was proposing, that we already have all the antibodies
00:07:04;26	within us before an antigen shows up.
00:07:07;22	And this seemed inconceivable, so most thinkers in the field moved away from Ehrlich,
00:07:14;18	and the first people to actually formulate this, they were Haurowitz and Breinl, who came up
00:07:20;07	with the model for a direct template.
00:07:23;11	Linus Pauling, the great chemist, took this further and essentially suggested that antibodies
00:07:28;10	came from one of these pink primordial proteins within the cell.
00:07:32;28	An antigen enters the cell.
00:07:34;28	The antigen then reacts with the antibody.
00:07:37;05	And the... the antibody, the preformed antibody, folds around the antigen, forms a new shape,
00:07:43;02	and then is secreted.
00:07:44;21	And this was the model that existed 'til 1957, because people couldn't accept that you could
00:07:50;26	already have pre-existing repertoires which went into 10^12 different possibilities,
00:07:58;18	and you have these many different antibodies within you.
00:08:03;05	In the 1950s, Jerne would first suggest that maybe there's a model where we already
00:08:08;12	have all the antibodies secreted from us and provided some data for that.
00:08:13;11	And then David Talmage would come up with another model, which is now called
00:08:18;13	the clonal selection hypothesis.
00:08:19;20	And this was further refined by... by Frank MacFarlane Burnet.
00:08:24;04	And the model essentially states that we already have a range of immune cells,
00:08:30;10	each with a different receptor.
00:08:33;07	An antigen comes by, identifies one of those cells, the cell gets triggered
00:08:38;10	-- so, this is now a clone of lymphocytes -- the clone expands -- so, we have clonal expansion --
00:08:45;04	and then after clonal expansion we have these cells secrete antibodies, which are then going to
00:08:50;08	be the effector molecules of the immune system.
00:08:53;01	So, this was a model, which is now the accepted model of the immune system.
00:08:57;11	We... we already have pre-existing immune cells, each with a different receptor.
00:09:04;14	And then antigen comes and triggers one of them.
00:09:06;18	And this holds true both for T cells and for B cells.
00:09:11;09	When the clonal selection hypothesis was put forward, we didn't know what the immune cells
00:09:18;03	for adaptive immunity were.
00:09:20;20	At this time in 1957, we thought these were just macrophages, or some macrophage-like cell.
00:09:28;02	We would soon learn that there were different cells that mediated these functions for the immune system.
00:09:35;14	So by the 1960s... so, antibodies were discovered in 1890... but by the 1960s, we have
00:09:43;17	the first ideas of the structure of antibodies.
00:09:46;23	And this is from the work of Gerald Edelman and Rodney Porter.
00:09:50;17	So, Edelman would show that antibodies are made up of two chains.
00:09:53;24	So, we have these two pink chains, which are the heavy chains, and these two orange chains
00:09:57;26	are the light chains.
00:10:00;06	And these were associated and linked to each other by disulfide bridges.
00:10:05;10	Porter would actually show that the antibody had different domains, and this is shown
00:10:11;06	on the next slide.
00:10:12;06	So, this is a much more modern slide.
00:10:13;24	And you can... you'll note that in the antibody there are heavy chains and light chains,
00:10:19;04	but they can be cleaved by certain proteases to give you fragments called Fab fragments.
00:10:24;24	So, Fab fragments are basically the part of the antibody that binds antigen.
00:10:30;10	And Fc fragments, which are really the tail of the antibody.
00:10:36;05	Crystal structures would then tell us the structure of an antibody domain.
00:10:41;13	And this will all come in the '70s.
00:10:43;08	So, if you looked at an antibody domain... so, here we have a schematic view of a domain,
00:10:50;01	in which you can notice that there is basically a ribbon.
00:10:53;05	So, it's a beta sheet, which is folded over to form a beta barrel.
00:10:59;12	And sticking out on top are some loops, labeled here a CDR1, CDR2, and CDR3.
00:11:06;08	Okay?
00:11:07;08	So basically, you have a ribbon, the ribbon is joined together in a fold, and then
00:11:13;04	each strand of the ribbon is linked the next by a loop.
00:11:16;28	And when you looked at the sequence of antibodies -- so, if you looked at this part of the slide --
00:11:21;10	you'll notice that there are three regions.
00:11:23;16	We've compared 100 antibodies... there are three regions of the antibody sequence
00:11:29;02	where the sequence varies a lot.
00:11:30;22	And these are called CDR1, 2, and 3.
00:11:34;00	And these correspond to those three loops that stick out from the top of the Ig domain.
00:11:39;24	So, immunoglobulin domains can be variable domains like this, with the CDRs,
00:11:44;19	or they can be constant domains, which make up the rest of the antibody molecule.
00:11:49;12	So, if you think about it differently, the CDRs are like fingers.
00:11:54;19	You have three fingers for the variable domain of the light chain, three fingers for the
00:12:00;20	variable domain of the heavy chain.
00:12:02;25	And the fingers can come close together to bind a small molecule, or they can
00:12:06;20	splay out widely to form a surface that can accommodate a protein.
00:12:10;27	So, on the right, you can see there is actually, over here, an antigen bound to an antibody.
00:12:16;03	It's a large protein antigen.
00:12:17;21	So, you can assume that CDR1, 2, and 3 -- the fingers -- have splayed out to form
00:12:23;00	a big surface to accommodate the antigen.
00:12:27;17	This is another view of the Fab part of an antibody, shown here in yellow and blue.
00:12:33;11	And all the red residues that you see on the Fab correspond to those six fingers,
00:12:39;20	CDR1, 2, and 3 from the heavy chain and CDR1, 2, and 3 from the light chain.
00:12:45;01	One the other side is hen egg lysozyme, in green.
00:12:49;03	And there too, the red residues are the residues on hen egg lysozyme that interact with
00:12:54;21	the red residues on the antibody, on the Fab of the antibody.
00:12:59;12	The glutamine that's shown in purple is actually going to burrow deep into the groove in
00:13:03;22	the middle of the red residues in the antibody.
00:13:09;02	Antibodies have many functions.
00:13:10;19	And most of... most of you are aware of them.
00:13:13;20	They can neutralize viruses and toxins.
00:13:16;22	So, neutralization means that the antibody binds to the virus or to the toxin and
00:13:23;05	prevents it from entering our cells or binding to a receptor.
00:13:27;12	Antibodies can medi... mediate opsonization.
00:13:30;17	So, opsonization really means that a pathogen might be coated with an antibody, and then
00:13:36;27	a receptor on the phagocyte -- which is for the Fc part of the antibody, so, the tail of the antibody --
00:13:43;17	recognizes the antibody and ingests the pathogen.
00:13:47;15	We call that opsonization.
00:13:49;24	There's another phenomenon called ADCC.
00:13:52;13	Now, in ADCC, you have a virally infected cell.
00:13:56;26	An NK cell recognizes an antibody that's coating the virally infected cell.
00:14:03;28	And this triggers a receptor on the NK cell, which then causes the killing of the
00:14:08;23	virally infected cell.
00:14:10;11	And so that's another function of antibodies, where the antibody coats the target cell,
00:14:15;01	and then an NK cell kills it.
00:14:17;04	And then finally, one of the other functions of antibodies is mediated through complement.
00:14:22;00	So, the complement proteins can lyse the microbes -- so, this is over here --
00:14:27;03	or fragments of complement can serve as opsins and help internalize pathogens.
00:14:32;26	And you can also have complement fragments which drive inflammation.
00:14:38;07	To summarize this part of the talk, antibodies protect you.
00:14:45;02	But antibodies kill.
00:14:47;24	Autoantibodies can make you very ill.
00:14:52;22	Antibodies coat pathogens; they coat the infected cell.
00:14:55;20	With NK cells and phagocytes, they give those microbes hell.
00:15:01;24	When antibodies fix complement, the shit really hits the fan.
00:15:05;27	You'll be blown to bits, little microbe.
00:15:09;05	Run away, if you can.
00:15:11;27	You take a set of beta strands, you get a beta-pleated sheet.
00:15:16;12	Fold the pleated sheet in two, then the barrel is complete.
00:15:20;04	Clip it with a disulfide bond and you have an Ig fold.
00:15:26;22	Found even in archaebacteria, this domain is old.
00:15:32;25	Loops make connections between the beta strands.
00:15:37;16	If you want to make an Fab, please use both your hands.
00:15:43;01	The loops are like fingers.
00:15:44;22	They stick out at the top.
00:15:48;14	Complementarity-determining regions, the cream of the crop.
00:15:54;16	Three fingers from the heavy, three fingers from the light, come together to create
00:16:01;25	an antigen-binding site.
00:16:04;07	Bring the fingers close together, you get a cleft with a purpose.
00:16:08;12	Splay the fingers widely, you get a protein-binding surface.
00:16:15;13	Antibodies protect you, but, baby, antibodies kill.
00:16:21;22	Autoantibodies can make you very ill.
00:16:25;22	Oh, you Y-shaped globular proteins, with a name that Ehrlich gave,
00:16:32;02	if the good doctor could hear this doggerel, he'd be turning in his grave.
00:16:38;21	The next question I'm going to turn to was long known as the GOD Question, and GOD refers to
00:16:46;11	the generation of diversity.
00:16:49;14	It was impossible, until the late 1960s, to imagine how this question would ever be answered.
00:16:56;09	The question essentially was this.
00:16:58;00	We know we have, now, about 20,000 genes.
00:17:01;14	But we can make maybe 10^14 different antibodies.
00:17:06;00	If you believe that one gene gives rise to one polypeptide, how is this ever going to
00:17:12;18	be conceived of being possible?
00:17:15;08	How can you use a limited number of genes to give you proteins that can recognize
00:17:21;20	10^14 different things?
00:17:23;14	So, this question boggled everybody's imagination.
00:17:26;28	It became one of the central questions of biology, because no one could understand
00:17:31;23	how this could actually happen.
00:17:33;08	And this question remained so mysterious that it was assumed -- I think this was assumed by the early 1970s --
00:17:41;13	that it would be one of those things that would never be solved.
00:17:45;07	We would never know the answer.
00:17:46;13	So, it was a religious question, okay?
00:17:49;23	Now, we did have some idea about antibody structure.
00:17:52;26	I told you that Porter and Edelman had out with the structure of two heavy chains and
00:17:57;24	two light chains, and the existence of light chains was understood.
00:18:01;25	There were plasmacytomas.
00:18:03;08	So, Henry Kunkel's lab at Rockefeller had described a lot of plasmacytomas.
00:18:07;12	These are tumors that make a single antibody.
00:18:10;11	So, there were... there was the ability to have a pure antibody, a pure light chain,
00:18:15;20	to try to sequence it.
00:18:17;17	But by the early '60s, people still couldn't sequence an entire light chain protein.
00:18:22;19	I mean, insulin had been sequenced, but this was a difficult task.
00:18:26;23	And many groups were trying to sequence light chains.
00:18:30;02	Now, Edelman... his mentor was Henry Kunkel, and Kunkel and Edelman didn't quite get along.
00:18:36;26	And there was a great meeting that was going to be held in California in Warner Springs
00:18:40;10	-- I mean, I... to me it's like Woodstock when I think about Warner Springs --
00:18:45;10	and everybody who was anything in biology... from Seymour Benzer, to Chris Anfinsen,
00:18:49;23	everybody was invited
00:18:51;03	to come and discuss and describe to others what they could think about how we
00:18:56;11	created diversity in the immune system.
00:18:58;26	So, Mel Cohn was the organizer of this meeting, and he got a call from Rodney Porter,
00:19:03;28	who was in Oxford.
00:19:05;25	And Rodney Porter made this call saying, there is this postdoc at Rockefeller,
00:19:11;07	you don't know who he is, but I've heard about his existence from Henry Kunkel.
00:19:16;14	And this postdoc actually was with a... worked with a friend of Kunkel's called Lionel Craig.
00:19:20;11	And he said, you should call him.
00:19:22;12	His name is Norbert Hilschmann, and he'll have something interesting to tell you.
00:19:26;14	Call him to your meeting in Warner Springs.
00:19:28;20	So, at Warner Springs, many people went and presented their knowledge, what they knew
00:19:35;06	about antibody light chains.
00:19:36;27	They had peptide maps.
00:19:38;06	They couldn't quite figure out what the maps told them.
00:19:41;10	They didn't have the sequence of a light chain.
00:19:43;26	The heavy chains were too difficult.
00:19:45;19	They were too big.
00:19:46;24	And then there was this talk from Norbert Hilschmann, this unknown person, never seen before,
00:19:52;17	never heard of before.
00:19:54;20	He shows up, and he gives his talk.
00:19:58;06	And on his first slide -- and these days... those days you had real slides -- he showed
00:20:02;20	the complete sequence of two antibody light chains.
00:20:06;19	Okay?
00:20:07;19	So, we have two antibody light chains.
00:20:10;27	And the remarkable thing about the sequence was that the light chains were
00:20:15;16	identical in sequence for most of the molecules, but the top third, the top... the N-terminal parts
00:20:21;15	were different.
00:20:23;16	This was a remarkable finding.
00:20:25;28	Everyone was excited.
00:20:27;03	For the first time, there was some sense about how antibodies differed from each other.
00:20:34;00	People stopped Hilschmann.
00:20:35;12	They asked him to go back to his earlier slides.
00:20:38;22	But Hilschmann did not.
00:20:41;20	He moved rapidly through the rest of the slides, he was not collegial, and he left the meeting.
00:20:47;21	And that's probably the reason why he was not the third recipient of the Nobel Prize,
00:20:53;15	along with Porter and Edelman.
00:20:55;06	Because he made a remarkable discovery.
00:20:58;05	But Hilschmann disappeared from public view.
00:21:00;24	He was somewhat concerned that his data would be taken up... taken on by others and so on.
00:21:06;09	But he didn't interact.
00:21:08;11	But someone in the audience listened carefully to this talk.
00:21:11;15	His name was Bill Dreyer, who's a polymath.
00:21:14;23	He's no longer around, but he was at Caltech as a professor, and an assistant professor
00:21:19;05	at the time.
00:21:20;09	And he looked at the data and said.
00:21:22;03	I understand how diversity is created.
00:21:26;26	So, he went back to his lab.
00:21:29;19	Most people didn't understand what he was trying to explain, but someone in his lab
00:21:33;06	understood him.
00:21:34;15	And they sat down and wrote a paper together.
00:21:36;25	And that's called the Dreyer and Bennett hypothesis, published in 1965.
00:21:42;07	And if I can give you an analogy to explain what the hypothesis says, imagine that you
00:21:47;22	have a... a lady with a strange wardrobe.
00:21:51;28	She has just one black skirt, but maybe she has a thousand different tops.
00:21:58;10	So, by mixing and matching a thousand tops with one black skirt she has a thousand outfits.
00:22:05;05	Now, imagine that she has twenty different, very different-looking belts.
00:22:09;16	Again, by mixing and matching these, she could get you twenty thousand different outfits.
00:22:14;20	And this is essentially what Dreyer and Bennett postulated, that genes might come in pieces,
00:22:21;09	and then you have these cassettes, and you can join together different variable cassettes
00:22:26;10	with one constant cassette and create different antibodies in different cells.
00:22:31;11	Okay?
00:22:32;11	You can prove this.
00:22:33;13	It was impossible to prove this at the time.
00:22:35;25	The recombinant DNA revolution hadn't occurred.
00:22:39;21	Finally, when recombinant DNA couldn't be performed in California or in... or in the
00:22:46;20	city of Cambridge, Massachusetts, Susumu Tonegawa went off and worked at
00:22:52;10	the Basel Institute of Immunology, and he did the crucial experiments
00:22:56;26	to show that immunoglobulin genes come in pieces.
00:23:00;08	This is just a Southern blot showing that DNA is that of a different size, so the DNA
00:23:07;19	for the antibody genes is of a different size in all cells other than B cells.
00:23:12;01	So, in a liver cell it's different from B cells.
00:23:14;19	And he did a slightly different experiment, but I'm using this Southern blot to show you this,
00:23:18;22	to show you that DNA had been cut and joined, and something had been done to it
00:23:24;15	in a B cell.
00:23:25;15	And this was Tonegawa's big discovery.
00:23:27;23	Okay?
00:23:28;23	So, when you look at immunoglobulin genes, we now know that these genes come in pieces.
00:23:34;00	We have a whole range of V segments, and then we have D segments and J segments.
00:23:39;20	We have these different settings for the heavy chain.
00:23:41;27	For the light chains, also, we have V segments and J segments.
00:23:45;09	And these can be joined together in different cells.
00:23:47;26	So... and I'll illustrate this in the next few slides.
00:23:50;15	So, this is the big picture view.
00:23:52;21	You have a bunch of V segments, D segments, and J segments upstream of a constant region segment.
00:24:00;01	And then these... we're gonna join together one V, one D, and one J to create an antibody gene.
00:24:07;02	And then at the junctions between the V's and D's, and D's and J's, we can
00:24:11;01	add or remove bases and create some junctional diversity.
00:24:14;10	And then when this DNA is transcribed, you're going to get a messenger RNA which actually
00:24:18;21	has this rearranged DNA encoding the variable part of the antibody.
00:24:24;00	And this is what's going to give you an antibody gene.
00:24:26;20	Okay?
00:24:27;20	So again, this is another view of light chains, where we have only V segments and J segments.
00:24:34;04	So, now the V's are going to be joined to the J's.
00:24:36;01	And seen below, you have V-kappy-29 joined to J-kappa-3.
00:24:40;21	And that creates a V-J exon.
00:24:44;07	And that's going to be upstream of the constant exons and give you a kappa light chain,
00:24:48;27	for instance.
00:24:50;03	Okay?
00:24:51;08	This is a view that...
00:24:52;18	Tonegawa looked carefully at the sequences of the antibody genes.
00:24:57;12	And what he was trying to ask was, how do you know where to cut?
00:25:00;22	How does the cell know where it should cut a V segment and a J segment so they can
00:25:05;10	be joined together?
00:25:06;18	And what he showed was that adjoining the constant... adjoining the coding region
00:25:11;22	-- so, if you look at the V in... in orange -- you'll notice that there is a 7-base pair sequence.
00:25:17;28	It's called a heptamer, CACAGTG.
00:25:20;20	Then, that's followed by a spacer, in this case it's about 12 base pairs.
00:25:25;14	And then it's followed by a nonamer, which is nine base pairs.
00:25:29;00	But the heptamer is always constant.
00:25:32;01	It's always CACAGTG.
00:25:34;08	And this is going to be joined to another gene segment downstream.
00:25:38;28	So, that's the one you see in green, down there.
00:25:41;21	So, the green segment has also next to it a CACAGTG, which is also the heptamer.
00:25:49;03	A spacer... this time, the spacer has 23 base pairs.
00:25:52;13	And then you have a 9-base pair region which is AT-rich.
00:25:56;00	So, this sequence, of a heptamer, a spacer, and a nonamer, was called a
00:26:02;04	recombination signal sequence.
00:26:04;13	And the spacer corresponds to either 12 base pairs -- that's one turn of the DNA helix --
00:26:10;02	or 23 base pairs -- it's two turns of the DNA helix.
00:26:14;16	And the rule was you always joined an RSS containing a 12-base pair spacer to
00:26:21;22	another coding segment which has an RSS which contains a 23-base pair spacer.
00:26:26;28	Never 12 to 12; never 23 to 23.
00:26:30;26	This assures that V joins to J, and V does... or D joins to J, and V doesn't join to J,
00:26:37;04	depending on the locus you're looking at.
00:26:40;16	Okay?
00:26:41;16	So, going back to this... so, the first stage of VDJ recombination... so, now,
00:26:45;21	in David Baltimore's lab, David Schatz and Marjorie Oettinger discovered RAG-1 and RAG-2.
00:26:51;25	And they showed that these two proteins actually helped the chromosome ribbon...
00:26:57;03	so, the first step of VDJ recombination, which requires RAG-1 and RAG-2, is it allows the chromosome
00:27:02;10	to ribbon between the two segments that are going to be joined.
00:27:05;07	So, you form a big loop.
00:27:08;00	And then you have synapsis.
00:27:09;13	Just bringing these two guys close to each other, though they were far away on the chromosome
00:27:13;19	to begin with.
00:27:15;17	Once this happens, RAG-1 then is gonna make a nick.
00:27:18;16	So, RAG-1 makes a nick, and the 3' hydroxyl then attacks the other strand and forms a hairpin,
00:27:26;05	so that the coding sequence now ends in a hairpin, whereas the signal sequence
00:27:31;04	with the CACAGTG is released as a clean double-strand break.
00:27:35;04	Okay?
00:27:36;04	So, the first step was synapsis, then there was cutting -- so we have synapsis, then we
00:27:42;00	have cutting... sorry, I have to go back to this slide.
00:27:44;25	The seconds step is cleavage.
00:27:46;24	The next step is opening up... you made hairpins, so you have a hairpin and you need to
00:27:52;11	open up the two hairpins, one from the coding region for the V, the other from the coding region for the J,
00:27:57;12	and join them together.
00:27:58;24	So, that's called hairpin opening and end-processing.
00:28:01;04	And the final step is repair, or ligation, where you join these pieces together.
00:28:06;15	Okay?
00:28:07;15	So, just to explain what happens in junctional diversity, you have two hairpins.
00:28:11;22	An enzyme called artemis cuts the hairpin maybe eccentrically.
00:28:16;00	So now you have a flap created, where you have DNA from the bottom strand going to
00:28:20;24	the top strand after the flap is created.
00:28:23;05	Polymerase fills in, so now you've filled in the gap.
00:28:25;25	So, those added nucleotides are called P nucleotides.
00:28:29;02	So, they were created in a templated manner.
00:28:31;26	And then at the blunt ends, we have another enzyme called TdT, terminal deoxynucleotidyl transferase,
00:28:38;05	which can add additional bases.
00:28:40;04	We call them N nucleotides.
00:28:41;15	So now, the two happens, instead of just joining them together, you actually have created
00:28:46;26	more diversity at the junction.
00:28:48;05	So, even if you join the same V and the same J in two different cells, the junctions
00:28:53;04	are going to be different.
00:28:54;22	Okay?
00:28:55;22	So, I'm going to summarize this part of the lecture.
00:29:00;15	CACAGTG, the generation of diversity.
00:29:06;14	A one-turn J kissed a two-turn V. They were brought into proximity by that lovely couple,
00:29:14;18	RAG-1/RAG-2.
00:29:16;01	RAG-1 says, I'm gonna cut you.
00:29:18;20	A pair of genes the RAGs do pick.
00:29:22;23	At each heptamer, they make a nick.
00:29:26;14	The 3' hydroxyl then must bend, to make a hairpin at the coding end.
00:29:32;21	Artemis cuts the pretty hairpin.
00:29:35;07	TdT puts N regions in.
00:29:38;17	It's time to shut the DNA door, bring in XRCC and ligase-IV.
00:29:45;14	CACAGTG, the generation of diversity.
00:29:51;01	Now you know your G-O-D.
00:29:54;19	Next time, all of immunology.
00:29:57;20	So, if you go back and think about B cell development in the context of understanding
00:30:03;21	VDJ recombination, we can understand that we have an early stage called a pro-B cell.
00:30:09;06	So, at the pro-B stage, you start to rearrange the antibody genes.
00:30:13;18	And you start with the heavy chain.
00:30:16;13	By the large pre-B stage, you have completed the rearrangement of the antibody heavy chain gene.
00:30:22;28	And you're gonna form a structure called the pre-BCR.
00:30:25;00	I'll come back to that.
00:30:27;16	Then the cell is going to go on, eventually, to become an immature B cell with IgM on the surface.
00:30:33;10	So, it has heavy chains and light chains.
00:30:34;27	This cell is going to emigrate from the bone marrow to the spleen, and then become
00:30:40;09	a general garden-variety follicular B cell.
00:30:45;04	So, one important checkpoint during B cell development is called the pre-BCR checkpoint.
00:30:51;07	So, when you go through VDJ recombination, you reach this point where you become
00:30:56;22	a large pre-B cell.
00:30:57;22	The large pre-B cell is a cell that has correctly rearranged the antibody heavy chain gene.
00:31:03;07	You know, when you add these bases to the junctions, you can go out of frame, so only
00:31:07;25	roughly half the cells are going to do this right.
00:31:10;01	So, the cells that have done it right on one chromosome are going to make a heavy chain protein,
00:31:14;09	they're going to make something called a pre-B receptor, and these cells are
00:31:19;15	going to survive and expand, and become a huge population of selected pre-B cells.
00:31:26;15	Each one of them will then... will then go on to rearrange a different light chain,
00:31:30;28	so that you now have B cells which have heavy chain and light chain.
00:31:34;09	And the pre-BCR checkpoint is very important in the context of B cell development and disease.
00:31:40;08	So, the pre-BCR... so, when the heavy chain is made, it associates with something called
00:31:46;03	surrogate light chains, which will be described in a subsequent lecture in some detail.
00:31:50;07	And it associates to form a receptor.
00:31:53;02	And this receptor signals constitutively.
00:31:55;03	The moment it's made, it's not looking for a ligand.
00:31:57;25	It's saying, you've done it right.
00:32:00;03	You have the right reading frame.
00:32:01;16	You deserve to live.
00:32:02;17	So, the signals cause the expansion and survival of the cell.
00:32:06;20	It also mediates a phenomenon called allelic exclusion, which I'll describe later in
00:32:11;10	a subsequent lecture.
00:32:14;13	So finally, the last question I'm going to talk about very briefly is about self-nonself recognition.
00:32:19;16	I'm going to give you a narrow view of this.
00:32:20;21	So, you create this diverse repertoire of B cells and T cells.
00:32:25;13	Each sees a different antigen.
00:32:27;08	But sometimes these are going to be self-reactive.
00:32:29;09	In fact, about 75% of the time they are self-reactive.
00:32:33;01	So, how do you get rid of the self-reactive guys?
00:32:35;25	I'm blood group A. If I make a blood group a B cell, it has the potential to kill me.
00:32:40;28	I need to do something about it.
00:32:42;21	So, one of the mechanisms that this happens during development is that at this immature B cell stage
00:32:48;11	you actually have this phenomenon of tolerance, or central tolerance.
00:32:54;08	And central tolerance in B cells is mainly mediated by a process called receptor editing.
00:33:00;13	And what I'm showing you here is that... imagine that this is a self-reactive B cell over there,
00:33:05;17	which has this orange light chain.
00:33:08;07	It sees a self-antigen -- let's say that's a red blood cell with that group A on it.
00:33:12;20	It triggers the cell and the cell then changes its light chain.
00:33:17;22	It no longer expresses the orange light chain.
00:33:20;13	It rearranges a new light chain, so now it has the yellow light chain.
00:33:24;08	And this combination of the heavy chain and the yellow light chain may be no longer self-reactive.
00:33:29;00	This process is called receptor editing.
00:33:31;03	It's a politically correct approach to tolerance.
00:33:34;00	You just don't bump off the self-reactive cell; you allow to reform itself.
00:33:38;24	Okay?
00:33:39;24	So, here you see in receptor editing, you'll notice that we have, let's say, V-kappa-29
00:33:44;04	has rearranged to V...
00:33:45;21	J-kappa-3.
00:33:46;24	But if I... if this cell were to edit, it might use V-kappa-25 to go and rearrange
00:33:52;02	to J-kappa-2, something down... or, sorry...
00:33:55;15	J-kappa-4 or J-kappa-5, something downstream of J-kappa-3.
00:33:58;20	And this would delete the bad light chain and bring in a new light.
00:34:02;08	This could also happen on the other kappa light chain chromosome, or it could happen
00:34:06;11	on a lambda light chain chromosome.
00:34:08;04	So, if a cell is self-reactive, it has a few opportunities to reform itself,
00:34:13;08	make a new light chain, and become no longer self-reactive.
00:34:17;11	So, this is one mechanism of central tolerance, which is in... in B cells, this is a major mechanism.
00:34:24;07	It's called receptor editing.
00:34:25;28	The other mechanism is deletion.
00:34:27;03	So, if you look at a big picture view of what happens during lymphocyte development
00:34:31;11	for B and T cells, we have assembly of receptors through VDJ recombination; then cells go through
00:34:38;16	an immature stage, when they are like teenagers, where they have to be tested;
00:34:42;20	and the ones which are self-reactive are going to be eliminated or edited.
00:34:47;23	And that's tolerance.
00:34:49;13	And then the cells are allowed to mature and become naive cells, go to the lymph nodes,
00:34:54;20	and be ready to do battle with pathogens.
00:34:56;13	So, in this lecture, we talked about the three central questions that have shaped immunology.
00:35:04;00	We first discussed the phenomenon of having pre-existing or induced antibodies.
00:35:10;25	How does a vertebrate actually make complementary shapes?
00:35:15;11	And we described this phenomenon as being explained best by the clonal selection hypothesis,
00:35:22;22	which is now an accepted fact in immunology.
00:35:26;13	We then asked the question... we have these different immune cells, each with
00:35:30;15	a different receptor... how do you create this incredible diversity?
00:35:34;22	And we answered by explaining that this has now been explained by VDJ recombination.
00:35:41;10	And finally, we asked, if you have this incredible diversity, which can see almost every shape
00:35:46;02	known to man -- and in fact, any shape that might be created in the next century as well,
00:35:51;05	we have receptors to recognize them -- how do you get rid of cells which are self-reactive?
00:35:56;07	How do you mediate this phenomenon of self-nonself recognition?
00:36:00;26	And we explained that in B cells the major way this was done in central tolerance
00:36:06;17	was through receptor editing.
00:36:07;18	There is another mechanism called deletion, which occurs in B cells but is much more important in T cells.
00:36:14;18	What we didn't talk about, because we didn't cover this, was the phenomenon of
00:36:19;24	peripheral tolerance, which happens after you've made your B and T cells.
00:36:24;00	And T regulatory cells, or regulatory T cells, are described as the cells that mediate a
00:36:30;19	peripheral lot of peripheral tolerance, which basically squelch self-reactive B and T cells in the periphery.
00:36:39;06	In the following two lectures, I'm gonna talk about some research results that go back to
00:36:45;13	the discovery, a few decades ago, of the pre-B receptor and of BTK signaling,
00:36:52;06	in the next lecture, which will explain some of the concepts of the early stages of B cell development.
00:36:59;06	And then, in the final lecture, I will talk about how we have used this kind of knowledge
00:37:04;10	about VDJ recomb...
00:37:06;06	VDJ recombination and signaling and other things to try to understand human disease.