microRNAs Create Regulatory Tension in Mammalian Blood Cells

Top

00:00:14;28 I'm David Baltimore from the California Institute of Technology.
00:00:21;07 And I want to talk today about microRNAs.
00:00:26;09 MicroRNAs were discovered back in the 1990s, but really only since the year 2000 have we
00:00:38;15 understood how many microRNAs there are and how important they are.
00:00:45;05 And I'm going to talk about them in the context of a particular kind of biology, which is
00:00:50;16 the biology of mammalian blood cells.
00:00:53;26 Because, in that context, you begin to see how important microRNAs are.
00:00:59;09 They're generally made as part of long precursors and cut out of those long precursors to
00:01:06;26 give rise to the 22-nucleotide microRNA.
00:01:14;13 We were interested in microRNAs because we had been interested in the process of inflammation.
00:01:21;02 Inflammation is something that occurs in humans and... and in all mammals it's a
00:01:25;25 response to infection, it's a response to... to all sorts of pathologic circumstances.
00:01:33;26 And we found, years ago, that a single family of transcription factors are responsible for
00:01:42;27 controlling inflammation.
00:01:47;09 So that when a cell is exposed to some kind of inflammatory stimulus, this family of
00:01:54;11 transcription factors turns on a whole panoply of genes that change the nature of the cell and
00:02:00;11 make the cell able to counter an inflammatory situation.
00:02:09;09 That involves the activation of something like 1000 genes.
00:02:13;03 And so this transcription factor, which is called NF-kappaB, is responsible for activating
00:02:20;00 some 1000 genes.
00:02:22;21 So, here's a... an indication of how you turn on NF-kappaB.
00:02:29;12 It's the exposure of cells so a whole range of things that are pro-inflammatory,
00:02:36;10 that are signals of inflammation: chemicals like IL-1 in the cell, or external things like
00:02:46;00 the lipopolysaccharide from a bacterium.
00:02:50;19 Those things all cause NF-kappaB to be released.
00:02:55;18 So, NF-kappaB ordinarily sits in cells with an inhibitor on it called I-kappaB, that orange thing.
00:03:03;07 And what all of these things do is cause the degradation of I-kappaB, the release of NF-kappaB,
00:03:10;26 which is two subunits.
00:03:13;04 It then goes to the nucleus and turns on these 1000 target genes.
00:03:19;15 We don't really know the number terribly well.
00:03:23;11 And that produces an inflammatory response and counters infection.
00:03:28;28 So, we wanted to understand whether microRNAs play any role in this process.
00:03:36;01 And years ago, we simply asked, was there any involvement of microRNAs in the
00:03:45;13 inflammatory response.
00:03:48;16 So, let's first talk about how microRNAs work.
00:03:52;11 MicroRNAs, and there's one shown here in this oval, one that happens to be illustrative...
00:04:04;10 microRNAs, again, very short, but there are about 8 nucleotides that are absolutely key.
00:04:13;25 Those 8 nucleotides allow a microRNA to interact with a messenger RNA.
00:04:19;17 So, here we have a messenger RNA.
00:04:22;03 It happens to be for a protein called IRF4.
00:04:27;05 The blue part of it is the coding region.
00:04:29;15 The black part of it is a non-coding region.
00:04:34;04 And buried in there is a target for a microRNA.
00:04:38;18 And the sequence of that target is shown down here.
00:04:41;18 It's exactly complementary to the 8 nucleotides on the microRNA.
00:04:46;28 So, a microRNA seeks out a messenger RNA, binds to it, and that prevents the translation
00:04:55;02 -- in fact, leads to the degradation -- of the messenger RNA.
00:04:59;13 So, microRNAs are inhibitors of messenger RNA function.
00:05:05;05 And they can be so effective that even though the cell is making a lot of a messenger RNA
00:05:09;22 it's making very little protein.
00:05:12;22 Because its messenger RNA is being degraded, really, at the same time as being...
00:05:18;08 or, right after it's being synthesized.
00:05:24;17 So, microRNAs were originally thought to be a curiosity that modulated gene expression
00:05:32;13 in some way.
00:05:33;13 But we now know that it's actually a partner in the whole regulatory process,
00:05:39;09 that microRNAs and transcription factors are part of a circuitry in cells that allow for precise decision-making
00:05:49;25 about how much of a given protein is made.
00:05:55;08 And so, we would like to know, then, for any given microRNA, what's its physiologic role?
00:06:03;04 That is, what are its targets and how do they fit into normal physiology?
00:06:08;16 And what proteins are controlled by it?
00:06:15;12 So, we did this experiment, as I said, many years ago.
00:06:21;08 And simply said, when there's an inflammatory process going on in a cell, are there
00:06:30;00 any microRNAs whose concentration goes up?
00:06:33;21 So, the inflammatory process was... was caused by lipopolysaccharide, which is a part of bacteria
00:06:43;28 that causes mammalian cells to excrete pro-inflammatory molecules.
00:06:52;02 And we just asked, of a couple of hundred microRNAs, all strung out here,
00:07:00;20 were any of them increased?
00:07:02;03 And three were increased.
00:07:04;14 They... those three are 132, 146, and 155.
00:07:12;23 And so we said, that's a nice number.
00:07:15;18 We can work on that.
00:07:16;18 That's not so big that we can't find enough graduate students to work on it,
00:07:22;16 and it's not so small that it's inconsequential.
00:07:26;16 And so let me tell you about these three microRNAs.
00:07:30;08 MicroRNA-132 is the first one I'll talk about.
00:07:33;27 It's highly conserved.
00:07:35;05 So, there's the sequence of it in many different animals.
00:07:40;03 They're not actually even all mammals.
00:07:43;13 And it's the same -- at least the seed sequence, that key sequence is the same -- in all of those.
00:07:51;14 And what it's doing, then, is searching out a messenger RNA.
00:07:57;23 And we'd like to know what messenger RNA that is.
00:08:01;09 But first, we characterized the nature of this microRNA.
00:08:05;24 We call them "meers", miR.
00:08:10;02 And we found, first of all, that it was enriched in the stem cells of the blood.
00:08:18;05 So, what's a stem cell?
00:08:21;20 This is a stem cell.
00:08:23;16 Stem cells are cells that give rise to all of the different elements of the blood.
00:08:32;26 And they give rise to it in a hierarchical differentiation scheme that leads to
00:08:37;22 lymphoid cells and myeloid cells and... and red blood cells and platelets.
00:08:44;12 But all of these derive from one stem cell.
00:08:50;17 It's the only cell that can renew itself in the blood.
00:08:53;28 It's actually found in the bone marrow, but then those cells go out into the blood.
00:09:00;09 And it does renew itself.
00:09:03;16 And so it's constantly making more of itself, but also making all of these downstream elements.
00:09:09;18 And that's a true stem cell: self-renewing, and giving rise to all of the elements of
00:09:17;23 a tissue -- in this case, a liquid tissue.
00:09:21;16 Now, how do we figure out whether a microRNA is able to influence these blood cells?
00:09:32;01 We do that by, first of all, expressing a large amount of the microRNA in bone marrow cells,
00:09:40;14 in these stem cells, and seeing what the consequence of that is.
00:09:44;24 So, we do that using virus vectors.
00:09:49;07 So, here's a mouse.
00:09:52;28 We... we treat that mouse to increase its number of stem cells.
00:09:58;18 We then sacrifice that mouse, take the bones -- long bones from the mouse --
00:10:03;20 extract the cells from that, and then we put into those cells virus particles that encode,
00:10:13;15 in this case, microRNA-132.
00:10:16;26 Now the stem cells are making large amounts of 132 and there's no control over it,
00:10:22;03 because we've put it in there under the control of the virus' genes.
00:10:29;22 We take those now modified cells and we put them back in a mouse.
00:10:36;03 But we've treated that mouse to kill off its ordinary bone marrow cells, blood-generating cells,
00:10:43;04 and now all the blood will be generated from these modified cells that we modified
00:10:48;26 in the laboratory.
00:10:50;14 We wait a few weeks/months, until the new cells have taken over.
00:10:58;14 And now we characterize the animal.
00:11:02;19 And so if we do that, one thing we see, for instance, is that, as that orange line shows,
00:11:13;09 there are more blood cells initially than in a control animal, as shown by
00:11:20;25 the black cell... line.
00:11:22;27 There's more cells for a while, for a couple of... couple of weeks.
00:11:28;21 And then suddenly the bone marrow begins to fail.
00:11:33;11 And by a few months, there is... there are many fewer blood cells than normal.
00:11:41;07 So, there's an enormous pathology generated by the overexpression of this microRNA.
00:11:51;28 We find increased numbers of cells over short periods of time,
00:11:59;13 and then decreased numbers of cells over longer periods of time.
00:12:06;08 That's the first clue as to what's going on.
00:12:09;15 But this overexpression method... very interesting, but not physiological.
00:12:20;02 And to really understand the physiology, what we want to do is ablate microRNA-132.
00:12:26;17 And you can do that.
00:12:27;17 You do that with genetic trickery.
00:12:32;05 And we ablate 132 and then we look at the animals.
00:12:35;25 And what we saw, particularly after about a year, was the same sort of pathology that
00:12:42;09 we saw by overexpression.
00:12:45;00 And that didn't make sense, because you'd think you'd get the complementary picture,
00:12:50;07 not the identical picture.
00:12:52;26 You get fewer blood cells.
00:12:54;27 Actually, the animals fail because of fewer blood cells.
00:13:00;06 But when we looked in the bone marrow at the stem cells themselves, what we saw was that
00:13:06;21 there weren't less of them; there were more of them.
00:13:08;23 It's just that they weren't differentiating.
00:13:11;07 They weren't giving rise to product cells.
00:13:16;21 So, if we compare overexpression and deletion, we can see that although they have
00:13:23;08 the same indication -- reduced numbers of cells -- they're for two different reasons.
00:13:30;06 In the overexpression case, you're wearing out the bone marrow.
00:13:34;15 They're producing more and more cells for a while, and then less and less because
00:13:39;01 they are literally worn out.
00:13:41;28 In the case of knocking out the gene, knocking out the mess... microRNA, you get
00:13:49;24 an increased number of stem cells, but they don't function right.
00:13:54;17 And we realized that there must be a key protein that was being affected by 132.
00:14:04;08 And so we looked for that.
00:14:06;04 And we looked for it by overexpressing the microRNA and just asking which gene was...
00:14:14;19 was downregulated.
00:14:15;22 And we found two.
00:14:19;20 One of them, a SOX gene, I won't talk about because it's part of another story.
00:14:25;04 It's interesting.
00:14:26;13 The gene that is interesting here is FOXO3.
00:14:30;17 FOXO3 -- and here's the messenger RNA for FOXO3 -- has a perfect sequence for interaction
00:14:38;24 with 132.
00:14:41;21 And 132 causes a major downregulation in FOXO3.
00:14:47;14 And you can see that here, where we look at the protein levels in overexpressing cells
00:14:55;02 or... or wild type cells.
00:15:00;14 And we can also overcome this effect by expressing FOXO3 without its regulatory region.
00:15:09;14 And if we do that, we overcome the inhibition.
00:15:13;20 And so the inhibition is evident in the orange bar, and the overcoming of it is evident
00:15:22;14 in the blue bar.
00:15:27;08 And so what happens is that 132 actually buffers FOXO3 expression.
00:15:36;15 And the way we look at it is this.
00:15:39;05 It's that if you have too much FOXO3 expression it's bad.
00:15:44;23 That's the knockout of 132.
00:15:47;12 And you don't get differentiation.
00:15:50;07 If we have too little FOXO3 -- that's with the overexpression of 132, and so we make
00:15:55;26 a lot of the inhibitory microRNA -- it's also bad.
00:16:00;27 Now the bone marrow works too hard.
00:16:03;13 FOXO3 is a brake that we're... we're taking off.
00:16:10;03 And so it's a... it's a sort of... and I'll...
00:16:13;02 I'll describe it later as a Goldilocks situation, situation in which you don't want too much
00:16:17;21 and you don't want too little.
00:16:19;04 You want just enough.
00:16:22;07 And so what I view it as is setting up a regulatory tension, a tension between too much and too little,
00:16:30;00 in which we have to find the midpoint.
00:16:34;01 And that's where the physiology is appropriate to the situation.
00:16:41;19 So, let me talk about the other two microRNAs that I introduced to you that are
00:16:50;03 increased by inflammation: 146, and I'll talk about 146a, which is the one that we actually study;
00:16:57;19 and 155.
00:16:58;19 Both of them are induced by inflammatory stimuli, but the work that we've done over the years,
00:17:05;11 and I'm only going to summarize it, has shown that they play opposite roles.
00:17:12;03 And you can see those roles here.
00:17:14;24 So, let me describe what this is.
00:17:18;18 Up here, we show a surface of a cell that's interacting with a pro-inflammatory stimulus,
00:17:26;07 in this case, as... as I've been describing, lipopolysaccharide.
00:17:30;27 That then sends a signal into the cell that is evident... or, that is activated by proteins
00:17:40;24 which we know about.
00:17:42;13 Those proteins are TRAF6 and IRAK1.
00:17:46;04 They lead finally to that IKK complex that I talked about earlier.
00:17:53;08 And that in turn strips off the inhibitor from NF-kappaB and allows NF-kappaB
00:18:00;09 to go into the nucleus.
00:18:01;20 And two of the genes that NF-kappaB activates are these microRNA genes, for 146 and 155.
00:18:12;18 But by careful dissection, we can show that the role of 146 is as a negative controller
00:18:20;01 of these intermediate proteins.
00:18:23;08 So, it downregulates these proteins and downregulates the stimulus
00:18:29;01 which has come from the lipopolysaccharide.
00:18:31;20 155, by contrast, upregulates this... the system actually by down-regulating inhibitors.
00:18:43;15 That's how it works, mechanistically.
00:18:47;02 And so it's a positive feedback regulator.
00:18:49;23 So, we have positive feedback and we have negative feedback coming from the same stimulus.
00:18:57;05 That sounds confusing, but what it leads to, again, is regulatory tension.
00:19:03;10 And let me show you how we know that.
00:19:06;05 So, the first thing we do is to knock out 146a.
00:19:10;11 Those animals are born normal, but over six months they develop a picture that looks very
00:19:16;07 similar to what happens with the 132 overexpression.
00:19:19;26 You... you overstimulate the bone marrow, you make too many cells, and finally
00:19:28;07 the bone marrow crashes and you stop making cells at all.
00:19:34;19 And what you...
00:19:35;23 actually, the hallmark of this is you get these huge spleens that form,
00:19:43;00 because the animal is trying to make blood cells in the spleen because it can no longer use
00:19:48;13 its bone marrow.
00:19:54;02 So, how does 146a work?
00:19:56;24 It works, as I told you, by inhibition of TRAF6.
00:20:01;17 I also mentioned that what TRAF6 is activate NF-kappaB.
00:20:07;11 That activates secretion of pro-inflammatory molecules like IL-6.
00:20:14;16 They stimulate hematopoietic stem cells.
00:20:18;14 And so if we knock out 146, which is an inhibitor, you stim... overstimulate the stem cells and
00:20:26;00 get too much product.
00:20:29;07 That's the nature of knocking out an inhibitor.
00:20:34;28 Now, how about 155?
00:20:36;23 I said it's a positive regulator.
00:20:40;21 And so what happens if we knock 'em both out?
00:20:44;08 And we did that by knocking out 155, crossing it with the 146 knockout, making animals
00:20:51;01 that had neither of those.
00:20:54;00 And we found that that suppressed the phenotype from 146.
00:21:00;01 So, over here we can see that 146 causes a big increase in IL-6 secretion and that,
00:21:09;26 if we knock out 155, as the green bar shows, that goes back to normal.
00:21:18;12 And there are many other parameters in which the same thing is evident.
00:21:22;00 In fact, if we infect those animals with a bacterium, they handle that bacterium better,
00:21:29;05 because they're making more cells that limit the infection, more proteins that limit the infection.
00:21:40;06 And if we look at 155 production in the knockout, which is the red bar, compared to the ordinary situation,
00:21:51;16 you actually get more 155.
00:21:55;17 Because it's in the feedback loop from... from 146.
00:22:00;25 So, 155 deficiency rescues 146.
00:22:08;12 Overexpression results also imply this.
00:22:10;22 And it implies that 155 is essential for the 146 phenotype.
00:22:19;09 And an experiment we should have done years ago we finally did when we realized all of this,
00:22:26;13 which is just look at what happens when you treat cells with lipopolysaccharide,
00:22:34;00 with a pro-inflammatory stimulus, to both 155 and 146.
00:22:39;24 155, as I showed you before, as is evident here in the black line, goes up and then comes down.
00:22:47;21 But 146 goes up much more slowly and stays up over time, literally over days.
00:22:57;20 And so the inhibitory influence is the dominant influence at long periods of time.
00:23:03;08 The stimulatory influence is the dominant effect at short periods of time.
00:23:10;00 And that allows us here to look at the same diagram I showed you before.
00:23:18;19 But now we see that negative feedback is slow and positive feedback is fast.
00:23:29;00 That then generates a regulatory tension, just as I showed you for 132.
00:23:34;19 But this is now a tension between two microRNAs, that evolves over time in this inflammatory milieu.
00:23:46;15 That perception, which we had some time ago, was reinforced by a comment that was made
00:23:57;08 in an article in the New England Journal of Medicine.
00:24:00;19 It was an article about something entirely different, something called the inflammasome,
00:24:07;14 which causes inflammatory bowel disease.
00:24:12;16 But in the process of describing this in the journal, they pointed out that there's a microRNA,
00:24:22;27 shown right here, which is actually inhibitory to the inflammasome.
00:24:28;25 And they described what was going on in the following way.
00:24:34;01 They said, in the aggregate, the evidence points to a relationship
00:24:39;13 between the health of your colon and inflammatory activation
00:24:45;14 that's reminiscent of the porridge in the story of Goldilocks.
00:24:49;25 Goldilocks and the three bears.
00:24:51;21 Inflammasome activation must be neither too hot nor too cold, but just right.
00:24:58;04 And so, this notion that we developed of tension, regulatory tension, may have very wide implications
00:25:05;26 in many different situations.
00:25:08;27 In general, it may be that in biology you're never working at the extremes.
00:25:14;19 You're always working somewhere in the middle, because that's the sweet spot where the physiology
00:25:22;25 allows you to live a long and healthy life, and still be able to respond to external stimuli.
00:25:33;01 In summary, microRNAs play many different regulatory roles, just in what we call hematopoiesis,
00:25:40;04 in the formation of blood cells or in the functions of blood cells, but also in the
00:25:46;02 whole rest of the body: in the heart, in the lungs, in the... in the kidneys.
00:25:52;23 In this case, we see 146 downregulating, 155 upregulating, and they work coordinately to
00:26:00;08 produce a tension.
00:26:02;02 132 buffering a transcription factor, again, keeping stem cells just right.
00:26:09;03 And in the article in the New England Journal, miR-223 keeping the inflammasome at just
00:26:16;20 the right level of response to allow for normal colonic health.
00:26:24;06 All of these things... many of these things, the ones I talked about, when they're dysregulated
00:26:29;04 actually cause leukemia, something which I haven't mentioned, because you get
00:26:36;22 dysregulated blood cells that go on to become really dangerous to the body.
00:26:42;19 And with that I thank you.

This Talk

Speaker: David Baltimore

Audience:

Educators of H. School / Intro Undergrad
Student
Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: November 2017

Talk Overview

What is the involvement of microRNAs in the inflammatory response? In this seminar, Dr. David Baltimore shows that the expression of three microRNAs (miR-132, miR-146, and miR-155) increase upon activation of the inflammatory pathway. Baltimore characterizes these microRNAs and shows how they are involved in the precise tuning of the inflammatory response.

Speaker Bio

David Baltimore

After serving as President of the California Institute of Technology for nine years, in 2006 David Baltimore was appointed President Emeritus and the Robert Andrews Millikan Professor of Biology. Born in New York City, he received his B.A. in Chemistry from Swarthmore College in 1960 and a Ph.D. in 1964 from Rockefeller University, where he… Continue Reading