The Discovery of Dynein

Top

07:33:43.18 So welcome. Today we're talking with Ian Gibbons, who nearly
07:33:47.03 50 years ago discovered the molecular protein, dynein.
07:33:51.04 So dynein is a force generating ATPase that causes cilia and
07:33:56.09 flagella to beat. It's also found in the cytoplasm of most eukaryotic
07:34:01.17 cells, where it transports all kinds of different cargo, such as small
07:34:05.13 membrane vesicles, the golgi apparatus, the nucleus, it transports
07:34:10.04 mRNAs and protein complexes. And in mitosis it also plays an important
07:34:15.01 role in constructing the mitotic spindle, so it's a very important molecule
07:34:19.29 in cell biology. And well, I've known Ian for many years. We've collaborated
07:34:29.03 on several occasions. But I've never had a chance really to sit down and
07:34:33.12 talk to you about how this whole story came about, how you
07:34:37.02 went about discovering dynein. So, well, let's go back to when it all
07:34:43.10 started in 1963, and maybe you can tell me about what captivated
07:34:48.23 your interest in studying cilia and flagella in the first place.
07:34:52.10 I was working as an electron microscopist at Harvard, in the biology
07:34:56.21 labs. And I was collaborating on a project that involved looking at
07:35:02.23 flagella in various flagellated protozoa. And one of the days, we
07:35:10.08 happened to take some pictures by accident of the set of cross-sections
07:35:16.16 of cilia and flagella and the transition between the basal body or centriole
07:35:20.08 up into the shaft of the actinin. And there were enormous variety of
07:35:26.18 fibrillar structures there that had never been described previously.
07:35:31.15 And I felt that they must be doing something important. I knew that
07:35:36.10 cilia and flagella were motile organisms, my supervisor at Cambridge
07:35:41.10 where I was getting my PhD was John Bradfield, and he had actually written a paper
07:35:45.22 on cilia and flagella mechanisms at that time. But then he left the field
07:35:49.25 completely and went into finance. So I decided that when I took
07:35:55.13 these beautiful structures, I really wanted to know how they might work.
07:36:02.04 There was so much structure there that it was far from obvious as to
07:36:06.15 what it might do, how the different bits might interact.
07:36:10.04 So which of these structures did you think might be the force
07:36:13.26 generating entity? And how did you actually go about figuring that out?
07:36:18.23 Well I didn't get there all in one go, it took several more times. Several
07:36:22.16 stages. And it one of the things was that after I'd spent couple of years doing
07:36:29.08 electron microscopy, I realized that one only got a rather limited distance
07:36:35.22 looking at structure alone because there was almost nothing known
07:36:38.24 about the chemistry. And not a great deal about the detailed motility.
07:36:42.23 And so, I decided to take on the biochemistry. Find out for myself that I had
07:36:51.07 no experience in biochemistry, but I did have my wife, who was an experienced
07:36:55.14 protein biochemist working in John Edsall's lab at the time.
07:36:59.24 She certainly provided me with lots of help in getting me started
07:37:02.26 and knowing the important questions that should be considered.
07:37:06.19 So once I did that, then I decided that the Tetrahymena cilia were
07:37:13.13 interested model organisms to isolate cilia, since one can grow them up
07:37:18.12 by the pound and isolate them. Actually, I bought myself a cream separator
07:37:24.06 to harvest them. We did maybe about a pound in Tetrahymena in each culture.
07:37:31.28 So then from that, I used the method for isolating the cilia that had been
07:37:39.11 developed by Sir John Randall's lab in London, and used that to isolate the cilia.
07:37:44.06 His people had tried to look at the proteins, but they hadn't gotten very
07:37:49.28 far, because they found that the proteins appeared to be too almost completely
07:37:55.22 insoluble. I suspected the problem had been that the isolated
07:38:02.27 cilia after they fractured from the cell, the cilia membrane healed
07:38:06.09 over the isolated tip. And so the cilia was completely enveloped by
07:38:11.29 the membrane, so the proteins couldn't get out. When I got these preparations
07:38:15.03 of isolated cilia, I decided that first thing to do get rid of the membrane.
07:38:22.28 So I applied a mild detergent, digitonin, that did a wonderful job
07:38:27.23 with taking off the membranes, and also showed that the structure
07:38:31.18 remained intact. All the internal bits of ciliary apparatus remained intact.
07:38:38.16 And very importantly also, the cilia retained their axoneme and ATPase activity.
07:38:46.14 Very little disappeared with the membrane, so we had the nuded
07:38:50.13 cilia. I just put them into a concentrated solution of salt, and they completely dissolved.
07:38:57.17 So you had access to all the protein, but it wasn't what I wanted.
07:39:02.19 I wanted to get some fractionation, and that I achieved by extracting
07:39:08.18 with low ionic strength with EDTA, to completely remove any divalent
07:39:13.04 cations. When I did that, I found I got a very interesting separation
07:39:18.21 of the protein into two fractions, a soluble fraction and the residual
07:39:25.12 pellet. And comparing those two structures in the electron microscope,
07:39:29.14 I saw that the extraction with low salt EDTA had removed all this central
07:39:35.20 apparatus, the spokes, the two central tubules, and in addition,
07:39:40.17 had removed the arms that ran around the length of the doublet
07:39:46.10 microtubules. When I looked at the solubility of protein and ATPase,
07:39:53.02 I found that this had indeed achieved a good separation, because
07:39:58.27 the soluble fraction contained about 80% of the total initial ATPase
07:40:10.02 activity. But only had about 30% of the total protein, so I'd obtained
07:40:15.00 something at a 3-fold purification just by doing this simple extraction.
07:40:18.14 And so, since I had most of the axoneme model and their ATPase
07:40:24.03 present in the soluble fraction, the obvious thing to do was to look at
07:40:27.16 it. And the only technique that one had at the time, and this was before
07:40:31.04 gel electrophoresis, the only technique available was centrifugation.
07:40:35.09 Started with an analytical centrifuge, and there were three peaks.
07:40:40.26 One slow one at 4S, one at 14S, and one at 30S. When I ran
07:40:46.12 this same material on a sucrose density gradient and assayed for
07:40:50.17 ATPase activity, then I found that the ATPase activity was completely
07:40:55.11 confined to the 14S and 30S fractions, the 4S had gone essentially null.
07:41:01.08 Though it was the largest fraction in total protein. So then the question
07:41:06.28 was, which of these various structures that had disappeared in the
07:41:10.28 axoneme did the ATPase actually come from? And to find that out,
07:41:16.00 I tried recombining. I took the soluble fraction and the pellet fraction and
07:41:21.17 mixed them together, in the presence of magnesium. And fortunately,
07:41:26.07 I found that a major fraction of the ATPase activity that had been
07:41:31.20 extracted in solution became rebound to the pellet fraction. Looking in the
07:41:38.19 electron microscope, this showed that the structures that had reappeared
07:41:43.19 when the ATPase activity became rebound were the two arms that
07:41:47.16 stick out from each of the doublet tubules. So this was pretty
07:41:51.05 strong evidence that the axonemal ATPase was indeed
07:41:55.07 localized in these arms, around the doublet tubules.
07:41:59.29 So now you had this ATPase, you could show that you could add it back
07:42:03.27 and it could reconstitute this really interesting structure between
07:42:07.05 the two outer doublets. So next step, how did you come up with the
07:42:12.23 name of this interesting ATPase?
07:42:15.05 Once we had the protein more purified, we gave it the name dynein.
07:42:18.13 Dynein means, "dyne" was the unit of force, the only unit of force present
07:42:23.20 when I went through school. Actually, Barbara was the one who
07:42:26.23 suggested the name dynein, she suggested it one evening at home.
07:42:30.11 So at that time, how did you think it worked?
07:42:33.04 Well, there were two theories running. And there were two possible
07:42:38.03 mechanisms. One would be a mechanism involving localized
07:42:41.21 contractions, that somehow the localized sections on the length
07:42:48.14 of the microtubule might change the crystallized packing arrangement
07:42:53.10 of subunits, in a way similar to that of the bacterial flagellum when it
07:42:58.03 changes its shape from a left handed helix to a right handed helix
07:43:01.01 under the influence of a chemotactic attractant. That will be one
07:43:05.29 possible mechanism. Second possible mechanism was that microtubules would slide
07:43:11.12 relative to one another, and when this opposed by some resistance to bending,
07:43:16.11 would cause a formation of a new bend. The difficulty with that mechanism
07:43:21.08 was that to have bends of opposite directions at different positions
07:43:25.27 along the length of the axoneme meant that the individual microtubules
07:43:31.04 -- the microtubules run straight along the length of the axoneme,
07:43:34.05 but it meant that sliding has to be in opposite directions in order to
07:43:38.23 generate bends of opposite directions at different positions.
07:43:43.13 Even where you're only thinking of a single microtubule.
07:43:48.23 That seemed to be a major obstacle. And so, I didn't take it any further.
07:43:58.05 At that time, I actually looked at the text of the paper and was actually
07:44:01.14 leaning toward a contractile hypothesis.
07:44:04.14 So then what did you do next? It was time to do an experiment
07:44:08.29 and so what kind of experiment to discern between these two mechanisms?
07:44:13.08 After we'd taken the formality of publishing the more purified protein,
07:44:16.25 and the name, which was in 1965, then I realized that we needed to study
07:44:24.02 yet more into the motility. And the Tetrahymena cilia were extremely
07:44:28.00 inconvenient for doing that, because they're only 5um long and
07:44:31.27 though you can see them wagging if you use a high power light
07:44:37.08 microscope lens, but you could never measure accurately the parameters
07:44:43.05 of waveforms. So that was when I decided that we should shift to
07:44:47.28 sea urchin sperm flagella. Because these are about 40um long.
07:44:51.26 About 8 times as long as the Tetrahymena cilia. Their motility
07:44:56.03 had already been studied by Sir James Gray, who'd shown that
07:45:00.05 they propagate waves along the length of the flagellum.
07:45:03.24 But we didn't have a really good source of them in Cambridge, Massachusetts,
07:45:07.22 where I was then. And so we waited and an opportunity came up
07:45:13.10 for me to move to Hawaii, the University of Hawaii. And I decided that
07:45:19.00 would be an opportunity to switch completely to using sea urchin sperm
07:45:23.05 since the sea urchins in Hawaii were sexually ripe all year round.
07:45:28.24 You would have material for working with them year round, without having
07:45:32.18 the complications that would happen if you tried to do something like
07:45:36.20 California or on the East Coast. So we moved to Hawaii and then we started
07:45:41.09 studying the motility in more detail. We first measured the parameters of
07:45:46.09 the bending waves of the live sperm, and then we took membranes
07:45:50.09 off the way we had been doing for the Tetrahymena cilia.
07:45:52.19 Showed that they could be reactivated with ATP and propagated
07:45:56.16 bends very nicely. Then the question was the mechanical force actually
07:46:02.12 going on in the axoneme? And Keith Summers, who had
07:46:10.24 -- my graduate student at the time, had the idea of using
07:46:15.00 trypsin to weaken the structure, he thought that might do
07:46:18.27 something. I had earlier observed that with the Tetrahymena cilia,
07:46:22.25 they become swollen when you gave them ATP, they didn't slide.
07:46:26.18 But they became swollen. So he tried to make this happen also with the
07:46:32.08 sea urchin sperm flagella, but was completely unsuccessful. They didn't
07:46:36.19 do the same trick. So he treated them with a little bit of trypsin to see
07:46:41.04 if that would get them to do the same trick, and well it did, but
07:46:44.29 it also did something much more dramatic because it just seemed to
07:46:48.07 disappear completely. If you took the demembranated isolated
07:46:51.20 axonemes, treated them weekly with trypsin, and then diffused in
07:46:55.25 ATP slowly from the edge of the microscope slide, as the front of ATP
07:47:01.04 reached the axonemes, it just seemed to fade and disappear.
07:47:04.07 So something quite dramatic was going on, something that
07:47:09.13 did not happen with the Tetrahymena cilia.
07:47:10.29 And you wanted to look at that directly and observe it.
07:47:14.11 I decided to use light and dark field microscopy to watch
07:47:17.21 the disappearing of these flagellar axonemes. So Barbara and I
07:47:24.05 went back one evening and we looked and we saw it, and we were
07:47:28.25 amazed that they really split apart. It was really dramatic.
07:47:33.21 And that was really one of the high points, because it was obviously
07:47:38.13 demonstrating a very fundamental mechanism between
07:47:45.03 the interaction of a motor protein and microtubule for the first time.
07:47:49.15 And it was happening under our own eyes. One of the things we
07:47:54.16 did was still to get more evidence that dynein was actually
07:47:57.21 the motor driving the motility. The way we did that was by
07:48:02.12 partially extracting the dynein arms, when we did that, we found
07:48:08.09 that the beat frequency decreased proportionally to the
07:48:12.20 fraction of dynein arms that had been removed. And then
07:48:16.14 we could add the solubilized dynein back again and the beat
07:48:20.07 flagella beat frequency increased almost back to its original
07:48:23.23 level as the arms became rebound. That was really the
07:48:27.06 confirming evidence, the last nail in the coffin, or whatever you want to call it.
07:48:33.12 So we've covered the isolation and discovery of the ATPase, dynein,
07:48:38.11 to the in vitro demonstration that dynein can slide microtubules, and that
07:48:46.03 was the underlying principle mechanism that gave rise to ciliar
07:48:51.24 and flagellar beating. So maybe you can give us a quick tour
07:48:55.08 of dynein after that.
07:48:57.28 So at that point, we wanted to characterize it biochemically.
07:49:00.12 Which we did, and by that time, membrane gels, gel electrophoresis
07:49:06.12 had been discovered and developed. So we were able to start looking
07:49:09.20 at the polypeptides that were present in our preparations.
07:49:12.19 Something that had not been possible when we did the original
07:49:15.23 extraction of the dynein, because no gel was available
07:49:19.14 that were appropriate at that time. So that was one of the things, we
07:49:25.02 worked up and purified the ATPase and started looking at its
07:49:29.09 polypeptides. We eventually cloned and sequenced one of the polypeptides
07:49:35.03 from the outer arm, and then after we cloned and sequenced it, we
07:49:41.17 started looking at the different forms of it. Because it turned out,
07:49:50.01 as most people discovered with gel electrophoresis, preparations were
07:49:53.24 always more complicated than you'd like to think. And a lot more
07:49:57.02 proteins were present than one really liked to think. So we discovered
07:50:01.00 there were about 8 or 9 bands in the electrophoretic
07:50:04.25 position of the dynein heavy chains. So we isolated different
07:50:09.16 ones of them, studied how many there were, and that led to
07:50:13.00 realizing that the dynein family of the heavy chains alone consists
07:50:19.26 of at least 9 subfamilies. And two of these correspond to the two
07:50:26.01 cytoplasmic dyneins, the other seven are axonemal proteins, and
07:50:31.20 sometimes there's more than one represented of a family present at one
07:50:35.04 time. So the usual flagellar axonemes that we were using contained
07:50:39.10 13 actual gene products. So, we'd cloned and sequenced one of these
07:50:46.04 genes. The heavy chain, the beta heavy chain, the outer arm dynein.
07:50:49.25 Cloned little bits of others. But of course as time went by, other people joined in
07:50:55.17 and so by now, there's a really elaborate tree showing the nine
07:51:00.21 subfamilies of the dynein heavy chain form independent subfamilies
07:51:06.27 that were present even in the last common ancestor of all eukaryotes.
07:51:11.08 In 1997, we wanted to retire from the University of Hawaii, less than full time.
07:51:18.28 So we moved to California to be closer to our family, our children, and
07:51:26.08 there I started working on the cytoplasmic microtubule binding domain
07:51:33.06 of cytoplasmic dynein. And by taking just that separate small piece
07:51:39.01 of just microtubule binding domain with just its coiled coiled stalk, which is only about
07:51:43.28 5% of the total heavy chain of the mass of the whole dynein heavy chain.
07:51:49.14 And we were able to express it in bacteria, study its properties.
07:51:53.18 We were able to show that by changing the register between
07:51:57.15 the two strands of the coiled-coiled stalk, we could modulate
07:52:00.13 the affinity constant with which this expressed microtubule
07:52:09.09 binding domain would bind to native cytoplasmic microtubules.
07:52:14.09 Having got there and having this expressed functional cytoplasmic
07:52:20.26 microtubule binding domain, the next step obviously was to
07:52:25.18 go on and crystallize it and get a structure. Took a year,
07:52:29.18 but Joan Garbarino in my lab did a wonderful job and
07:52:34.10 then helped especially with the crystallography, we joined
07:52:39.13 Andrew Carter from your own lab and together, we managed to get
07:52:44.11 a reasonable atomic level structure for microtubule binding.
07:52:50.27 Are there any particular thoughts you have? Just looking back over this?
07:52:55.20 My general modus of operandi, I think has been to work with a
07:53:00.04 technique for a while, and if it seems to be sticking, try another
07:53:03.17 side -- try climbing another side of the mountain. Change your
07:53:06.27 technique and start a fraction where you approach it in a completely
07:53:09.27 different way. And I think that's useful, I definitely found it interesting.
07:53:15.19 I don't know how well it would work today, where there are many more
07:53:18.14 people in science. And most top problems are more popular,
07:53:21.07 competition might make it difficult for you to do that. But certainly, it
07:53:24.19 worked very well for me, so I would encourage anyone else
07:53:27.09 to try doing it if they have the opportunity. But the main thing is
07:53:32.06 to have fun, enjoy what you're doing, and then with any kind
07:53:37.01 of decent luck, you'll get the right result.

This Talk

Speaker: Ian Gibbons

Audience:

Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: November 2011

Talk Overview

In an interview with Ron Vale, Ian Gibbons describes how his initial chemical dissection of the internal structure of Tetrahymena cilia led him to seek and then identify the force-producing ATPase, which he named dynein. Gibbons also describes his key experiment showing that dynein produces sliding between microtubules, which gives rise to the beating of cilia and flagella, as well as his more recent experiments aimed at understanding dynein’s motility mechanism at an atomic level.

Speaker Bio

Ian Gibbons

Ian Gibbons, together with his wife Barbara, identified and named the motor protein dynein. Gibbons has continued to work on the function and structure of dynein throughout his long career. After receiving his PhD in biophysics from the University of Cambridge, he took a position at Harvard and then at the Kewalo Marine Laboratory at… Continue Reading