Top

iBiology

Bringing the World's Best Biology to You

Understanding Cell Shape

Part 1: Understanding Cell Shape: Big Insights From Little Plants

00:00:05.09 Hello. My name is my Magdalena Bezanilla. I'm
00:00:08.08 from the University of Massachusetts at
00:00:10.04 Amherst and today I want to tell you
00:00:12.03 about the work that we do in my lab. So
00:00:14.08 I'm a cell biologist and my primary
00:00:17.03 interest is in understanding how cells
00:00:19.08 shape themselves, and we study this in
00:00:23.06 plant cells because plants have to build
00:00:27.03 their cell walls and the cell walls
00:00:29.03 ultimately dictate what their shape is.
00:00:31.02 And that shape then determines what the
00:00:35.05 organism looks like. What I'm showing you
00:00:37.05 in this movie is a movie of a plant, my
00:00:41.26 favorite plant, a moss, and what you're
00:00:46.06 seeing is the development of the gametophore.
00:00:47.04 The gametophore is a leafy shoot. If
00:00:48.10 you were to walk in the woods, what you
00:00:51.18 would see in a patch of moss is a bunch
00:00:53.16 of these leafy shoot. So, what I hope to
00:00:56.03 show you today is this organism, with its
00:00:59.28 amenability to look at its single cells
00:01:01.01 as they are growing, really helps us
00:01:02.08 understand how plants ultimately shape
00:01:06.01 themselves. So, this is another example of
00:01:10.05 the beautiful and dramatic differences
00:01:12.05 in cell shapes that plants can have. This
00:01:15.03 is an emerging leaf from Arabidopsis, a
00:01:19.00 very well-known model organism in the
00:01:21.08 study of plant molecular biology. What
00:01:24.08 you see in the different colors off of
00:01:27.01 the little green leaf are leaf hairs and
00:01:29.06 these leaf hairs have little sort of
00:01:33.00 horns off of them -- they're called
00:01:34.09 trichomes. They're very dramatic cell
00:01:37.01 shapes also in the epidermis of the leaf.
00:01:40.06 You can see different shapes of cells
00:01:43.03 Ultimately these things will pattern the
00:01:46.07 organism. So the fundamental question in
00:01:49.08 my lab is, how does a cell from the
00:01:53.03 inside help to pattern the entire
00:01:56.03 organism. So in other words, how does it
00:01:58.06 template the extracellular geography
00:02:01.01 that's important for the whole organism
00:02:03.04 to function and to have the right
00:02:05.07 morphology? So, the reason plants are
00:02:11.01 really great for this is that it's been
00:02:13.05 known that... for example here you have on
00:02:16.05 the top a wild-type epidermis
00:02:18.09 and on the bottom you have a mutant
00:02:21.00 epidermis, and what you can see is that
00:02:23.07 the shapes of those cells are very, very
00:02:26.00 different.
00:02:26.07 So, does this ultimately impact the shape
00:02:31.10 of the organism? Well it does, in fact. So
00:02:32.25 right here you can see the wild-type
00:02:34.04 leaf looks normal and round and circular,
00:02:37.00 like you would expect it to, but the
00:02:40.06 mutant is shrunken and shriveled. And
00:02:42.05 this actually also translates to the
00:02:45.05 entire plant. So, as I mentioned in the
00:02:49.09 introduction, the cell wall that the
00:02:53.27 plant cell built is ultimately
00:02:54.08 constraining the shape of the cell, and
00:02:56.06 there are two processes by which you
00:02:59.00 could imagine this could happen.
00:03:00.05 So one process, which my lab is very much
00:03:03.05 interested in, and in Part 3 of this
00:03:05.08 series you'll hear more about it, is cell
00:03:08.08 division. So when a cell divides you can
00:03:13.03 see it divide in the middle in which
00:03:14.06 case... which is over here closer to the
00:03:18.02 arrow... you can see that it would give you
00:03:20.24 two equally rectangular cells, or you can
00:03:23.24 make a triangular cell, or you can make
00:03:24.02 very rectangular cells, or you could make
00:03:26.26 unequal cells. And that ultimately can
00:03:29.07 impact what the shape is, but coupled
00:03:32.29 with that you also have cell expansion.
00:03:34.03 So, cell expansion can by itself
00:03:38.27 completely remodel the shape of the cell,
00:03:41.13 but it also could remodel the shape
00:03:42.02 based on how it was templated during
00:03:45.29 cell division. So, my lab really wants to
00:03:47.08 understand how molecules inside the cell
00:03:52.01 help to pattern the overall shape of
00:03:54.04 the cell and that's really fundamentally
00:03:56.06 what we do in my lab. In order to do this
00:03:59.08 you need a system that allows you to
00:04:01.09 look at the shapes of the cells, look at
00:04:04.09 the shape of the whole organism, but also,
00:04:06.02 very importantly, you need to be able to
00:04:09.01 manipulate the organism and you need to
00:04:11.05 be able to do that very, very precisely.
00:04:14.27 And so, as a postdoc, I discovered a
00:04:17.05 beautiful model system that allowed us
00:04:21.12 to really do all of these things. And so
00:04:22.04 what I hope to do in this section is to
00:04:24.00 tell you more about this model system,
00:04:26.09 how it works, all of the benefits and
00:04:29.02 extreme measures that we can take to manipulate it.
00:04:33.06 And hopefully I'll convince you that
00:04:35.08 this is a great system to work on. So,
00:04:39.20 this system is Physcomitrella patens.
00:04:40.08 As I mentioned, it is a moss and the
00:04:43.05 wonderful thing about this organism is
00:04:45.07 it's very easy to grow: all you really
00:04:48.05 need is sort of a refrigerator-sized
00:04:51.05 incubator in your lab that has some
00:04:54.20 lights in it that you can control, and
00:04:55.08 you can control the temperature. And so,
00:04:58.09 what you're seeing here is a plant that
00:05:01.05 was growing in this incubator. It is
00:05:04.04 about the size of a dime, just so you
00:05:06.22 have an idea of the reference, and on the
00:05:08.05 outskirt of the plant, so all along the
00:05:11.00 outskirt here, is a filamentous tissue
00:05:13.05 which I will refer to as protonemata
00:05:16.01 so that you're aware of the terminology.
00:05:18.04 And then inside you can see the little
00:05:22.04 sort of shoots that you saw developing
00:05:23.07 in that first video and those shoots are
00:05:26.05 all are called gametophores. So this is
00:05:29.00 the life cycle of Physcomitrella patens.
00:05:33.01 It starts... one of the benefits of this
00:05:33.19 model organism is the fact that it's
00:05:36.18 predominantly haploid, so most of the
00:05:37.01 tissues shown on this slide are actually
00:05:39.00 haploid tissues. In fact, it starts out as
00:05:42.25 a haploid spore, which is over there at
00:05:44.00 the far end of the slide, and that spore
00:05:46.09 germinates into protonemal filamentous
00:05:49.03 tissue, the protonemata. There are two
00:05:51.05 tissue types: the one on the top here is
00:05:53.08 the chloronemal tissue -- it grows slowly
00:05:57.29 relatively and it has lots of
00:05:58.03 chloroplasts; and then here are the
00:06:00.06 caulonemal cells, which grow fast and
00:06:02.05 have fewer chloroplasts. This is
00:06:06.24 essentially a two-dimensional array of
00:06:07.05 tissue that branches. During development,
00:06:11.07 you will branch, but then you can also
00:06:14.02 form a bud. And that bud is shown here. It
00:06:17.00 is the predecessor to the leafy shoot or
00:06:21.06 the gametophore, which as I mentioned is
00:06:22.15 what you would see if you were walking
00:06:23.00 in the woods. So, at the apex of the
00:06:26.00 gametophore, shown with the circle
00:06:28.04 there, is where you get formation of male
00:06:30.16 and female organs. And those are shown in
00:06:36.04 the next panel. The next circle there is
00:06:36.07 where the male organs are. You actually
00:06:38.07 have flagellated sperm made in these
00:06:40.06 organs that then are released and then
00:06:42.27 swim up to and fertilize the egg, which
00:06:45.04 is circled here. Then you get for
00:06:48.02 fertilization, formation of the only
00:06:50.08 diploid part of this lifecycle called
00:06:53.00 the sporophyte, which is bracketed in the
00:06:55.07 next panel, and that sporophyte then
00:06:59.03 undergoes meiosis and in the capsule you
00:07:01.08 have spores. So, I just wanted to show you
00:07:06.00 that we can actually watch these
00:07:07.08 developmental processes happening in
00:07:11.05 our microscopes. And so the next movie
00:07:14.01 that I'm going to show you is actually
00:07:15.06 the development of one of the buds and
00:07:17.09 it's going to look like it was branching,
00:07:20.03 but then it turns into a leafy gametophore
00:07:22.06 So, here it is. You're seeing the
00:07:25.07 formation of the bud and you'll see some
00:07:29.05 very characteristic cell division
00:07:31.02 patterns that will lead to a triangular
00:07:35.07 cell at the top. That triangular cell is
00:07:38.04 actually, interestingly, the single stem
00:07:40.03 cell for the entire structure and it
00:07:43.04 will continually divide, replenishing
00:07:45.03 itself, and then you'll see that you'll
00:07:47.04 have formation of leafy gametophores
00:07:50.00 on the side, and you can see a rhizoid
00:07:53.03 growing out. That one happens to explode
00:07:56.05 at the end, there, and then it'll start to
00:07:59.02 regrow next to that very green filament.
00:08:01.09 So, we can watch this one more time just
00:08:04.04 to see all those things happen.
00:08:05.08 So, here is the bud emerging. The first
00:08:09.00 division is very characteristic, and the
00:08:11.09 formation of that triangular apical stem
00:08:14.01 cell, and then you can see a leaflet on
00:08:17.00 one side, and then the rhizoid,
00:08:20.00 which will emerge off the base, which is
00:08:21.06 also a very polarized growing tissue, and
00:08:25.05 that one is very interesting because as
00:08:27.00 it grows to the edge of the field, at the
00:08:28.09 very edge of the field, for some reason
00:08:30.03 it explodes, right there, and then it will
00:08:34.09 be able to regenerate a new tip. What I
00:08:38.02 love about this movie is it shows that
00:08:40.02 even though a cell exploded this tissue
00:08:42.07 is very developmentally plastic and can
00:08:45.01 regenerate itself. So, I want to
00:08:49.06 focus on these protonemal cells, these
00:08:51.06 filamentous cells on the edge of this
00:08:53.03 plant, because they are an excellent
00:08:55.05 system to study the two processes I'm
00:08:57.07 most interested in. As I
00:08:59.01 said, I'm very interested in cell
00:09:00.05 division and cell expansion, and really
00:09:02.00 this tissue is fantastic for that.
00:09:05.02 So, as a tool, we developed a line that
00:09:09.09 expresses a green fluorescent protein in
00:09:12.09 the nucleus. What that allows us to do is
00:09:15.04 watch cell division events under the
00:09:17.06 microscope. So, you can see that line here
00:09:21.02 and you can see an apical cell division
00:09:24.05 that is symmetric and then the next time
00:09:27.00 the movie plays I'll point out a
00:09:29.06 sub-apical cell that is an asymmetric cell
00:09:32.04 division. So, this is a wonderful system
00:09:35.03 to see both kinds of divisions and
00:09:37.08 they're actually different. One, you place
00:09:40.00 your division plane right smack in the
00:09:42.09 middle and the other one you actually
00:09:44.05 have to move the nucleus to a new place
00:09:46.06 and then divide there. So, they're very
00:09:48.09 different signaling mechanisms,
00:09:50.05 potentially, that we can investigate and
00:09:52.00 explore to understand these two
00:09:53.08 different types of cell divisions. Now,
00:09:56.02 what about cell expansion? Well, these
00:09:59.05 protonemal cells are fantastic for
00:10:02.00 studying cell expansion. And so what I'm
00:10:04.04 showing you here is a video of these
00:10:06.06 cells growing and what you can see is
00:10:09.05 that they are very persistently growing
00:10:12.01 in one direction, and that's because
00:10:14.00 their growth machinery is directed to
00:10:17.03 one particular place in the cell and one
00:10:19.07 of the fundamental things that my lab
00:10:21.04 wants to understand is how is that
00:10:23.03 growth machinery so precisely maintained
00:10:25.08 in one place, so that the cell can only
00:10:28.03 expand at that one place. So, hopefully
00:10:31.05 I've convinced you that imaging this
00:10:34.05 organism is very simple to do, but of
00:10:36.07 course we now need to be able to
00:10:38.00 manipulate it genetically as well. And we
00:10:41.02 also need to be able to grow it in the
00:10:43.04 lab. And while that life cycle I showed
00:10:46.01 you only takes about two, two and a half
00:10:48.07 months, it still takes some time. So, we'd
00:10:51.02 like to be able to short-circuit that
00:10:52.04 and here is where the developmental
00:10:54.02 plasticity of this organism is so key. So,
00:10:57.00 you can take a plant that looks like
00:10:58.07 this, that has many different tissue
00:11:00.05 stages, and you can literally put it in a
00:11:02.05 blender with water and just chop it up.
00:11:04.02 And everywhere where it's been damaged it
00:11:06.00 will be de-differentiate to the most basal cell
00:11:08.00 type, those chloronemal filaments that
00:11:11.00 are full of chloroplasts
00:11:12.07 and grow relatively slowly. And so then
00:11:16.00 you can have a very fresh supply of
00:11:18.03 these cells on a weekly basis, on a daily
00:11:21.00 basis, depending on how often you want to
00:11:23.00 blend it in a blender. So, we have easy
00:11:25.07 access to single cells, fantastic cell
00:11:28.00 biology. But to do reverse genetics and
00:11:31.05 to do molecular genetics it would be
00:11:33.01 very useful to have all the information
00:11:35.05 about the genome, and luckily in 2006 the
00:11:38.01 complete genome was published, and we
00:11:40.01 have a very well-annotated genome at
00:11:41.00 this point. Now, the reason that this
00:11:44.01 organism was chosen for sequencing the
00:11:47.00 genome is because, among plants, it's
00:11:49.02 actually the only known plant that
00:11:51.02 undergoes efficient homologous
00:11:52.09 recombination when transformed with a
00:11:55.00 piece of DNA. And so this opens doors for
00:11:58.07 many, many tools. So, you can you can knock
00:12:01.04 out a gene, you can knock in a gene, and
00:12:04.00 so you can do very precise genetic
00:12:05.08 manipulations. So, I'll talk about that,
00:12:08.07 but I also want to talk about another
00:12:10.05 way where it's very developmentally
00:12:11.00 plastic, and that is that you can
00:12:14.00 actually regenerate a whole plant from a
00:12:15.09 single cell. So, what I'm showing you here
00:12:18.00 is a plate of moss tissue after one week
00:12:21.00 of growth, after it was blended in a
00:12:23.04 blender. And then we can scrape it off
00:12:25.05 the tissue and the reason we can do that is
00:12:28.00 we actually overlay the plate with a
00:12:30.00 piece of cellophane, and that allows the
00:12:32.06 tissue to just grow right on top of the
00:12:34.00 surface without actually penetrating the
00:12:36.04 agar. because the agar is really like
00:12:38.03 soil for this moss -- it could grow right
00:12:40.02 into it. But the cellophane blocks that
00:12:42.03 and then allows us to take all the
00:12:44.01 tissue off. Then, we can treat the tissue
00:12:46.06 with enzymes that digest off the cell
00:12:49.00 walls and we get a suspension of single
00:12:51.00 cells, and these are called protoplasts.
00:12:53.06 So, then the protoplasts, we can place them
00:12:56.02 onto appropriate medium, and that we can
00:12:59.08 watch regeneration. So, I have a little
00:13:01.07 animation here, a cartoon, where you can
00:13:04.03 see that the cells grow, you get these
00:13:06.02 symmetric cell divisions, and then you
00:13:08.03 get that asymmetric cell divisions that
00:13:10.02 make the branch. And so this is actually
00:13:13.00 what a plant looks like imaged with a
00:13:15.03 fluorescence microscope. We've labeled
00:13:17.06 the cell walls with the fluorescent dye
00:13:19.07 so that we can outline the whole plant.
00:13:21.06 In the center of the cell, in the center
00:13:23.08 of this plant, you can actually see the
00:13:25.04 initial protoplast,
00:13:26.05 but then you can see that the cells have
00:13:28.03 divided and you formed a beautiful
00:13:30.00 little plantlet. This is a six-day old
00:13:32.09 plant, so you can actually see phenotypes
00:13:35.00 very quickly when regenerating from a
00:13:37.08 single cell. So, it's very powerful.
00:13:41.04 Alright. So, what is efficient homologous
00:13:44.07 recombination mean? Well, let's just look
00:13:47.00 at a normal gene, okay?
00:13:49.02 And in this gene we've got boxes that
00:13:51.08 are the exons... and then
00:13:54.01 those exons are separated into coding
00:13:56.00 and non-coding sequence. So, the white
00:13:58.00 parts of it are the coding sequence and
00:14:00.06 then the colored boxes are the
00:14:02.07 non-coding sequences. And the lines are
00:14:05.00 introns of this gene. Now, we want to
00:14:07.07 manipulate this gene; we actually want to
00:14:09.00 knock out all of the white boxes, right?
00:14:12.01 We want to get rid of the coding
00:14:13.06 sequence. So, how can we do that? Well,
00:14:16.00 using DNA techniques, we can make a
00:14:18.07 construct and what that construct has is
00:14:21.07 regions of sequence identity in the
00:14:23.09 colored regions, but in the middle we've
00:14:26.04 replaced it with something else.
00:14:27.09 And that something else is something
00:14:29.03 that we can actually select for in the
00:14:31.04 lab. So, that something else could be an
00:14:33.00 antibiotic resistance gene, for example,
00:14:35.09 So, when this piece of DNA, this one here,
00:14:38.07 gets into the organism then a homologous
00:14:42.00 recombination event in the regions where
00:14:44.02 there's sequence identity occurs and you
00:14:46.05 select for that purple piece of DNA,
00:14:49.02 because you can do that in the
00:14:50.06 laboratory, and then you get an altered
00:14:53.00 locus. This then allows you to manipulate
00:14:57.04 the genome precisely and what you've
00:14:59.06 done is gotten rid of all of those white
00:15:01.00 boxes, and you've turned it into an
00:15:03.04 altered locus. So, now, the protein coding
00:15:06.04 region of that gene is gone and you've
00:15:08.03 made a null allele. Now, you can imagine
00:15:10.08 doing lots of other things with this
00:15:12.04 technology. You could precisely put in a
00:15:14.08 GFP
00:15:15.06 where you want it to be, a green
00:15:17.02 fluorescent protein where you want it to
00:15:18.07 be. You could precisely put in a point
00:15:21.04 mutation. So, it's a very, very powerful
00:15:23.04 technique, and it allows us to manipulate
00:15:27.03 the genome. Now, I mentioned that this is
00:15:30.05 a haploid organism, right? And so that
00:15:34.04 provides some challenges and the
00:15:37.05 challenges are that sometimes a gene
00:15:41.28 essential, so if you try to knock it out
00:15:42.19 to find out what its function is you
00:15:44.29 actually can't recover the organism.
00:15:46.00 So, that's a real problem. But, also
00:15:49.12 haploid organisms tend to have genes in
00:15:52.25 multiple copies. So, it turns out moss
00:15:54.20 underwent a couple of duplications... ancient
00:15:56.14 genome duplications and so there's a lot
00:15:59.03 of genes that live in small gene
00:16:01.00 families, which means that oftentimes you
00:16:05.19 can get rid of one copy but then the
00:16:07.14 other copy comes in and takes its place --
00:16:08.21 that's functional redundancy. So, that's a
00:16:10.00 real hassle. Well, are there other
00:16:14.05 techniques that we could use to try to
00:16:15.27 circumvent this and get more manipulability?
00:16:17.08 And so, one of these
00:16:19.00 other techniques is to use RNA
00:16:23.02 interference. So, RNA interference doesn't
00:16:23.07 actually affect the genome at all. What
00:16:27.06 it does is it affects the expression of
00:16:29.25 genes, and so that allows us to then look
00:16:33.08 at phenotypes based on lots of
00:16:37.09 expression of particular genes without
00:16:37.18 actually affecting the whole genome. So,
00:16:40.09 as a postdoc, what I did was to develop a
00:16:45.00 very rapid transient RNAi assay that we
00:16:46.07 still use every day in my lab because
00:16:48.09 it's a very powerful technique. So, we're
00:16:52.22 going to go back to this line that I
00:16:54.27 showed you earlier that expresses that
00:16:56.22 GFP in the nucleus. In fact, it expresses
00:16:59.16 a fusion; it expresses GFP fused to
00:17:00.01 another reporter called
00:17:02.03 beta-glucuronidase that I'm calling GUS for
00:17:04.27 short. This allows us... the reason for this is
00:17:08.17 that GFP would actually just leak right
00:17:10.24 back out of the nucleus, so you can make
00:17:11.04 it bigger and then it doesn't come out
00:17:13.22 of the nucleus, so we can see it nicely
00:17:14.20 concentrated in the nucleus. But that
00:17:16.06 actually gave us a little bit of a tool.
00:17:17.00 So, we can make an RNAi construct and
00:17:20.02 this is what our RNAi constructs look
00:17:22.07 like. So, we have a strong constitutive
00:17:25.00 promoter that expresses inverted repeats
00:17:28.15 of regions of GUS and regions of
00:17:31.00 whatever gene it is that you want to
00:17:34.10 target. So, the regions of GUS will
00:17:37.10 actually go and target that GFP-GUS
00:17:38.09 fusion, so it'll turn that gene off, right?
00:17:42.00 It'll silence it. And it also has pieces
00:17:46.01 of the target gene and those pieces of
00:17:49.22 the target gene will be silenced
00:17:50.28 simultaneously, because it's all in the
00:17:51.02 same construct.
00:17:52.03 So, when this makes an RNA, what happens
00:17:55.02 is you get folding back of those
00:17:57.09 inverted repeats and so that you
00:17:59.04 actually get a double-stranded RNA. And
00:18:01.05 that activates the silencing machinery.
00:18:04.06 So, you have these enzymes that come in
00:18:06.06 and splice this double strand... long
00:18:09.00 double-stranded RNA into small
00:18:11.02 interfering RNAs. Those small interfering
00:18:14.03 RNAs then go and target an enzyme
00:18:16.06 complex that goes and searches out for
00:18:18.00 the mRNA, and either degrades the mRNA or
00:18:21.06 halts translation. So, therefore you don't
00:18:24.02 get expression of that gene anymore. And so
00:18:27.01 what we can do, then, is we can transform
00:18:29.04 this into our protoplasts, because that's
00:18:31.01 where we transform DNA into cells in
00:18:33.04 this particular organism, and then we can
00:18:35.06 allow those protoplasts to regenerate, we
00:18:37.06 can select for the piece of DNA, and then
00:18:39.08 we can look for plants that no longer
00:18:42.00 express the GFP. And then we can ask the
00:18:45.00 question, does this plant look normal or
00:18:47.00 does it look mutant?
00:18:48.09 Alright. And so we can do this and it's
00:18:51.00 a very rapid assay. So, this right here is
00:18:53.00 a seven-day old plant. You may be
00:18:55.02 wondering, why are those plants red? Well,
00:18:57.00 we're imaging these plants with a
00:18:59.01 microscope, a fluorescent microscope,
00:19:01.03 so that we can see the GFP in the
00:19:03.00 nucleus, and these plants have
00:19:05.01 chlorophyll and the chlorophyll
00:19:06.07 fluoresces red under these conditions.
00:19:08.05 So, that allows us to see the outline of
00:19:10.08 the plant. It's actually very useful
00:19:12.08 because what we can do is we can
00:19:14.04 quantify the area of the chlorophyll
00:19:17.02 autofluorescence and that actually tells
00:19:18.09 us how big the plant is, and we can do
00:19:21.00 that fairly rapidly and we can do it
00:19:22.07 automated. So, let me tell you a little
00:19:25.08 bit about how we've dissected polarized
00:19:29.00 cell expansion using this fantastic cell
00:19:32.05 biology and the molecular genetics. So,
00:19:35.04 what do we know about this ability to
00:19:37.09 grow from one particular place. So, this
00:19:40.03 type of growth has two names: polarized
00:19:42.07 cell expansion or tip growth. And I use
00:19:45.01 those two fairly interchangeably. So, what
00:19:47.08 we know is that tip growth requires
00:19:49.08 secretion of cell wall material to the
00:19:53.04 tip of the cell, right here, okay? So, those
00:19:55.05 are little vesicles that have cell wall
00:19:57.02 material in them and they're going to be
00:19:58.08 deposited here to the tip. Alright. So,
00:20:01.07 that particular material is actually
00:20:05.00 very flexible
00:20:06.00 material, so that when it gets put there
00:20:09.02 then turgor pressure, which is very high
00:20:11.09 in these cells, pushes against that
00:20:14.09 flexible material. Everywhere else in the
00:20:17.08 cell you have inflexible material the
00:20:20.06 turgor pressure can't change -- only at the
00:20:23.00 tip can you change it. And that's
00:20:24.05 essentially how these cells continue to
00:20:26.05 grow: delivery of flexible material to
00:20:28.09 the tip and then turgor pressure-driven
00:20:31.03 growth. So, how is this regulated? Well, we've
00:20:37.04 known for a long time that the actin
00:20:39.06 cytoskeleton is very important and this
00:20:41.08 is important in these cells as well as
00:20:43.07 other tip-growing cells in other plants.
00:20:45.08 So, let me tell you a little bit about
00:20:48.01 actin. So, actin is a fantastic protein.
00:20:51.03 It can self-assemble. So, it starts... you
00:20:54.06 have a monomeric protein that can then
00:20:56.08 dimerize and trimerize, and it's that
00:20:58.00 trimer that is the seed for the filament.
00:21:01.06 And the filament can then grow and it
00:21:04.01 actually can grow at both ends. Now, the
00:21:08.02 reactions that make the dimer and the
00:21:10.04 timer are not very well favored, and
00:21:12.02 that's shown by the very large arrows
00:21:14.03 going towards the monomeric actin
00:21:16.00 subunit. But once you make that trimer
00:21:18.09 it becomes a seed and it's actually more
00:21:21.01 favorable to make a filament. So, this
00:21:23.08 actually happens in a test tube with
00:21:25.07 purified actin, okay? It's amazing that
00:21:28.08 this molecule can do this. And in the
00:21:31.07 cell all of these potential steps can be
00:21:34.08 regulated by proteins that bind to actin.
00:21:37.04 So, you can regulate whether you're going
00:21:40.04 to make a little seed, or you can
00:21:42.04 regulate whether you're going to add
00:21:43.09 more to either end, or you can regulate
00:21:46.03 whether the length of the filament -- you
00:21:48.02 can cut it apart. And so, actin-binding
00:21:51.09 proteins in the cell are going to be
00:21:53.06 very critical for all of these processes.
00:21:57.01 We know actin is important because we
00:21:59.09 can actually depolymerize actin
00:22:01.07 filaments and then we can see an effect
00:22:03.09 on growth. So, that plant that I'm showing
00:22:06.03 you here in the box here is a normal
00:22:08.01 plant under... again, protoplasts that
00:22:11.01 regenerated into a plant after seven
00:22:13.00 days and now you're looking at the
00:22:15.03 chlorophyll autofluorescence. With
00:22:17.01 increasing concentrations of a drug
00:22:19.00 that specifically
00:22:19.08 depolymerizes actin filaments,
00:22:21.28 called Latrunculin B, you
00:22:23.00 can see that the plants get smaller and
00:22:26.08 smaller, and eventually end up being
00:22:27.07 incredibly small, composed of small
00:22:30.04 spherical cells. So, actin is absolutely
00:22:32.08 essential for this type of cell
00:22:34.02 expansion. So, we can see this in cells as
00:22:39.16 they're growing, so we have a cell here
00:22:41.00 that was growing, now we're going to flow
00:22:42.05 in this drug, and you can see what
00:22:44.00 happens is that the tip loses its tip-like shape
00:22:46.04 and just swells. And that's it;
00:22:49.01 it just swells. So, the polarity, that
00:22:52.03 place where we knew we were going to put
00:22:54.17 those flexible cell wall materials, gets
00:22:57.01 diffused along the tip. So, in order to
00:23:00.06 understand what actin is doing we need
00:23:02.01 to look and see what actin looks like in
00:23:03.08 these cells. And so we used a small actin-binding
00:23:08.02 protein called Lifeact fused to
00:23:11.02 a green fluorescent protein and that
00:23:12.07 allowed us to look in live cells at
00:23:14.07 what the actin cytoskeleton looks like.
00:23:16.03 And so these are just still images, three...
00:23:19.15 maximum projections from a confocal [microscope]
00:23:22.15 and you can see that there's a lot of
00:23:24.05 actin. But really this doesn't tell you
00:23:27.16 the crux of the issue, and the crux
00:23:30.00 of the issue is that actin is
00:23:32.24 incredibly dynamic. So, what this other
00:23:34.02 image is is a movie and I'm going to
00:23:37.10 play it in a second, but this movie was
00:23:38.29 taken by total internal reflection
00:23:39.08 fluorescence microscopy. What that allows
00:23:44.07 us to do is it gives us a very high
00:23:44.29 signal-to-noise right underneath the
00:23:46.05 cell membrane. So, we're looking at just a
00:23:49.10 subset of the actin filaments and we're
00:23:50.09 looking at their dynamics. But you can
00:23:53.05 actually see really exciting events -- you
00:23:56.24 can see bundling, you can see buckling,
00:23:58.03 you can see growth, you can see
00:24:00.06 depolymerization. I'm not pointing any of
00:24:02.29 these out right now because it's
00:24:04.15 actually really hard to go and do that,
00:24:06.01 but just... I want you to get a feel for
00:24:08.00 how dynamic this is. As I'm talking, this
00:24:11.04 movie is actually playing back at real
00:24:13.00 time, so this is not set up in any way at
00:24:16.07 all and by the end of the time that I've
00:24:18.03 talked to you the entire cortical array
00:24:22.20 that I showed you has remodeled. So, it's a
00:24:23.00 very, very dynamic process. So, that's
00:24:25.03 what's going right going on right
00:24:28.10 underneath the plasma membrane. What
00:24:29.07 about during tip growth? And so, in this
00:24:32.07 movie,
00:24:33.03 you can see a cell that's growing and
00:24:37.15 you can see that there's an apical spot
00:24:39.12 of highly dynamic actin filaments and
00:24:40.08 you can see that, as it moves around,
00:24:43.00 where the cell is growing follows it. So,
00:24:47.12 this apical actin spot seems to predict
00:24:51.19 where that flexible cell wall material
00:24:52.06 is going to be put and where turgor
00:24:54.03 pressure then pushes up against the cell
00:24:56.03 and growth occurs. So, we have two actin
00:25:00.05 populations, at least, and these are the
00:25:02.03 two that I'm going to talk about. We have
00:25:04.04 the very dynamic actin filaments that
00:25:10.01 are being remodeled at high rates, right
00:25:11.19 underneath the plasma membrane -- that's
00:25:12.29 called cortical actin -- and then we also
00:25:14.04 have this tip-focused actin spot that is
00:25:18.21 important for... that seems to predict where
00:25:21.24 the cell is going to grow. Okay. So, I told
00:25:24.05 you that actin is remodeled based on how
00:25:28.08 it interacts with other proteins in the
00:25:30.24 cell. So, one of the first questions that
00:25:32.01 I asked in my lab was, what... can we alter
00:25:35.03 some of these proteins and can we see an
00:25:37.28 effect on growth? So, for example, there
00:25:43.06 are proteins that balance... that alter the
00:25:46.18 balance between filamentous and
00:25:47.19 monomeric actin, and one of these
00:25:49.11 proteins, actin depolymerizing factor,
00:25:51.02 what it does is it binds to filamentous
00:25:53.08 actin and it severs it, and it also
00:25:56.00 induces depolymerization. And so if you
00:26:00.24 don't have actin depolymerization factor
00:26:02.07 you would have less actin severing, less
00:26:03.08 depolymerization. So, what happens if you
00:26:06.05 were to get rid of the actin
00:26:07.07 polymerizing factor? It turns out ADF is a
00:26:11.12 single copy gene in moss and so we
00:26:13.28 decided to use RNA interference to get
00:26:18.02 rid of the expression of this gene. And
00:26:18.05 what I'm showing you here is the result
00:26:19.08 of an RNAi experiment. These are both
00:26:21.09 seven-day old plants regenerated from
00:26:23.07 protoplasts. The top plant is wild type.
00:26:26.00 So, it grows just fine, it's regenerated,
00:26:28.09 it's showing all those beautiful
00:26:30.00 filaments. And the bottom plant is what
00:26:33.12 would have what happens when you remove
00:26:35.14 actin depolymerization factor. They're very,
00:26:37.08 very small plants composed of small
00:26:39.06 spherical cells, and in fact, if we let
00:26:41.00 them grow even a day or two later, they
00:26:42.09 actually start to die. The chlorophyll
00:26:45.14 fluorescence senesces
00:26:47.12 and they die. So, this is an essential
00:26:48.28 gene. We would never actually have been
00:26:51.27 able to use homologous recombination to
00:26:52.04 knock it out. We can also analyze the
00:26:55.00 dynamics of the actin cytoskeleton in
00:26:57.08 these two plants. So, in the top there you
00:27:01.22 can see the cortical actin in the
00:27:03.01 control plants and you can see that, yes,
00:27:05.02 it's very dynamic, it's being remodeled,
00:27:07.24 and in the bottom plant you can see
00:27:09.19 that, in fact, nothing is happening.
00:27:11.06 There's still actin, but nothing... nothing
00:27:15.13 is happening. It's a very static array. So,
00:27:19.10 this tells us that the dynamics of this
00:27:23.01 network really rely on disassembly of
00:27:24.08 actin filaments. Without this disassembly
00:27:27.09 factor, you get a very static array. So,
00:27:32.09 ADF doesn't always work by itself; it has
00:27:36.01 helper proteins. So, there's a protein
00:27:38.02 called actin interacting protein, AIP1,
00:27:40.08 and it binds to and enhances ADF's
00:27:44.24 disassembly activity. Interestingly, this
00:27:46.12 gene is also just a single copy in the
00:27:49.22 moss genome, and RNAi showed that you
00:27:51.04 got smaller plants, but they were much
00:27:54.20 more viable -- they lasted longer. So we
00:27:56.11 thought, oh... maybe we can knock this gene
00:27:58.27 out. And so we did. We used to homologous recombination
00:28:00.08 and we knocked AIP1 out.
00:28:03.13 What did it look like? Well, we were
00:28:05.04 actually able to get complete plants
00:28:07.04 right? We were able to get the wild type ones
00:28:10.10 shown on the top, and you can see
00:28:11.11 the knockout of AIP1 down there.
00:28:13.01 And AIP1 regenerated, it actually formed
00:28:17.09 the leafy shoots, it pretty much did
00:28:18.03 everything you expected the plant to do,
00:28:19.09 but it was very, very, small in comparison.
00:28:22.07 And it grew very, very slowly. And if you
00:28:25.09 looked at its arrays, actin arrays, you
00:28:29.02 could see, compared to control, that they
00:28:31.04 were actually quite static. Alright? Not
00:28:35.04 as static as the ADF RNAi arrays, but
00:28:40.05 pretty static anyway. Well, it's nice to
00:28:44.02 be able to look at these movies and say
00:28:45.06 this one's fast and this one's slow, but
00:28:46.12 when you do cell biology you want to be
00:28:48.01 able to put a number to it or at least
00:28:50.21 have some way of quantifying this. And so
00:28:52.01 we did that. One way to do this is
00:28:55.29 actually just looking at these images
00:28:56.14 here, or we just took points from our
00:29:00.27 movies, three seconds apart, and then we
00:29:02.09 color-coded them. We made the first one
00:29:06.28 red, the second one blue, the third one
00:29:07.01 green, and then we overlaid those, okay? So,
00:29:11.10 if the filaments move around a lot,
00:29:12.17 you'll be able to see the blue, the red,
00:29:16.06 or the green. And so, as you can see in
00:29:18.22 these two controls on the top, you see a
00:29:19.10 lot of color. And you see a lot of color
00:29:21.11 because the filaments are moving in
00:29:24.22 those three seconds, but here in the AIP1
00:29:27.13 mutant it's a lot whiter, because
00:29:30.04 the filaments are much more static, and
00:29:31.26 in the ADF RNAi plant it's completely
00:29:35.02 white. So, that's nice. It's a nice
00:29:38.17 representation of the dynamics, but it
00:29:40.18 still doesn't get at a good
00:29:42.04 quantification. So, how could we quantify this?
00:29:43.17 Well, we could look at these time
00:29:45.21 points and we could ask, how much have
00:29:48.25 the filaments changed from one to the
00:29:49.07 other? So, we could ask, how well
00:29:52.00 correlated are those images. So, if you
00:29:53.12 take zero seconds and three seconds and
00:29:55.22 you look at them, and you can quantify
00:29:58.21 the correlation coefficient of those two
00:30:01.09 images. And then you can see, as you
00:30:03.28 increase that time point -- so you have
00:30:05.07 three seconds apart, you have six seconds
00:30:06.26 apart, you have ten seconds apart, etc --
00:30:09.09 you can see... so, if something stays very,
00:30:12.21 very correlated it's going to stay high,
00:30:14.10 but if something dramatically changes
00:30:16.04 you're going to see a very big drop-off in
00:30:18.05 that correlation coefficient. And so this
00:30:20.21 is just data showing you that. So, the
00:30:23.16 correlation coefficient is on the y axis
00:30:24.14 and different time intervals are on the
00:30:29.05 x axis. And so you can see the purple is
00:30:31.05 the ADF RNAi plants -- they stay highly
00:30:33.02 correlated for the duration of the time
00:30:37.07 intervals that we were looking at,
00:30:38.09 whereas the grey, blue, and red are the
00:30:42.05 controls, and they very rapidly decay. But
00:30:43.16 that green line is the AIP1 mutant,
00:30:47.10 right? The AIP1 mutant is not an ADF
00:30:51.08 RNAi, it is viable, we can make plants,
00:30:52.11 but if it has a much more static array.
00:30:55.07 Recently, we've been looking at this
00:30:57.09 AIP1 mutant in terms of, how does it grow
00:30:59.23 and what does the actin look like? And so
00:31:02.15 here I'm just showing you some recent
00:31:04.13 data where you can see that, in the
00:31:08.06 control you have that beautiful actin
00:31:10.01 spot, but in the AIP1 mutant you have
00:31:11.27 this bizarre lack of
00:31:13.00 actin as a tip and then massive bundles
00:31:16.16 that seem to be flowing backwards to the
00:31:18.07 base of the cell. So, we don't understand
00:31:21.17 this yet. This is our next step. What are
00:31:23.15 we going to do, how are we going to
00:31:24.29 figure this out? It's just more genetics
00:31:26.15 and more cell biology to understand how
00:31:28.00 this cell can still sort of figure out
00:31:30.29 where its tip is. Now, we know that one of
00:31:32.07 the effects is that the cell grows much,
00:31:35.01 much slower. So, what I've shown you today
00:31:39.24 is that using the powerful cell biology
00:31:42.24 and genetics in this organism we can
00:31:45.04 analyze the function of these actin-binding
00:31:46.28 proteins, and what we learned is
00:31:48.29 that you absolutely need rapid turnover
00:31:51.00 of that cytoskeleton, both at the cortex
00:31:55.11 and at the tip, in order for rapid tip
00:32:01.01 growth to occur. And I want to end by
00:32:01.05 thanking the people in the lab who
00:32:03.01 contributed. So, Robert Augustine was
00:32:05.00 actually my first PhD student and he did
00:32:08.10 all the beautiful work on ADF and AIP1.
00:32:10.13 CJ Bascom is a new student who is now
00:32:12.08 trying to figure out what those weird
00:32:15.03 flowing bundles are doing. And thank you
00:32:19.03 for watching.

Part 2: Using Reverse Genetics to Dissect the Formin Gene Family

00:00:05.00 Hi. I'm Magdalena Bezanilla. I'm from the
00:00:08.00 University of Massachusetts at Amherst
00:00:10.00 and my lab studies how cells obtain
00:00:14.00 their shape. And today I'm going to tell
00:00:17.00 you about using the powerful reverse
00:00:21.00 genetics that's afforded to us by
00:00:24.00 the moss Physcomitrella patens to essentially
00:00:27.00 look at a gene family and be able to
00:00:29.00 dissect what parts of the gene family are
00:00:32.00 responsible for polarized growth versus
00:00:35.00 what parts of the gene family are not.
00:00:37.00 And what I'm showing you in this video
00:00:40.00 is the kind of growth that we're really
00:00:42.00 interested in understanding at the
00:00:44.00 molecular level. These are cells from the
00:00:48.00 moss Physcomitrella patens that are
00:00:49.00 growing at their tips by delivering
00:00:53.00 flexible cell wall material to the tip,
00:00:55.00 and then turgor pressure pushes them and
00:00:57.00 they grow. We can do a lot with this
00:01:01.00 organism and today RNA interference is
00:01:05.00 going to really be the star of the show
00:01:06.00 in helping us to understand the function
00:01:09.00 of this gene family that we're going to
00:01:11.00 talk about. So, first let me just back up
00:01:15.00 a little bit and tell you that this kind
00:01:18.00 of growth is dependent on the actin
00:01:20.00 cytoskeleton. So, actin is a protein that
00:01:22.00 goes from a monomer to a filament and
00:01:25.00 there are two major populations in these
00:01:27.00 cells: one is a very dynamic array of
00:01:31.00 actin filaments that are rapidly
00:01:33.00 remodeling underneath the plasma
00:01:35.00 membrane -- this is cortical actin array; and
00:01:37.00 then there's also a tip-focused actin
00:01:40.00 spot. And so, we want to understand how
00:01:44.00 these arrays are established and in Part 1
00:01:47.00 I told you that depolymerizers were
00:01:49.00 very, very important for this array and
00:01:51.00 for growth. What about proteins that
00:01:54.00 help to make actin filaments? We want to
00:01:57.00 know, do they play a role and what could
00:01:59.00 they be doing? So, one of a family of these
00:02:03.00 proteins are called the formins. They
00:02:05.00 nucleate and elongate actin filaments
00:02:07.00 and what I'm showing you here is a model
00:02:10.00 of a small actin filament in blue, and
00:02:14.00 then the formin FH2 domain, so this is
00:02:18.00 the formin homology 2 domain
00:02:19.00 in red, and that is sitting on
00:02:23.00 one end of the actin filament. And then
00:02:26.00 it has these arms that are waving off of
00:02:28.00 the FH2 domain. That's called the FH1
00:02:30.00 domain or the formin homology 1
00:02:32.00 domain, and that's characterized by
00:02:34.00 stretches of proline residues.
00:02:37.00 Those stretches of proline residues bind to a
00:02:39.00 small protein called profilin.
00:02:42.00 Profilin is a small actin monomer binding
00:02:44.00 protein that likes to bind to actin. And
00:02:47.00 so what basically happens here is you've
00:02:49.00 got profilin bound to an actin filament...
00:02:51.00 an actin monomer and this increases
00:02:54.00 the local concentration of actin
00:02:56.00 monomers at the growing end of the actin
00:02:58.00 filament, and allows for actin elongation.
00:03:01.00 And so we're just going to blow up this
00:03:04.00 molecule. You can see that, then, the actin
00:03:06.00 monomer can be added on to the actin
00:03:08.00 filament. What I'm not showing you here
00:03:10.00 is it then that little red dimer sort of
00:03:12.00 shimmies to the top of the filament, and
00:03:14.00 then this process can be repeated over
00:03:16.00 and over again. So, this is how formins
00:03:19.00 help to elongate actin filaments. What
00:03:22.00 about the formins in moss -- how many are there?
00:03:24.00 So, formins belong to gene
00:03:27.00 families and they belong to gene
00:03:28.00 families in humans. They also
00:03:31.00 belong to gene families in plants. And so
00:03:33.00 it turns out that the model system Arabidopsis thaliana
00:03:36.00 has 21 formin genes and they
00:03:40.00 group up into two different families. It's
00:03:43.00 really hard to study the function of a
00:03:45.00 family where you have so many gene
00:03:47.00 members because oftentimes functional
00:03:50.00 redundancy makes it difficult to
00:03:51.00 identify a phenotype. So, in moss, we
00:03:55.00 also have a formin gene family and
00:03:57.00 it's slightly smaller, so that helps us a
00:03:59.00 little bit. So, here I'm showing you the
00:04:02.00 six members of the class one formins.
00:04:04.00 They are all characterized because they
00:04:06.00 have the red domain, which is the formin
00:04:08.00 homology 2 domain, so if you basically
00:04:10.00 have an FH2 domain you are a formin.
00:04:13.00 They also have the green domain, which is
00:04:15.00 the FH1 domain -- that's where the prolines
00:04:17.00 are. And then they have variable
00:04:19.00 N-terminals... N-termini. So, they have,
00:04:22.00 for example, some of them have a signal
00:04:24.00 sequence, some of them have a
00:04:25.00 transmembrane domain, so they're actually
00:04:27.00 predicted to be integral membrane
00:04:30.00 proteins. And there's also
00:04:32.00 a very interesting one which has a Sec10
00:04:35.00 domain on the N-terminus. There's also
00:04:39.00 the class 2 formin. The class 2
00:04:41.00 formin, which are the stars of today's talk,
00:04:43.00 I have this FH2 domain, have the green
00:04:47.00 FH1 domain, and then have a domain on the
00:04:49.00 N-terminus called a PTEN domain. This is
00:04:52.00 homologous to the PTEN protein in mammals,
00:04:55.00 which is really, really important in
00:04:57.00 cancer. It is a lipid and protein
00:05:00.00 phosphatase but the PTEN domains on these
00:05:03.00 particular proteins don't appear to have
00:05:06.00 the important residues for lipid and
00:05:10.00 protein phosphatase activity.
00:05:13.00 So, potentially those could just be
00:05:14.00 important for lipid binding and
00:05:16.00 hopefully I'll show you that that's the
00:05:17.00 case today. And then in mosses, which are
00:05:21.00 a basal type of land plant, there's
00:05:24.00 actually a third type of formin on the
00:05:27.00 bottom there. It is a formin -- it has the FH2
00:05:29.00 domain -- it has a very, very tiny FH1 domain
00:05:32.00 and it has a different N-terminus.
00:05:34.00 We're really not going to talk about
00:05:36.00 this guy because it turns out that this
00:05:38.00 one is not expressed in the tissue that
00:05:40.00 we're interested in studying. But how do
00:05:43.00 you tackle a family of this many genes.
00:05:47.00 How do you do functional analysis
00:05:49.00 rapidly? And so we turned to RNA
00:05:52.00 interference because this would allow us
00:05:54.00 to develop constructs that would express
00:05:58.00 different regions of the different
00:06:00.00 genes and then we could build larger
00:06:02.00 and larger constructs to then
00:06:05.00 multiplex it, and then allow us to target
00:06:07.00 multiple family members at the same time.
00:06:10.00 In order to understand the phenotype we
00:06:13.00 needed to be very quantitative about it; we
00:06:15.00 needed to have a very good analysis. And so
00:06:17.00 what we can do is we can take pictures
00:06:18.00 of our plants and the pictures that we
00:06:21.00 take are fluorescent images, and those
00:06:23.00 images show us a lot of the fluor...
00:06:26.00 chlorophyll autofluorescence, and so you
00:06:28.00 get these red plants. We can measure the
00:06:30.00 area of the chlorophyll autofluorescence
00:06:32.00 and that tells us how big the plant is.
00:06:33.00 But how big the plant is doesn't tell us
00:06:36.00 what shape it is, and so we use these
00:06:38.00 morphological parameters... particularly
00:06:42.00 we are using circularity to be able to
00:06:44.00 distinguish between, for example,
00:06:46.00 a circular, very abnormal plant, because
00:06:50.00 the plant that's circular would not be
00:06:51.00 normal, but a plant that has lots of
00:06:53.00 elongated projections like a star would
00:06:56.00 be more normal. And so these two shapes
00:06:58.00 actually have exactly the same area, but
00:07:01.00 they are different in terms of their
00:07:03.00 morphology and you can distinguish them
00:07:05.00 because the circular plant has a
00:07:07.00 circularity of one but the star has a
00:07:10.00 circularity of 0.2. So, you can use
00:07:12.00 this to get a good dynamic range on your
00:07:15.00 morphology analysis. Okay, so how do we do
00:07:19.00 this analysis? Well, we work on this
00:07:21.00 organism that can be regenerated from
00:07:23.00 single cells. We can transform those
00:07:25.00 cells with a piece of DNA, that DNA has
00:07:28.00 an RNAi construct, and then we can
00:07:30.00 analyze the phenotype a week later to
00:07:33.00 look at what the plants look like. So,
00:07:34.00 these are our control plants. Our control
00:07:37.00 plants are silencing a non-essential
00:07:39.00 reporter gene called GUS, and what you
00:07:42.00 can see is that the plants are nice and
00:07:44.00 large, they're red because you can see
00:07:46.00 the chlorophyll autofluorescence, and you
00:07:48.00 can see lots of polarized extensions, okay?
00:07:49.00 So, that's what a normal plant looks
00:07:52.00 like and the construct will always be at
00:07:54.00 the top of the slide, so you can see that
00:07:56.00 this is the GUS-RNAi construct. And so
00:07:58.00 this is our quantification. So, we can
00:08:00.00 look at the area and we normalize all of
00:08:03.00 our experiments to our GUS control, so
00:08:06.00 our GUS control will have an area of 1,
00:08:08.00 and then the circularity is down here and
00:08:11.00 you can see it has a very low
00:08:12.00 circularity because they're very
00:08:13.00 polarized plants. So, now we can begin our
00:08:16.00 analysis and we basically grouped
00:08:19.00 the formins into the formins that
00:08:20.00 were most similar to each other. So, we
00:08:22.00 developed a RNAi construct that had the
00:08:27.00 formin 1 A,B, and C sequences to it, and
00:08:30.00 then we transformed that in and we looked
00:08:32.00 at the phenotype. And if you were to
00:08:34.00 just look at those images you would say,
00:08:36.00 oh my gosh, there's no phenotype, but in
00:08:38.00 fact there's a very small decrease in
00:08:40.00 area. Okay, so they contribute a little
00:08:42.00 bit, potentially, to growth. And then if you
00:08:44.00 look at the next two that are similar to
00:08:47.00 each other, because they each have a
00:08:49.00 transmembrane domain, but no signal
00:08:50.00 peptide, then you can see that, again,
00:08:52.00 there's a small decrease in area, but the
00:08:55.00 plants are still very polarized. And then
00:08:58.00 we looked at the other group, which
00:09:00.00 in this case was the very strange gene
00:09:02.00 that had the Sec10 domain on it, but
00:09:05.00 also we included formin 3. At the
00:09:07.00 time we did these experiments -- this
00:09:09.00 always happens -- we didn't know it wasn't
00:09:10.00 expressed, so we didn't know we didn't
00:09:12.00 have to include it, but we included it
00:09:13.00 just to be safe. And we can see that this
00:09:16.00 actually has the most dramatic effect on
00:09:19.00 total area and this all makes a lot of
00:09:22.00 sense, because it turns out if you look
00:09:24.00 at the expression of these genes, formin 1F,
00:09:27.00 which is the one that's being
00:09:29.00 silenced here, is the one that's most
00:09:30.00 highly expressed. So, they all contribute
00:09:33.00 to growth and they contribute based on
00:09:36.00 how well expressed they are. Okay, so now
00:09:40.00 we... sort of separated the class 1
00:09:42.00 formins into their little groups of how
00:09:44.00 similar they are, but now we can sort of
00:09:46.00 concatenate them and start putting them
00:09:48.00 together, and see what happens. So, here is
00:09:50.00 a construct that silences everything
00:09:52.00 except formin 1F and formin 3,
00:09:54.00 and you can see that that one is small
00:09:57.00 but not very, very small. Now, we add the
00:10:00.00 formin 1F and the formin 3, and it gets
00:10:02.00 smaller, but still with plants are
00:10:04.00 polarized. So, this construct actually
00:10:07.00 silences all the class 1 formins. So,
00:10:09.00 what this tells us is that, while the
00:10:12.00 class 1 formins are not important
00:10:14.00 for polarity, because these plants are
00:10:16.00 very normal looking -- they're just a
00:10:17.00 little bit smaller. So, they're smaller in
00:10:19.00 that... that means that they are important...
00:10:22.00 these formins play a role in growth
00:10:24.00 but they don't play a role in
00:10:25.00 determining polarity. So, that was
00:10:28.00 something that was really cool to find
00:10:31.00 out. Like, within three or four weeks of
00:10:33.00 making all the constructs we could
00:10:35.00 immediately tell class 1 formins are
00:10:38.00 not essential for cell polarity,
00:10:39.00 something that in any other system would
00:10:42.00 have taken a much, much longer time.
00:10:44.00 Alright. So, what about the class 2 formins?
00:10:46.00 Well, I told you they had that
00:10:48.00 different N-terminus, so they had that PTEN domain,
00:10:50.00 and there's only two of them.
00:10:52.00 So, we put both of them, because they're very
00:10:54.00 similar to each other, on the same
00:10:55.00 construct. And lo and behold, we got an
00:10:59.00 amazing phenotype. So, you get rid of all
00:11:02.00 the class 2 formins and all of a sudden
00:11:04.00 now you have these very small plants
00:11:05.00 composed of spherical cells. These are
00:11:08.00 absolutely essential for cell polarity
00:11:10.00 because this is exactly the same
00:11:12.00 phenotype that we get
00:11:13.00 if we get rid of the actin cytoskeleton,
00:11:14.00 which we know is essential for cell
00:11:17.00 polarity. And so you can look at the
00:11:19.00 analysis here of the area -- you can see
00:11:21.00 they're very small compared to the
00:11:23.00 control -- and look at the circularity. The
00:11:25.00 circularity has gone through the roof,
00:11:27.00 right? So, again, it's that signal-to-noise
00:11:30.00 this gives us to look at morphology as
00:11:33.00 well as the area, alright? So, this is
00:11:36.00 the analysis that we did and from that
00:11:37.00 we learned that, yes, class 2 formins
00:11:40.00 in Physcomitrella patens, in the moss
00:11:42.00 Physcomitrella patens, are essential for
00:11:44.00 cell polarity. Well, we had all the pieces
00:11:47.00 of DNA we thought, why not? Let's string
00:11:50.00 them all together and so we did, and we
00:11:52.00 had construct now that silences all the
00:11:55.00 formins. Well, we got very sad-looking
00:11:57.00 plants. So, these plants, you know, don't
00:12:01.00 live very long, they make three or four
00:12:04.00 cells, the cells are enormous compared to
00:12:07.00 normal cells, and in the... in the analysis, here,
00:12:11.00 they don't look very different from
00:12:13.00 the class 2 formins RNAi, but they
00:12:16.00 really are quite dead. And they don't
00:12:19.00 last much longer, so if you look at them
00:12:21.00 at nine or ten days after transformation,
00:12:23.00 they're pretty much gone from the plate.
00:12:24.00 So, that tells us that formins are
00:12:26.00 essential for viability and that was
00:12:29.00 also a very important piece of
00:12:31.00 information, and that this phenotype is
00:12:33.00 not necessarily the same phenotype of
00:12:36.00 silencing just the class 2 formins. So,
00:12:39.00 they do have different roles, the class
00:12:41.00 1 versus the class 2. Alright. So,
00:12:44.00 now we know that the class 2 formins
00:12:46.00 are critical for cell polarity and
00:12:48.00 there's two of them. So, I think the next
00:12:50.00 question is, are they redundant? Do they
00:12:52.00 need to work together or do they have
00:12:54.00 different roles? So, in order to do that,
00:12:57.00 what we ended up doing was to look at a
00:12:59.00 region in the sequence that we could use
00:13:01.00 that would be very specific for each gene,
00:13:04.00 because we didn't... if you have regions
00:13:07.00 of sequence that are very similar
00:13:08.00 between two genes in the coding sequence,
00:13:11.00 you can actually get silencing of the
00:13:13.00 other member, and so we wanted to ensure
00:13:15.00 that that wasn't happening. So, we went to
00:13:17.00 the untranslated regions of the genes
00:13:19.00 that are very, very different. And so, our
00:13:22.00 construct here, a strong promoter, expresses an
00:13:27.00 inverted repeat of the GUS control gene, and the
00:13:30.00 untranslated region of formin 2A.
00:13:32.00 And so, what does that do?
00:13:34.00 That specifically silences the expression of
00:13:36.00 formin 2A, and so then we can use our
00:13:39.00 RNAi assay and we can ask, what do the
00:13:41.00 plants look like when you silence just
00:13:43.00 formin 2A? Well, they look just like the
00:13:47.00 wild type. They look very
00:13:48.00 indistinguishable. So, we can essentially
00:13:51.00 do exactly the same thing -- now make a
00:13:52.00 construct that silences just formin 2B --
00:13:54.00 and then again we can look at the
00:13:57.00 RNAi and we see that, again, they're very
00:13:59.00 large, very polarized plants. Okay, so it's
00:14:03.00 possible these untranslated regions are
00:14:04.00 targeting anything, right? It's possible
00:14:06.00 they're just not working, right? So, how do
00:14:08.00 we figure that out? How do we test that?
00:14:10.00 Well, the way to test that is to take the
00:14:12.00 UTR of 2A and the UTR of 2B, put them
00:14:15.00 in the same construct -- this should
00:14:17.00 silence both of the genes -- and this
00:14:20.00 should look just like silencing the
00:14:22.00 coding sequence, if those pieces of UTR
00:14:24.00 actually work. And, lo and behold,
00:14:27.00 that's exactly what we saw, right?
00:14:29.00 You silence the UTR of 2A, the UTR of 2B,
00:14:32.00 and then you see these very small
00:14:34.00 unpolarized plants, and again very small
00:14:37.00 and very unpolarized. Okay? So, that tells us
00:14:40.00 that these two genes are completely
00:14:42.00 functionally redundant and that they
00:14:46.00 are essential for cell polarity. This
00:14:49.00 construct, the UTR construct, actually
00:14:52.00 provides us with the ability to do
00:14:54.00 functional genetics, right? We can start
00:14:57.00 asking, what regions of the protein are
00:15:00.00 important? Because we can do a
00:15:02.00 complementation assay. So, how does that
00:15:04.00 complementation assay work? Well, we have
00:15:06.00 our construct that specifically silences
00:15:09.00 the endogenous gene using their
00:15:11.00 untranslated regions. And then we can
00:15:14.00 co-transform a construct that lacks any
00:15:17.00 untranslated regions, so just has the
00:15:19.00 coding sequence. And then we can see, will
00:15:21.00 that be able to rescue the phenotype? Okay?
00:15:25.00 So, this is our complementation analysis.
00:15:28.00 So, this is... again, the control plants are
00:15:30.00 large, the UTR-silenced plants are
00:15:32.00 very small and very unpolarized. And then,
00:15:35.00 if we co-transform, now, the full-length
00:15:38.00 formin 2A coding sequence,
00:15:40.00 we can largely rescue the area and the
00:15:45.00 circularity phenotype. Now, we can ask,
00:15:48.00 okay, well, we want to know where this
00:15:50.00 molecule is in the cell... that's what we
00:15:52.00 really want to know because we're... if we
00:15:54.00 know where it is in the cell we can
00:15:55.00 really start to understand how it works.
00:15:56.00 So, let's add a green fluorescent
00:15:59.00 molecule to this coding sequence and we
00:16:02.00 can do that using molecular genetic tools,
00:16:03.00 and then we can ask, does it rescue the
00:16:05.00 phenotype? And it does. So, we can actually
00:16:08.00 tag this molecule with three tandem GFP
00:16:10.00 molecules and we had to do that just
00:16:12.00 because it was expressed at low levels,
00:16:14.00 and so we need to have three GFPs there
00:16:17.00 to get enough signal, and you can see
00:16:19.00 that it restores the area and it
00:16:21.00 restores the polarity phenotype. So, now
00:16:24.00 we have this powerful tool because we
00:16:26.00 can look at where this protein is
00:16:28.00 expressed in cells. Now, one of the
00:16:30.00 things that Physcomitrella patens does
00:16:32.00 really, really well is it undergoes
00:16:34.00 homologous recombination. So, instead of
00:16:36.00 doing these experiments transiently,
00:16:39.00 where you transform in and you look at
00:16:41.00 seven-day old plants only, we can
00:16:43.00 actually generate a stable line that
00:16:46.00 expresses the protein that we want to
00:16:48.00 see. So, it has a coding sequence with the
00:16:50.00 GFP tag on it and then we can actually
00:16:53.00 express it from its own genomic context
00:16:56.00 from its own promoter. And so, how do we
00:16:59.00 do that? Well, we use homologous
00:17:01.00 recombination, so this is just a very
00:17:02.00 schematic diagram of the 3' end
00:17:05.00 of the formin 2A gene. The formin 2A
00:17:07.00 coding sequence is in yellow and then
00:17:09.00 the UTR, the 3' UTR, is in
00:17:11.00 blue. And so we can make a DNA construct
00:17:14.00 that has the coding sequence fused to
00:17:17.00 those three tandem GFP molecules and
00:17:19.00 then beyond that we put an antibiotic
00:17:21.00 resistance cassette, so that we can
00:17:22.00 select for that piece of DNA. And then we
00:17:25.00 allow the organism to do its great job,
00:17:27.00 which is that it uses homologous recombination and
00:17:29.00 then we select for this altered locus,
00:17:33.00 and we can see now that instead of
00:17:35.00 having the wild-type formin 2A gene in
00:17:37.00 there, we have a formin 2A gene that
00:17:39.00 has three... a piece of DNA that will
00:17:42.00 encode for three tandem GFPs in frame.
00:17:45.00 So, this will allow us to look at the
00:17:47.00 localization, and to make sure that the
00:17:50.00 plant isn't affected in any way we can
00:17:53.00 do quantitative growth assays and
00:17:54.00 this plant grows just like the wild type.
00:17:56.00 So, we really know that this plant is
00:17:58.00 behaving very similarly during tip
00:18:00.00 growth. Okay, so where does it localize?
00:18:02.00 It localizes to the tip of the cell, which
00:18:05.00 is important because it's required for
00:18:07.00 tip growth. And the other thing that it
00:18:10.00 does... if we wanted to look at how it
00:18:12.00 interacts with actin it would be very
00:18:14.00 useful to be able to image formin at the
00:18:16.00 same time as we image actin, and so we
00:18:19.00 developed a line that has formin
00:18:21.00 labeled, in this case, 3 mCherrys,
00:18:23.00 and an actin probe, Lifeact, that's labeled
00:18:27.00 with green fluorescent protein. So, we can
00:18:29.00 look at that in growing cells and what
00:18:32.00 you can see is that wherever the red
00:18:34.00 accumulates then you see sort of a
00:18:36.00 yellow burst and that yellow burst is
00:18:38.00 because, right where the red, the GFP...
00:18:41.00 the Lifeact is polymerizing... is binding to
00:18:44.00 polymerized actin, alright? And so this
00:18:46.00 clearly tells us that polymerization, the
00:18:49.00 ability of these formins, potentially,
00:18:51.00 to polymerize actin are an important
00:18:53.00 aspect of its function. So, we wanted to
00:18:55.00 probe that with our complementation
00:18:57.00 assay, but we had to sort of understand
00:19:00.00 how is it that this formin is
00:19:02.00 specifically localizing to this area of
00:19:05.00 the cell? And so we guessed. We figured that
00:19:08.00 the N-terminus was important for this
00:19:10.00 and so we made a construct that just
00:19:12.00 expresses the N-terminus fused to GFP,
00:19:14.00 put that into plants, and then looked to
00:19:16.00 see what happened when you just expressed
00:19:18.00 the N-terminus or the PTEN domain.
00:19:21.00 And this is what happens. So, those are the
00:19:24.00 images i showed you before, which is the
00:19:26.00 full-length formin 2A fused to three GFPs,
00:19:28.00 and then the these images right here are
00:19:31.00 the PTEN fused to GFP, so you can see that
00:19:34.00 the PTEN domain is sufficient to target
00:19:37.00 the formin molecule to the tip of the
00:19:39.00 cell. So, this allows us to now play with
00:19:43.00 the polymerization part of the protein,
00:19:45.00 but targeting it properly to
00:19:48.00 the right place in the cell, because if you
00:19:50.00 just started expressing different kinds
00:19:52.00 of formins you might not be able to
00:19:54.00 target it to the right place in the cell.
00:19:57.00 So, we used our complementation analysis to
00:19:59.00 functionally dissect these genes. So, we
00:20:02.00 know actin polymerization is important
00:20:04.00 for growth and that these formin
00:20:05.00 molecules are important. So, we
00:20:08.00 wanted to understand what region is the
00:20:09.00 molecule is important and one of the
00:20:11.00 questions that we wanted to ask is this
00:20:14.00 one: is polarization mediated by any FH1/FH2
00:20:17.00 domain sufficient for function?
00:20:20.00 So, it had been known for a very long time
00:20:21.00 that the FH1/FH2 domains... as i showed
00:20:24.00 you in that earlier diagram, the
00:20:26.00 FH2 domain sits on the barbed end of
00:20:28.00 the actin filament and the FH1 domain
00:20:29.00 helps with profilin to add more actin
00:20:31.00 monomers. So, these are the polymerization
00:20:35.00 engine and so, does it have to be a class
00:20:39.00 2 polymerization engine? Could it be a
00:20:42.00 different polymerization engine? Would it
00:20:44.00 still work if we targeted it to the
00:20:46.00 right place? So, we generated chimeric
00:20:48.00 molecules to essentially test this
00:20:50.00 question. And so this is one of the
00:20:53.00 chimeric molecules we made, which had,
00:20:55.00 again, the PTEN domain to localize the
00:20:58.00 protein to the right place, but then we
00:20:59.00 took the FH1/FH2 domain from a class 1
00:21:03.00 formin in Physcomitrella patens and
00:21:06.00 we just chose formin 1D... it's expressed
00:21:08.00 well in tip growing cells and we used
00:21:10.00 its FH1/FH2. But then we wanted to ask
00:21:14.00 a more complicated question: what if we
00:21:17.00 sort of use different combinations of FH1 and FH2
00:21:20.00 domains? And so we generated
00:21:24.00 super chimeric molecules where we had
00:21:26.00 the PTEN domain, then the class 2 FH1
00:21:29.00 and the class 1 is FH2 and then the
00:21:32.00 converse. And then we used our
00:21:34.00 complementation assay to ask, do any of
00:21:37.00 these work? Can we restore polarized
00:21:39.00 growth? And how well will they restore
00:21:42.00 polarized growth? So, this is same layout
00:21:46.00 as the previous slide, so the construct
00:21:47.00 is at the top ,and as you can see this is
00:21:50.00 the phenotype that would result from
00:21:52.00 silencing the class 2 formins. So, if
00:21:55.00 you put in formin 2, you can see
00:21:59.00 that it restores area and restores
00:22:01.00 polarity, but if you put in the class 1
00:22:06.00 FH1/FH2, very small plants, pretty much
00:22:10.00 the same size as the RNAi alone, and,
00:22:14.00 again, very high circularity. There are
00:22:17.00 some tiny little polarized extensions
00:22:19.00 that try to emerge on some of the plants, but
00:22:21.00 mostly it's very unpolarized. So, that
00:22:24.00 tells us that class 1 FH1/FH2 doesn't
00:22:27.00 work. Well, can we dissect it a little bit
00:22:29.00 more? Is there a part that is more
00:22:31.00 important than the other? And so, here we have
00:22:34.00 the FH2 domain from the class 2 and it's
00:22:37.00 super unpolarized, alright? So, it's even
00:22:39.00 worse than using the class 1 FH1/FH2 and
00:22:43.00 this one is also pretty unpolarized. So,
00:22:47.00 really, what this tells us is that the
00:22:49.00 FH1/FH2, they have to work together in
00:22:52.00 order for the function of the molecule
00:22:54.00 to be intact and to be able to rescue
00:22:57.00 the polarity. Well, we wanted to
00:23:01.00 understand at sort of the biochemistry
00:23:03.00 level, are these all generating actin
00:23:06.00 filaments at the same rate? Do they all
00:23:08.00 function? And so we collaborated with
00:23:10.00 Lauren Blanchoin and we purified the
00:23:14.00 proteins, the FH1/FH2 domains of all of
00:23:16.00 these molecules, and we measured the
00:23:19.00 ability to elongate actin filaments in
00:23:22.00 in vitro assays. And we used single
00:23:25.00 molecule total internal reflection
00:23:26.00 fluorescence microscopy assays to do
00:23:29.00 this, and what I'm just going to show you is
00:23:31.00 the ultimate result of that. And so this
00:23:35.00 is a graph showing you the rate of actin
00:23:37.00 elongation and... for all of our molecules.
00:23:41.00 So, actin alone is in blue and formin 2A FH1/FH2,
00:23:44.00 so that was the one
00:23:46.00 that fully rescued, is a very rapid
00:23:49.00 acting elongator, but all of the other
00:23:51.00 molecules -- so, the formin 1B FH1/FH2,
00:23:54.00 in yellow, the chimeras in purple and in green --
00:23:57.00 they really don't make very rapid
00:24:01.00 actin filaments. So, what this told us is
00:24:04.00 that, yes, you need an actin elongator
00:24:07.00 and you need an actin elongator
00:24:09.00 that's very, very rapid, okay? So, we then
00:24:14.00 wanted to look at the other end of the
00:24:16.00 molecule; what about the PTEN domain?
00:24:18.00 So, this PTEN domain, I told you, is very
00:24:20.00 similar to a mammalian lipid and protein
00:24:23.00 phosphatase, and we don't think it's a
00:24:27.00 phosphatase because it doesn't have the
00:24:28.00 right catalytic residues, and so what...
00:24:31.00 what... what... what.. is it doing? So, we did
00:24:33.00 some in vitro
00:24:34.00 assays and we took the PTEN domain, we
00:24:36.00 purified it, and we found that it
00:24:38.00 specifically binds to a phosphoinositide.
00:24:41.00 It specifically binds to PI(3,5)P2. And
00:24:46.00 it helps to localize the formin to the
00:24:50.00 cell cortex. So, this particular movie that
00:24:53.00 I'm going to play here is a movie taken
00:24:56.00 with total internal reflection
00:24:57.00 fluorescence microscopy, where we can
00:24:59.00 sort of get a very detailed view of the
00:25:03.00 cell cortex. And what you can see is that
00:25:05.00 this formin molecule very, very
00:25:07.00 beautifully localizes to the cell cortex
00:25:09.00 in these spots and they're very dynamic.
00:25:12.00 And we know that this localization is
00:25:14.00 mediated by that PTEN domain as well.
00:25:17.00 And this PTEN domain binds this very
00:25:20.00 specific phosphoinositide. So, we
00:25:23.00 basically wanted to ask the question,
00:25:25.00 okay, if we replace the PTEN domain,
00:25:30.00 again, make Frankenstein molecules,
00:25:32.00 and replace it with something that binds
00:25:35.00 a different phosphoinositide, could we
00:25:38.00 rescue function? How important is that
00:25:40.00 PI(3,5)P2 to its function? So, we made a
00:25:43.00 molecule that would bind to PI(3,4)P2
00:25:47.00 and that molecule could not localize to
00:25:50.00 those cortical dots, and it could not
00:25:53.00 rescue the function. Then we made a
00:25:56.00 molecule that could bind PI(4,5)P2. It
00:25:59.00 also could not rescue and could not
00:26:02.00 localize to cortical dots. Then we took a
00:26:06.00 yeast protein that binds PI(3,5)P2, we
00:26:10.00 put it on the N-terminus of the FH1/FH2,
00:26:13.00 and it rescued and it localized to
00:26:16.00 cortical dots. We thought, okay, so that's
00:26:19.00 probably a mistake... let's try a different
00:26:22.00 protein. So, we took a mouse protein but
00:26:24.00 also bind specifically to PI(3,5)P2, we
00:26:27.00 put it on the N-terminus, and it also
00:26:29.00 rescued. And it rescued very, very well.
00:26:31.00 And so this tells us that PI(3,5)P2
00:26:35.00 binding is very important for the
00:26:37.00 function. So, now we've dissected this
00:26:39.00 molecule down. We know that it
00:26:41.00 has to bind PI(3,5)P2 in order for it
00:26:44.00 to work and we also know that it has to
00:26:47.00 rapidly elongate actin filaments. So, we
00:26:52.00 wanted to look at its activity at the
00:26:54.00 cell cortex in vivo, okay? So, this required
00:26:57.00 us to lose this total internal
00:26:59.00 reflection fluorescence microscopy to do
00:27:01.00 this imaging. So, this is a similar movie...
00:27:05.00 it's actually the same movie that I
00:27:06.00 showed you earlier when it was really
00:27:08.00 blown up, but what we did was look at the
00:27:11.00 behavior of these cortical dots. And so
00:27:13.00 we basically put them into little groups,
00:27:15.00 so we have some that are relatively
00:27:17.00 stationary, some that moved relatively
00:27:19.00 randomly, but then some that move in linear
00:27:21.00 trajectories. And those linear
00:27:23.00 trajectories are really obvious if you
00:27:24.00 take this movie and you take three
00:27:26.00 seconds of time and you just make a
00:27:28.00 maximum projection of that time, of those
00:27:30.00 three seconds. And that's shown here.
00:27:33.00 So, you can see in this still, now, that's
00:27:37.00 basically adding up all of the time
00:27:39.00 points, you can see very nicely linear
00:27:42.00 trajectories that are made by this
00:27:43.00 molecule. Now, remember, these molecules
00:27:46.00 very rapidly elongate actin filaments.
00:27:49.00 So, you can envision this molecule is
00:27:51.00 sitting on the end of an actin filament
00:27:53.00 generating a linear molecule. If that's
00:27:56.00 the case then those trajectories should
00:27:59.00 disappear if we get rid of the actin
00:28:01.00 filaments. And so we treated with Latrunculin,
00:28:03.00 which is a drug that depolymerizes
00:28:05.00 actin filaments, and lo and behold
00:28:07.00 you get a bunch more stationary, some
00:28:11.00 random, and absolutely no linear
00:28:14.00 trajectories, as shown here when we do
00:28:15.00 that maximum projection. So, yes, what
00:28:18.00 we're seeing when these molecules move
00:28:21.00 in linear trajectories is most likely
00:28:23.00 the generation of an actin filament.
00:28:25.00 That's really cool, to actually be able
00:28:26.00 to see that in a living cell. So, we
00:28:30.00 wanted to see if we could do that with
00:28:32.00 two colors and so these things are very,
00:28:35.00 very rapid. We had to actually use
00:28:38.00 special imaging so that we could
00:28:41.00 actually acquire both of the colors at
00:28:43.00 the same time, so that there was no time
00:28:46.00 delay. And so we again imaged actin with
00:28:50.00 Lifeact, this molecule that binds to
00:28:52.00 filamentous actin, and this is a movie
00:28:54.00 showing formin in purple, Lifeact in
00:28:58.00 blue, and I made a
00:29:00.00 little circle to show you where... and this
00:29:02.00 is looping continuously... where a formin
00:29:06.00 molecule appears on the cell cortex and
00:29:08.00 then if you follow it once it appears
00:29:10.00 you'll see that there's a blue actin
00:29:12.00 filament that fills in behind it. Also, we
00:29:16.00 saw that these formin molecules, when
00:29:18.00 they move in linear trajectories, they
00:29:20.00 also move along pre-existing actin
00:29:22.00 filaments, and we measured the intensity
00:29:25.00 of the pre-existing actin filament
00:29:27.00 before and after the formin molecule
00:29:29.00 went by it, and we saw that the intensity
00:29:32.00 of the actin filament went up. So, that
00:29:34.00 means that as it's moving along a
00:29:35.00 pre-existing actin filament it's
00:29:37.00 actually generating a new acting
00:29:38.00 filament. And we quantified these events
00:29:41.00 and we found that, in the imaging that we
00:29:45.00 did, 80% of the linearly
00:29:47.00 moving formin dots could be correlated
00:29:49.00 with an actin filament. So, that's pretty
00:29:51.00 good most of those linearly moving dots
00:29:53.00 are most likely making an actin filament.
00:29:56.00 And then we measured their rates and
00:29:59.00 those rates are very similar to rates of
00:30:02.00 actin polymerization in the cell cortex.
00:30:04.00 And they did... don't differ between
00:30:06.00 generating a new actin filament and
00:30:08.00 moving along a pre-existing actin
00:30:10.00 filament. And then you can see that there
00:30:12.00 is an increase in actin fluorescence that's
00:30:14.00 similar between generating a new actin
00:30:16.00 filament and also generating one along a
00:30:19.00 pre-existing actin filament. So, we're
00:30:21.00 very confident that, when we see a
00:30:24.00 linearly moving trajectory of a formin
00:30:27.00 molecule, it's making in actin filament
00:30:29.00 right there in front of your eyes. So, in
00:30:32.00 summary, we know that these class 2
00:30:35.00 formin molecules are very, very
00:30:37.00 important for cell polarity, we know that
00:30:39.00 they localize to the cell cortex via
00:30:42.00 the ability to bind to PI(3,5)P2, we
00:30:45.00 know that they rapidly generate actin
00:30:48.00 filaments, and so now the question is,
00:30:50.00 what are they doing? What are they
00:30:52.00 polarizing? Are they polarizing the
00:30:55.00 vesicles? Are they polarizing other
00:30:57.00 aspects of the membrane? And so we are
00:31:02.00 looking at this very carefully and we
00:31:04.00 think that actin polymerization, mediated
00:31:07.00 by these class 2 formins, may be
00:31:09.00 promoting directional membrane trans...
00:31:12.00 membrane turnover,
00:31:14.00 so this membrane, when it gets
00:31:16.00 incorporated into tip, it also has to be
00:31:18.00 brought back into the cell, and we think that
00:31:20.00 the formins are helping to make sure
00:31:22.00 that that's happening right here at the
00:31:24.00 tip. And so, with that, I just want to
00:31:27.00 thank the people in the lab who has
00:31:29.00 helped with this. So, this has been a very
00:31:31.00 long-standing project in the lab, started
00:31:33.00 my very talented postdoc Luis Vidali,
00:31:35.00 who now has his own position and is
00:31:38.00 continuing to work on myosins, and
00:31:40.00 then followed by another post doc, Ming
00:31:42.00 Lee, and a very talented graduate student,
00:31:44.00 Peter. Thank you very much.

Part 3: Myosin and Actin Steer Plant Cell Division

00:00:06.00 Hi. My name is my Magdalena Bezanilla and I'm
00:00:09.00 from the University of Massachusetts at
00:00:11.00 Amherst. And today I'm going to tell you
00:00:13.00 a story about the intersection between
00:00:16.00 two cytoskeletons and how they work
00:00:20.00 together to help position cell division
00:00:22.00 plane in plants. So, positioning the cell
00:00:26.00 division plane is very important because
00:00:28.00 it ultimately dictates the shape of the
00:00:30.00 cell. And so this is fundamentally an
00:00:33.00 interest in the lab because my lab
00:00:35.00 studies the molecular basis of cell
00:00:37.00 shape. So, this movie that I'm playing for
00:00:40.00 you here is a movie of the cytoskeleton
00:00:45.00 underneath the plasma membrane of a
00:00:48.00 filamentous cell within the moss
00:00:51.00 Physcomitrella patens, which is the organism
00:00:53.00 that my lab works on. And in red you see
00:00:56.00 microtubules and in green you see actin
00:00:58.00 filaments, and what I want you to take
00:00:59.00 away from this movie is just how
00:01:01.00 beautifully complex and integral the
00:01:04.00 cytoskeleton is, and how dynamic it is,
00:01:06.00 and that it's constantly remodeling.
00:01:10.00 So, first I need to tell you a little
00:01:11.00 bit about how plant cells divide. They
00:01:14.00 divide using the cytoskeleton and
00:01:18.00 their division is slightly different
00:01:19.00 from animal cells. So, in many plant cells
00:01:23.00 there's a microtubule structure called
00:01:25.00 the pre-prophase band, which is a
00:01:27.00 structure of microtubules right
00:01:29.00 underneath the plasma membrane. And what
00:01:32.00 that does is it essentially marks the
00:01:35.00 cell cortex and it says, this is where
00:01:38.00 the new cell division plane should
00:01:40.00 arrive. That's the pre-prophase band.
00:01:44.00 And then, as mitosis proceeds, and the nuclear
00:01:47.00 envelope falls apart, the microtubules in
00:01:51.00 the pre-prophase band disassemble and
00:01:53.00 they become part of the mitotic spindle.
00:01:55.00 As you can see here, this is a spindle in
00:01:58.00 metaphase, where the chromosomes are all
00:02:00.00 aligned along the metaphase plate. Then
00:02:03.00 anaphase occurs and the chromosomes move
00:02:06.00 to the two poles, and then those
00:02:08.00 microtubules get repurposed again and
00:02:10.00 they get repurposed into a structure
00:02:12.00 called the phragmoplast. So, the phragmoplast
00:02:14.00 is this beautiful structure based...
00:02:17.00 made of microtubules and what it does
00:02:20.00 it helps to move vesicles to the
00:02:23.00 position in the middle of the phragmoplast,
00:02:25.00 and those vesicles are full of
00:02:27.00 cell wall material. So, they fuse together
00:02:29.00 and they make this lipid bilayer and
00:02:34.00 then inside of it is new cell wall
00:02:37.00 material. So, what it's essentially doing is
00:02:39.00 it's templating a new cell plate that's
00:02:42.00 going to then separate the two daughter
00:02:44.00 cells. This phragmoplast sets up in the
00:02:48.00 middle between the two daughter nuclei,
00:02:50.00 but then has to expand from the middle of
00:02:52.00 the cell out to the cell cortex, to that
00:02:55.00 site that had been defined by the
00:02:57.00 pre-prophase band. There, that new membrane
00:03:00.00 and cell wall material fuses with the
00:03:02.00 mother cell wall and cytokinesis is
00:03:05.00 completed. So, microtubules are absolutely
00:03:08.00 essential for this process; if you were
00:03:10.00 to get rid of the microtubule
00:03:11.00 cytoskeleton you would halt mitosis, you
00:03:14.00 would help cytokinesis. But the other
00:03:17.00 cytoskeleton in plant cells, the
00:03:19.00 actin cytoskeleton, is also present. So,
00:03:22.00 the actin cytoskeleton is present in
00:03:24.00 the pre-prophase band. It is also found
00:03:27.00 in the phragmoplast... so, the actin
00:03:29.00 I'm showing you here is in blue and the
00:03:32.00 microtubules were in pink, and the
00:03:36.00 actin cytoskeleton is present throughout
00:03:37.00 the cell cortex, so right underneath the
00:03:40.00 plasma membrane throughout this entire
00:03:42.00 process. And it's been a little bit
00:03:44.00 challenging to figure out what the role
00:03:47.00 of actin is during cell division. What is
00:03:49.00 it doing in the pre-prophase band, what
00:03:51.00 is it doing in the phragmoplast, what
00:03:53.00 is it doing at the cell cortex. And the
00:03:55.00 main reason this has been challenging is
00:03:57.00 because you can actually depolymerize
00:03:58.00 the entire actin cytoskeleton and cell
00:04:01.00 division kind of continues and still
00:04:04.00 happens. And so it's not essential for
00:04:07.00 cell division, but in every single plant
00:04:09.00 species that's been looked at actin is
00:04:12.00 present, so it must be doing something.
00:04:13.00 We just haven't figured out quite what
00:04:16.00 it's doing. So, my lab works on this moss
00:04:21.00 Physcomitrella patens and we're
00:04:23.00 really interested in these beautiful
00:04:25.00 filamentous cells, and one of the things
00:04:28.00 about the moss that is actually sort of
00:04:30.00 attractive is that, in these filamentous
00:04:32.00 cells, they actually divide
00:04:34.00 without having a pre-prophase band. So,
00:04:37.00 the apical cells, the sub-apical cells, they
00:04:40.00 divide without having a pre-prophase band.
00:04:42.00 So, this is an image of microtubules in a
00:04:47.00 living cell and the blue dots would show
00:04:50.00 you that, in this particular image, which
00:04:53.00 is a mid-plane section of a nucleus
00:04:55.00 before nuclear envelope breakdown, that's
00:04:58.00 where you would see microtubule
00:04:59.00 fluorescence if it were there... but it's
00:05:02.00 not there, okay? So, there is no
00:05:05.00 pre-prophase band. Now, I'm going to play
00:05:07.00 this movie and you're going to see the
00:05:09.00 nuclear envelope breakdown, formation of
00:05:10.00 a mitotic spindle, and metaphase, and then
00:05:13.00 it's just going to loop over and over
00:05:14.00 again. So, there is no pre-prophase band
00:05:17.00 in these cells. So, what that allows us to
00:05:20.00 do is ask specifically, what is the role
00:05:23.00 of actin in the phragmoplast, right?
00:05:26.00 We can get rid of the complication of the
00:05:28.00 pre-prophase band because there is no
00:05:29.00 pre-prophase band, and we can dissect,
00:05:31.00 what is the function of actin in
00:05:33.00 phragmoplast? And we got a window into
00:05:37.00 this problem by analyzing a molecular
00:05:40.00 motor that walks on actin filaments.
00:05:43.00 So, myosins are molecular motors that bind to
00:05:45.00 and walk along actin filaments, and
00:05:47.00 they're a very large, diverse family of
00:05:50.00 molecules. This is just a phylogenetic
00:05:54.00 tree of myosins based on their motor
00:05:56.00 domains... in eukaryotes, and you
00:06:00.00 can see that there's at least 35 myosin
00:06:02.00 classes when this particular paper was
00:06:04.00 published, and what's really interesting
00:06:07.00 and has always been very intriguing to
00:06:09.00 me is the fact that plants only have two
00:06:12.00 classes of my myosins. They have a class
00:06:14.00 VIII and then a class XI myosin, and they
00:06:18.00 don't have this huge vast diversity that
00:06:20.00 you can find in other organisms. Humans
00:06:22.00 have many different classes of myosins,
00:06:24.00 for example. So, what are these myosins
00:06:28.00 doing in plants and what is their role?
00:06:30.00 And so, using the reverse genetics
00:06:33.00 afforded to us by the powerful system
00:06:36.00 Physcomitrella patens, we've actually
00:06:37.00 analyzed the function of both plant
00:06:39.00 plant myosin classes. So, myosin XIs are
00:06:44.00 very similar to class VIII myosins from
00:06:47.00 animals and fungi, and we think that
00:06:50.00 they're important in vesicle transport
00:06:51.00 and organelle transport. And we showed
00:06:54.00 using RNA interference that they are
00:06:56.00 essential for polarized growth in moss.
00:06:59.00 And in other plants and seed plants
00:07:02.00 they're important for cytoplasmic
00:07:04.00 streaming and organelle movement. Myosin
00:07:07.00 8s, which are the stars of today's show,
00:07:11.00 are less well studied and less well
00:07:14.00 understood. So, in seed plants, people have
00:07:19.00 not figured out a function yet. There
00:07:22.00 have been limited localization studies.
00:07:24.00 Using antibodies, they've shown that they
00:07:28.00 can localize to the wall after
00:07:31.00 cytokinesis and also to channels that
00:07:34.00 are found between plant cells called
00:07:36.00 plasmodesmata. So, we set out to analyze
00:07:41.00 the function of class VIII myosins in
00:07:42.00 plants... in moss and we used homologous
00:07:46.00 recombination to generate knockouts. So,
00:07:50.00 first I need to tell you that there are
00:07:51.00 five genes that compose the myosin VIII
00:07:54.00 family in moss and they can be grouped
00:07:56.00 into two groups based just on their
00:07:58.00 sequence similarity. So, the purple domain
00:08:01.00 is the motor domain and then the red
00:08:04.00 domain is where the light chains bind and
00:08:06.00 the blue domains are predicted to be
00:08:07.00 coiled-coils, so these are likely to be
00:08:09.00 dimeric myosins. So, we essentially...
00:08:15.00 also looked at the expression of these
00:08:17.00 genes in different tissue types and this
00:08:20.00 is shown here. So, the green bars are the
00:08:23.00 filamentous tissues, the protonemata,
00:08:26.00 and the purple bars are the
00:08:28.00 expression in the leafy shoots or the
00:08:32.00 gametophores. And what you can see is
00:08:34.00 that, with the exception of myosin VIII-D,
00:08:38.00 most of them are expressed in this
00:08:40.00 filamentous tissue stage, the
00:08:43.00 protonemata, and there's not dramatic
00:08:48.00 differences in their expression levels.
00:08:50.00 So, we're going to focus on cell division
00:08:53.00 events in these protonemal filaments
00:08:55.00 because it's cell division in these
00:08:57.00 protonemal filaments that are
00:08:58.00 happening in the absence of a
00:08:59.00 pre-prophase band.
00:09:01.00 Okay. So, we generated a panel of mutants.
00:09:06.00 Using homologous recombination, we
00:09:07.00 knocked out the myosin VIII genes.
00:09:10.00 On the top panel, we have knockouts,
00:09:13.00 single knockouts. We generated some
00:09:16.00 double knockouts, one triple knockout, two
00:09:18.00 quadruple knockouts, and then, at the
00:09:20.00 bottom there, you see regenerating plants
00:09:23.00 of the quintuple knockout. So, we actually
00:09:25.00 knocked out all five class VIII myosins.
00:09:28.00 So, the first surprise is that you can
00:09:31.00 live without class VIII myosins. But, why hold
00:09:34.00 on to five genes if you can live without
00:09:35.00 them? So, that's something that still
00:09:37.00 bothers me at night, but it is, you know...
00:09:40.00 it is a fact that these myosins are
00:09:43.00 not essential for viability. However,
00:09:45.00 there are some differences. So, these
00:09:47.00 plants are plants that were regenerated
00:09:49.00 from a single cell and there are about
00:09:51.00 five to six days old, and we're imaging
00:09:54.00 them with a fluorescent microscope and
00:09:56.00 we're looking at the fluorescence in
00:09:58.00 their wall, because we've added a dye
00:10:00.00 that binds to the wall. And so that
00:10:02.00 allows us to measure the area of the
00:10:04.00 plant because we measure the area of the
00:10:06.00 fluorescence, and so wild type we set to
00:10:09.00 one and we normalize all our data to
00:10:11.00 wild type. So, what you can see is that
00:10:13.00 the single mutants are all a little bit
00:10:16.00 smaller and then subsequent knockouts
00:10:19.00 eventually result in the quintuple knockout,
00:10:22.00 where you are a plant that's about
00:10:24.00 58% the size of
00:10:26.00 wild type. So, there is a growth defect. And
00:10:30.00 if you look really carefully at these
00:10:32.00 young plants, we notice that there is
00:10:34.00 also a defect in positioning of the
00:10:36.00 cell plate. So, wild type plants that are
00:10:40.00 five days old have what I'm calling
00:10:43.00 transverse cell plates or an angle of zero.
00:10:46.00 And the myosin VIII mutant, as you can
00:10:50.00 see from this histogram, has a large
00:10:53.00 population of cells that have oblique
00:10:55.00 cell plates. So, instead of positioning
00:10:58.00 the cell plate perfectly transversely to
00:11:00.00 the length... to the long axis of the cell
00:11:02.00 there's slop -- there's some mistake here.
00:11:05.00 So, to get an idea of how this myosin is
00:11:10.00 working, we wanted to generate a fusion
00:11:12.00 of this molecule with a green fluorescent protein,
00:11:14.00 so, we could analyze its localization.
00:11:16.00 And so we decided to look at myosin 8A
00:11:22.00 because... so the orange molecules are the
00:11:25.00 ones that are all very similar to each
00:11:26.00 other and the blue ones are the ones
00:11:27.00 that are similar to each other, so those
00:11:29.00 are the two groups, the group A and the
00:11:30.00 group B. And of group A, myosin 8A is the
00:11:34.00 most highly expressed. So, we thought, okay,
00:11:36.00 we'll try that one and will fuse it to
00:11:38.00 three tandem GFP molecules, and we'll
00:11:41.00 transform it back into the quintuple
00:11:43.00 knockout. By transforming it back into
00:11:45.00 the quintuple knockout, that allowed us
00:11:47.00 the opportunity to see if this molecule
00:11:49.00 was functional because we could see if
00:11:51.00 it rescued the phenotypes that we had
00:11:53.00 observed in the quintuple knockout. So,
00:11:56.00 this transgene largely rescues most of
00:12:00.00 the phenotypes that we observe that I
00:12:01.00 don't have time to talk to you about, but
00:12:03.00 the one phenotype that's important for
00:12:05.00 today is the cell plate positioning. And
00:12:07.00 as you can see in this green histogram
00:12:10.00 here, where we've taken the myosin 8
00:12:12.00 fused to the GFP and we've put it into
00:12:14.00 the quintuple knockout, we see a shift of
00:12:17.00 that distribution back to more wild type.
00:12:19.00 So, that tells us that this GFP fusion
00:12:22.00 protein is a functional protein and we
00:12:24.00 then analyze its localization. So, first I
00:12:28.00 need to tell you that this is what
00:12:30.00 wild-type cells look like when we put
00:12:32.00 them on our confocal microscope. So, those
00:12:34.00 are beautiful chloroplasts that
00:12:37.00 autofluoresce in this particular channel, and
00:12:39.00 so whenever you see chloroplasts in
00:12:42.00 many of the images that I’m going to
00:12:43.00 show you, that's just autofluorescence.
00:12:46.00 Myosin 8-GFP plants look like this.
00:12:48.00 So, you can see that there's a very big
00:12:50.00 difference between the wild type and the
00:12:51.00 myosin 8-GFP in that the cytosol now
00:12:54.00 lights up with the fluorescence, right? So,
00:12:57.00 those chloroplasts do not
00:12:59.00 have any myosin 8 in them but the
00:13:01.00 cytoplasm does have a lot of myosin 8-GFP,
00:13:03.00 and hopefully you can see that there
00:13:06.00 are little particles, little intense
00:13:08.00 particles, and some of those particles
00:13:10.00 seem to be more enriched at the tip of
00:13:12.00 the cell. So, this is what sort of the
00:13:15.00 global view looks like with, for example,
00:13:17.00 a spinning disk confocal microscope, but
00:13:20.00 we wanted to look right underneath the
00:13:21.00 plasma membrane and we wanted to analyze
00:13:23.00 the localization of these molecule at the
00:13:26.00 plasma membrane.
00:13:27.00 So, we use total internal reflection
00:13:29.00 fluorescence microscopy to do that and
00:13:31.00 what I'm showing you here is an image
00:13:33.00 from a TIRF microscope, where you can see
00:13:36.00 the myosin 8-GFP forms these beautiful
00:13:38.00 little dynamic particles that are moving
00:13:41.00 along the surface of the plasma membrane.
00:13:42.00 And if, then, you want to ask, what about
00:13:45.00 actin? You can then look at actin using a
00:13:48.00 probe called Lifeact that binds to
00:13:51.00 actin filaments, and what you can see is
00:13:53.00 that many of those particles are moving
00:13:55.00 along actin filaments. And this is what
00:13:58.00 we would expect. This is a myosin motor;
00:14:00.00 it's supposed to walk on actin filaments.
00:14:02.00 So, this is very gratifying to see.
00:14:04.00 We actually measured the motility
00:14:07.00 along these filaments and we were able
00:14:09.00 to find out how fast they moved, how long
00:14:12.00 their trajectories were as well. So, one
00:14:14.00 day my graduate student came to the lab
00:14:16.00 and said, I got this image last
00:14:18.00 night and this looks like a really
00:14:20.00 interesting structure, and I thought, oh
00:14:22.00 my goodness, this looks like mitotic
00:14:23.00 spindle. And I thought, how is it the myosin 8
00:14:26.00 is localizing to a mitotic spindle.
00:14:28.00 We know that actin's in the phragmoplast,
00:14:30.00 but we hadn't really ever seen actin in
00:14:33.00 a mitotic spindle -- that's made of microtubules.
00:14:34.00 So, in order to really validate
00:14:37.00 that this structure was a mitotic
00:14:39.00 spindle, we needed to generate a line
00:14:41.00 that had the fluorescent myosin, also had
00:14:44.00 fluorescent microtubules. So, we generated a
00:14:47.00 line that had both of these labeled with
00:14:49.00 two different colors. And this is going
00:14:53.00 to be a movie where I'm going to play
00:14:56.00 myosin 8 in green and mCherry tubulin in red
00:14:59.00 in the merge. And this
00:15:03.00 is right before nuclear envelope
00:15:04.00 breakdown, so you can see those beautiful
00:15:06.00 microtubules that are going right around
00:15:08.00 the nucleus, then the nucleus is going to
00:15:10.00 break down and you're going to form the
00:15:12.00 mitotic spindle. So, as we play this movie
00:15:15.00 you can see that my myosin 8 is clearly
00:15:16.00 localizing to microtubules throughout
00:15:19.00 the entire division process, and as the
00:15:23.00 nuclear envelope breaks down at the
00:15:25.00 beginning of the movie it immediately
00:15:26.00 binds to the spindle, and then, as it
00:15:29.00 turns into a phragmoplast, which
00:15:30.00 happens very quickly after that, you can
00:15:32.00 see a very fine band of myosin 8
00:15:35.00 that localizes to the leading edge of
00:15:37.00 the phragmoplast. So, that was very
00:15:39.00 exciting. So, we
00:15:41.00 to ensure that actin really isn't in the
00:15:43.00 mitotic spindle because maybe it is in
00:15:46.00 these cells and we hadn't looked
00:15:47.00 carefully enough. So, we looked at a line
00:15:49.00 that had myosin 8 and actin. And so, what you can
00:15:53.00 see here is that myosin 8 localizes to the
00:15:55.00 spindle, but it... when it turns into that
00:15:58.00 very, very narrow band, that's when
00:16:00.00 actin appears. Before then, really, actin
00:16:04.00 is not accumulating. So, actin is not in
00:16:07.00 the mitotic spindle but it seems to be
00:16:09.00 appearing right when that spindle
00:16:10.00 transitions into this phragmoplast
00:16:12.00 that's so important for generating the
00:16:14.00 new cell plate. So, then we thought, okay,
00:16:17.00 well, we have these tools, we have these
00:16:19.00 drugs that we can get rid of the actin
00:16:20.00 cytoskeleton, so what happens if we get
00:16:22.00 rid of the actin cytoskeleton? So, we
00:16:24.00 treated cells with Latrunculin,
00:16:25.00 which is a drug that depolymerizes
00:16:27.00 actin filaments, and then we looked at
00:16:29.00 myosin, and we had mCherry tubulin so
00:16:32.00 we could look at mitotic spindles and
00:16:34.00 phragmoplasts. And what you can see is
00:16:36.00 that myosin localizes just fine to the
00:16:38.00 mitotic spindle and it localizes just
00:16:40.00 fine to the mid-zone of the phragmoplast
00:16:42.00 during phragmoplast expansion.
00:16:45.00 So, this was very puzzling, we were a little
00:16:47.00 concerned, so we did a little recap. Okay.
00:16:50.00 Myosins are actin-based motors. We know
00:16:52.00 this and they're walking along actin
00:16:54.00 filaments in the cell cortex. But during cell
00:16:57.00 division they seem to be binding the
00:16:59.00 microtubules, not actin necessarily, and
00:17:02.00 they show up at the site of cell
00:17:04.00 division before actin ever shows up, and
00:17:07.00 it doesn't seem like our localization
00:17:10.00 depends on actin during cell division. So,
00:17:12.00 this was a little bit puzzling, we were a
00:17:14.00 little bit concerned, and so we asked the
00:17:16.00 question, does myosin 8 actually work
00:17:18.00 with actin during cell division or is it
00:17:20.00 doing something entirely novel? Did we
00:17:23.00 actually find a kinesin or something?
00:17:26.00 So, we looked more carefully. So, again,
00:17:29.00 this is this movie that I've already
00:17:30.00 shown you where myosin 8 is localizing
00:17:33.00 to the mitotic spindle, there, and then as
00:17:37.00 it turns into a phragmoplast,
00:17:38.00 actin accumulates in the phragmoplast zone. So, we
00:17:42.00 thought, maybe what's going on is that
00:17:44.00 myosin 8 needs to get the right area and
00:17:48.00 it gets to the right area by interacting
00:17:50.00 with microtubules but really its
00:17:52.00 function is during cytokinesis,
00:17:54.00 and that function works with actin. So,
00:17:58.00 to really address that what we did was
00:18:00.00 look very carefully at expanding
00:18:02.00 phragmoplasts. So, this is a wild-type
00:18:05.00 cell labeled with GFP-tubulin and the
00:18:08.00 membranes that are getting incorporated
00:18:10.00 into that new cell plate are labeled
00:18:12.00 with FM4-64 which is a lipophilic
00:18:14.00 membrane dye. And so what you can see is
00:18:17.00 this beautiful phragmoplast structure makes
00:18:20.00 a very linear sort of line that's very
00:18:24.00 uniform, and you form a new cell plate.
00:18:28.00 And then these are just examples, still
00:18:31.00 images, of a beautiful, nicely formed phragmoplast
00:18:35.00 with membrane in the center, in
00:18:37.00 wild-type cells. So, then we thought, okay,
00:18:40.00 well, let's look at what happens in the
00:18:42.00 myosin 8 mutant. So, in the myosin 8
00:18:45.00 mutant, again, there is your phragmoplast...
00:18:47.00 spindle and then it turns into a
00:18:50.00 phragmoplast about now. and you can
00:18:52.00 start to see membranes getting
00:18:54.00 incorporated into that phragmoplast. And
00:18:57.00 you can see that the phragmoplast is
00:18:59.00 messier, it's not as uniform, and you can
00:19:03.00 see that the membranes are not getting
00:19:05.00 incorporated into a nice linear line. And
00:19:07.00 over here we have these stills, alright?
00:19:11.00 And in the stills you can see that the
00:19:12.00 membrane is waves and then you can see
00:19:16.00 that you can form these sort of not very
00:19:19.00 normal-looking cell plates. So, it does
00:19:23.00 look like in the myosin 8 mutant
00:19:25.00 that phragmoplast expansion and the
00:19:28.00 building of the new cell plate is
00:19:30.00 aberrant. So, if myosin 8 requires
00:19:34.00 actin for its function during phragmoplast
00:19:36.00 expansion, then we should be able
00:19:39.00 to see a copy this mutant phenotype by
00:19:42.00 getting rid of actin. And so that's what
00:19:44.00 we did. We looked at cells treated with
00:19:47.00 Latrunculin during this stage of
00:19:49.00 phragmoplast expansion and what we saw was
00:19:52.00 that we got very similar results to
00:19:57.00 knocking out the five myosins. We see a
00:20:01.00 disorganized phragmoplast and we see
00:20:04.00 buckling of the membranes as the membranes
00:20:06.00 incorporate. And in these still
00:20:09.00 images, which are just other examples of
00:20:11.00 this happening, you can see that the
00:20:14.00 phragmoplasts are aberrant and you can see
00:20:16.00 that the planes of the membranes are not
00:20:19.00 uniform. So, that led us to think, okay, so
00:20:22.00 this is probably a myosin and it
00:20:24.00 probably walks on actin filaments, and
00:20:26.00 it's probably doing a job very
00:20:28.00 specifically during cytokinesis and
00:20:31.00 during cell expansion... during phragmoplast
00:20:34.00 expansion. So, this is a movie, now,
00:20:36.00 where we're looking at myosin 8
00:20:39.00 and microtubules, and we're going to play
00:20:42.00 it through once as the phragmoplast
00:20:44.00 expands, and then I want you to focus where
00:20:48.00 that circle is because there are a lot
00:20:51.00 of peripheral cytoplasmic... of peripheral
00:20:55.00 microtubules at the edge of this
00:20:56.00 structure. And those peripheral
00:20:59.00 microtubules, interestingly, all have
00:21:02.00 myosin 8s on their ends. So, it looks like
00:21:04.00 these peripheral microtubules interact...
00:21:07.00 have myosin 8 on the ends and we know
00:21:09.00 that in this structure the ends of these
00:21:11.00 microtubules are the plus ends of the
00:21:13.00 microtubules. So, it looks like myosin 8
00:21:15.00 is binding to the plus ends of these
00:21:17.00 peripheral microtubules, and they very,
00:21:20.00 very rapidly incorporate into the
00:21:22.00 expanding phragmoplast as the
00:21:24.00 phragmoplast expands. They also search the
00:21:27.00 cell cortex, where you can also see an
00:21:29.00 accumulation of myosin 8 at the cell
00:21:31.00 cortex. So, it looks like myosin 8 is
00:21:35.00 helping to direct the expansion of this
00:21:37.00 structure to the right place on the cell
00:21:39.00 cortex. If we look at a very similar cell,
00:21:43.00 but now a cell that is treated with
00:21:47.00 Latrunculin, so there are no actin
00:21:49.00 filaments, what we see is that now myosin 8
00:21:53.00 highly decorates those peripheral
00:21:56.00 microtubules. They stay associated with
00:21:59.00 the cell cortex for a very long period
00:22:01.00 of time, as those search and search
00:22:04.00 and search, but there are no actin
00:22:05.00 filaments around, so those myosin 8s
00:22:07.00 just stay bound to the plus ends of the
00:22:09.00 microtubules, and the whole structure
00:22:12.00 actually can slip. So, instead of being
00:22:16.00 stuck in one place in the middle of the
00:22:18.00 cell, the structure actually slips
00:22:20.00 up and down, and sometimes skews. So, our
00:22:24.00 hypothesis based on these images in
00:22:26.00 these movies was that we think that
00:22:28.00 actin is actually forming in the mid-zone,
00:22:30.00 the edge of the phragmoplast. So, we
00:22:34.00 work on molecules that polymerize actin
00:22:37.00 filaments and we thought, well, let's look
00:22:39.00 at them during cell division and see
00:22:40.00 where they are. And so we have a line
00:22:44.00 that has a functional fusion of formin 2A,
00:22:47.00 which polymerizes actin filaments
00:22:50.00 and... fused to GFP... and a line that has.
00:22:53.00 mCherry tubulin. And as this... this... this is
00:22:56.00 a spindle that's in metaphase and it's
00:22:58.00 going to go into anaphase when I play
00:23:00.00 the movie. So, you can see anaphase from
00:23:02.00 those shadows moving apart from each
00:23:04.00 other and then you can see formin 2A
00:23:06.00 accumulate beautifully in the mid-zone
00:23:09.00 of the phragmoplast as the phragmoplast
00:23:11.00 expands out to the mother cell
00:23:13.00 cortex. So, formin is exactly where it
00:23:16.00 needs to be if actin filaments are being
00:23:18.00 polymerized off of this expanding
00:23:20.00 structure. So, then we needed to just bite
00:23:23.00 the bullet and look at actin and
00:23:24.00 microtubules at the same time. Actin is
00:23:27.00 very challenging to image because it's
00:23:28.00 very, very dynamic, and so we were
00:23:31.00 fortunate enough to be able to get
00:23:33.00 good images of this using a fairly fast
00:23:36.00 microscope. And you can see here... actin
00:23:39.00 is in green, microtubules are in red in
00:23:42.00 this merge image, and the circle is just
00:23:44.00 pointing to that area where those
00:23:45.00 peripheral microtubules are and they're
00:23:48.00 embedded in a meshwork of actin
00:23:50.00 filaments, and that meshwork of actin
00:23:52.00 filaments is connected to the cell
00:23:53.00 cortex. So, the microtubules are being
00:23:57.00 templated by this meshwork to
00:23:59.00 incorporate in along a specific plane in
00:24:02.00 the middle of the cell, allowing for the
00:24:04.00 expansion to occur and so that that
00:24:06.00 structure doesn't slip up or down, and
00:24:08.00 doesn't tweak. And so it's a... and, remember,
00:24:11.00 those peripheral microtubules have
00:24:12.00 myosin 8 on their plus ends, so they
00:24:14.00 are then incorporating and interacting
00:24:17.00 with the actin filaments to then make
00:24:18.00 this structure incorporate the right way.
00:24:20.00 So, this led us to a model where we think
00:24:25.00 myosin 8 is binding to the mitotic
00:24:27.00 spindle essentially as a way to get to
00:24:29.00 the right place and be there at the
00:24:31.00 right time. And then umm... in anaphase,
00:24:34.00 myosin 8 concentrates at
00:24:36.00 the cell cortex, it also concentrates in
00:24:38.00 the middle of the spindle, and also in
00:24:40.00 the poles. We don't understand the polar
00:24:42.00 concentration, we haven't figured that
00:24:43.00 out yet, but we understand what's going
00:24:46.00 on with this in the middle, here. As this
00:24:49.00 turns into a phragmoplast, then actin
00:24:53.00 gets polymerized off of the middle there,
00:24:55.00 towards the cell cortex where myosin 8
00:24:58.00 can hold on to it, and that essentially
00:25:00.00 makes a network that then the
00:25:02.00 microtubules will connect to and then
00:25:05.00 incorporate along this plane, allowing
00:25:08.00 for cell division to occur along it in
00:25:10.00 the correct plane and divide the cell
00:25:12.00 properly. So, that's what's happening in
00:25:15.00 division in an apical cell, in these
00:25:18.00 protonemal filaments. So, we were
00:25:21.00 wondering about sub-apical cells, because
00:25:23.00 these sub-apical cells, as you can see
00:25:25.00 here in these division events, are
00:25:27.00 asymmetric events, where you actually
00:25:30.00 have to move the nucleus to the division
00:25:33.00 site and then division occurs at the
00:25:35.00 division site. So, we know that there are
00:25:38.00 no pre-prophase bands there, but what does
00:25:40.00 myosin 8 look like when you have this
00:25:42.00 branching event where you have to have
00:25:44.00 this large nuclear migration. One of the
00:25:48.00 things that we noticed right away in our
00:25:50.00 mutant analysis is that the myosin 8
00:25:52.00 knockout, so this is the quintuple
00:25:54.00 knockout, has dramatic defects in cell
00:25:58.00 plate positioning right at branch sites.
00:26:01.00 So, this is an image with Calcofluor
00:26:03.00 fluorescence, where Calcofluor binds very
00:26:06.00 specifically to new cell plates. And so
00:26:08.00 you can see that some of these cell
00:26:09.00 plates are incomplete, and some of them
00:26:12.00 are very aberrant in their positioning. So,
00:26:16.00 it seems as if the phenotype actually is
00:26:19.00 enhanced at branch sites. So, myosin 8
00:26:23.00 at branch site localizes to the emerging
00:26:26.00 branch neck, and it localizes to the
00:26:29.00 emerging branch neck before mitosis even
00:26:32.00 begins. And once this movie will loop
00:26:34.00 you'll see that as the cell goes into
00:26:38.00 mitosis and then there's a phragmoplast,
00:26:39.00 you actually have two rings of myosin 8
00:26:41.00 that are concentric: one that was
00:26:43.00 very static on the cell cortex and then
00:26:46.00 one that's on the phragmoplast,
00:26:47.00 that expands out and reaches that
00:26:49.00 cortical population. So, this actually
00:26:52.00 looks like a pre-prophase band... except it
00:26:56.00 doesn't have any microtubules -- it just
00:26:58.00 has myosin 8. So, myosin 8 is defining
00:27:01.00 the cortical region that the phragmoplast
00:27:04.00 needs to expand to, okay? And this requires
00:27:08.00 actin, because if we treat cells with
00:27:12.00 Latrunculin, you'll see that the myosin 8
00:27:15.00 doesn't have any actin filaments
00:27:18.00 to interact with and you see buckling of
00:27:21.00 the cell... of the cell plate as it's
00:27:23.00 expanding. And this is the phenotype that
00:27:26.00 we see so clearly in the myosin 8
00:27:28.00 knockout. Okay. So, that's what's happening
00:27:33.00 in moss, where there is no pre-prophase
00:27:35.00 band. But we wanted to ask the question,
00:27:38.00 well, what happens in a plant cell that
00:27:40.00 does have a pre-prophase band?
00:27:43.00 Where would my myosin 8 localize in that cell?
00:27:45.00 And the other thing is that we needed to
00:27:47.00 redraw what the actin looks like in an
00:27:50.00 expanding phragmoplast. So, the actin
00:27:53.00 actually is coming out of the center of
00:27:55.00 the phragmoplast and it's not
00:27:56.00 necessarily collinear with the phragmoplast,
00:27:59.00 which is what's in the textbooks. Okay.
00:28:02.00 So, in order to analyze myosin 8
00:28:05.00 localization in a cell that has a
00:28:07.00 pre-prophase band, we decided to turn to
00:28:10.00 tobacco BY2 suspension culture cells,
00:28:12.00 because this is a plant cell culture
00:28:14.00 line that has been used to analyze cell
00:28:17.00 division for years and years, and
00:28:18.00 cytokinesis as well. And it's a beautiful cell
00:28:21.00 type that is very large and very easy to
00:28:23.00 image. So, we actually decided to put the
00:28:27.00 moss myosin 8 into these cells and see
00:28:29.00 how the moss myosin 8 localizes. And, to our
00:28:32.00 delight, we got images that look like
00:28:34.00 this. So, this is a cell that's about to
00:28:39.00 undergo mitosis -- you can actually see a
00:28:41.00 halo of the nuclear envelope -- and you can
00:28:44.00 see a band of myosin 8 right
00:28:48.00 underneath the cell cortex. This is
00:28:49.00 exactly what the pre-prophase ban looks
00:28:52.00 like. So, myosin 8 from moss is binding
00:28:56.00 to the pre-prophase band of tobacco
00:28:58.00 cells. And what happens
00:29:00.00 in cytokinesis? So, in the cortical
00:29:05.00 division site... you can see it at the
00:29:06.00 cortical division site in a maximum
00:29:08.00 projection image, but if you look at the
00:29:10.00 midplane of that same cell what you see
00:29:14.00 is the phragmoplast... I'm going to point
00:29:16.00 out here in circles where the cortical
00:29:18.00 division site is right at the cell
00:29:20.00 cortex, but then you also see myosin 8
00:29:23.00 accumulating in the mid-zone of
00:29:24.00 the phragmoplast. So, myosin 8 is
00:29:27.00 both at cortical division site and in
00:29:29.00 the phragmoplast mid-zone. So, if we look
00:29:32.00 at a cell that's undergoing division and
00:29:34.00 we label myosin 8 and membranes, we
00:29:37.00 can see here that the myosin 8 is in
00:29:40.00 green and the membranes labeled by FM4-64
00:29:43.00 are in red... what you can see is that
00:29:47.00 the phragmoplast expands out to a site,
00:29:50.00 and that arrow is pointing to the myosin 8
00:29:53.00 accumulation at the cortical
00:29:54.00 division site. So, you have myosin 8 both
00:29:57.00 in the phragmoplast mid-zone and at the
00:29:59.00 cortical division site, providing a
00:30:01.00 mechanistic link between the expansion
00:30:04.00 of the phragmoplast and the site where
00:30:06.00 it needs to join the mother cell wall.
00:30:09.00 So, what I've told you today is that during
00:30:11.00 cell division myosin 8 marks the
00:30:14.00 future site of division in these
00:30:15.00 branching cells. In the apical cells that
00:30:18.00 divide symmetrically, that mark happens
00:30:20.00 around anaphase, but in these cells where
00:30:23.00 you have to actually move the nucleus to
00:30:25.00 the site of division, it happens before
00:30:27.00 mitosis begins. And then myosin 8
00:30:31.00 localizes to the mitotic spindle and to
00:30:34.00 the phragmoplast and you have this
00:30:36.00 internal ring of myosin 8 on the
00:30:39.00 expanding phragmoplast that grows out
00:30:42.00 and reaches the cortical population. So,
00:30:46.00 myosin 8, together with actin, provide
00:30:49.00 this mechanistic and spatial link
00:30:51.00 between the phragmoplast and the
00:30:53.00 cortical division site. So, with that, I
00:30:56.00 want to thank you for watching and I
00:30:58.00 want to thank you Shu-Zon Wu, a very talented
00:31:00.00 graduate student and postdoc, who has been
00:31:02.00 working on this project and has been just a
00:31:04.00 wonderful imager. Thank you.

Videos in this Talk
  • Part 1: Understanding Cell Shape: Big Insights From Little Plants
    Part 1: Understanding Cell Shape: Big Insights From Little Plants
    Audience:
    • Student
    • Researcher
    • Educators of H. School / Intro Undergrad
    • Educators of Adv. Undergrad / Grad
  • Part 2: Using Reverse Genetics to Dissect the Formin Gene Family
    Part 2: Using Reverse Genetics to Dissect the Formin Gene Family
    Audience:
    • Student
    • Researcher
    • Educators of H. School / Intro Undergrad
    • Educators of Adv. Undergrad / Grad
  • Part 3: Myosin and Actin Steer Plant Cell Division
    Part 3: Myosin and Actin Steer Plant Cell Division
    Audience:
    • Researcher
    • Educators of Adv. Undergrad / Grad
Speaker: Magdalena Bezanilla
Total Duration: 1:36:04
Recorded: July 2016
All Talks in Plant Biology

Talk Overview

Magdalena Bezanilla’s lab is interested in understanding what determines cell shape at the molecular level. In Part 1 of her talk, Bezanilla explains why the moss, Physcomitrella patens, is an excellent model system to address this question. P. patens grows quickly and entire plants can be vegetatively regenerated from single cells. The genome is sequenced and its cells undergo efficient homologous recombination- the only plant known to do this! In the filamentous tissue that establishes the plant, cells at the tips grow in a polarized manner as flexible cell wall material is incorporated and then pushed out by turgor pressure. To study the role of actin in determining cell shape, Bezanilla’s lab used a live-cell actin binding probe to label actin and imaged cell growth with fluorescence microscopy. They showed that polarized growth depended on the rapid disassembly and assembly of actin filaments; processes regulated by actin binding proteins.

Formins are a family of actin binding proteins that help to nucleate and elongate actin filaments. In Part 2 of her talk, Bezanilla describes experiments in which her lab used RNA interference to examine the role of the formin gene family in Physcomitrella. They showed that class II formins specifically are required for polarized growth and that they actively generate new actin filaments at the cell cortex and the cell tip.

In Part 3, Bezanilla examines the role of myosins in plant cell division. Thirty-five classes of myosins are known, yet, only two classes, myosins VIII and XI are found in plants. While Myosin XI is required for cell viability, myosin VIII is not. However, myosin VIII mutants have improperly positioned cell plates or cell division defects. Results from Bezanilla’s lab detail the molecular role of myosin VIII in plant cell division.

Speaker Bio

Magdalena Bezanilla

Magdalena Bezanilla

Magdalena Bezanilla’s lab investigates the role of the actin cytoskeleton in regulating cell growth and cell shape in plants. She has promoted the use of the moss Physcomitrella patens because of its ability to undergo homologous recombination, a trait that, among land plants, is unique to Physcomitrella. Her lab has developed numerous other tools that have allowed them… Continue Reading

Playlist: Cytoskeletal Proteins

Related Resources

Vidali L, Bezanilla M. Physcomitrella patens: a model for tip cell growth and differentiation. Curr Opin Plant Biol. 2012 Dec;15(6):625–31.

Augustine RC, Pattavina KA, Tüzel E, Vidali L, Bezanilla M. Actin interacting protein1 and actin depolymerizing factor drive rapid actin dynamics in Physcomitrella patens. Plant Cell. 2011 Oct;23(10):3696–710.

Vidali L, van Gisbergen PAC, Guérin C, Franco P, Li M, Burkart GM, et al. Rapid formin-mediated actin-filament elongation is essential for polarized plant cell growth. Proc Natl Acad Sci USA. 2009 Aug 11;106(32):13341–6.

van Gisbergen PAC, Li M, Wu S-Z, Bezanilla M. Class II formin targeting to the cell cortex by binding PI(3,5)P(2) is essential for polarized growth. J Cell Biol. 2012 Jul 23;198(2):235–50.

Wu S-Z, Ritchie JA, Pan A-H, Quatrano RS, Bezanilla M. Myosin VIII regulates protonemal patterning and developmental timing in the moss Physcomitrella patens. Mol Plant. 2011 Sep;4(5):909–21.

Wu S-Z, Bezanilla M. Myosin VIII associates with microtubule ends and together with actin plays a role in guiding plant cell division. eLife 2014 Sep 23;3:e03498.

 

Reader Interactions

Leave a Reply

Your email address will not be published. Required fields are marked *