Ruth Lehmann (NYU / HHMI) 2: Establishing Soma-Germline Dichotomy

Hello, my name is Ruth Lehmann I’m a professor at NYU School of Medicine, and I’m also a Howard Hughes Medical Institute Investigator This is Part 2 of telling you about the germ line In this part, I will focus on how germ cells, which make egg and sperm, become so different from the soma, which makes the rest of the body So, let’s look at this question In the Drosophila embryo, at the very early stages, cells have to make a decision whether to make soma, which is the whole embryo, pretty much, and then how to set aside a few cells which will make the germ line And those few cells will be able to — through egg and sperm — give rise to a whole new generation So, what I want to tell you in this episode is, how do germ cells actually form? Because in flies, there is a very specific way how they form So, the differences between soma and germ line are so great that even the cells of the body of the fly and the cells of the germ line of the fly form by completely different mechanisms And then I will tell you how these future germ cells are making sure that they don’t become somatic cells, and they don’t share the eventual deadly fate of the germ line of the soma So, we’re talking about germ line/soma distinction And the germ line is set aside in Drosophila, as I explained in Part 1, by germ plasm, which is maternally synthesized and deposited in the egg And then, what happens during the early stages of development is nuclei divide so, there are no cells yet formed, and then these nuclei move out to the surface of the egg, or zygote at that point, and then those nuclei which migrate into the germ plasm become specialized as germ cells And they will be surrounded by cells at that time, and form the primordial germ cells Then, there are another four synchronous divisions of nuclei And at that point, the somatic cells are formed So, the germ line, including the formation of the germ cells, is solely dependent on these maternal transcripts Indeed, you can inhibit transcription completely in the early embryo and germ cells will still form On the other hand, the soma requires new somatic transcription And that means the embryo has to transcribe genes to allow the cellularization event of the soma to occur So, there are two separate genetic pathways which regulate these two fates of cells And so the maternal transcripts will lead to the formation of these cells at the posterior pole, which are also sometimes referred to as pole cells And there are two genes that I will be talking about today — germ cell-less and polar granule component — which play an important role in setting the germ cells apart And if you have listened to Part 1, these are two of the localized RNAs that I was talking about in Part 1 The soma is formed by the ingression of cell membranes in between the nuclei that have moved to the surface And there are a number of transcripts which are synthesized by and expressed by the zygote, by the embryo, that are absolutely necessary for this ingression of the membranes, and the formation of the soma at a later stage And this is sort of an interesting story in trying to observe a process that is happening And this was done by Ryan Cinalli using two-photon microscopy And Ryan was showing us this movie during a group meeting, and I want you to get sort of the feeling of trying to understand what is happening when you’re watching a movie And so what you see here is… there is an opening, which is an opening of the cells

which are… the nuclei have migrated into the cytoplasm, there’s a bud, and this opening is where the embryo is gonna be… the rest of the embryo is going to be And then the germ cells will be forming over there But that is only one of the constrictions which are occurring When you see the movie, you will also see a constriction which is orthogonal, which is an anaphase constriction, which leads to cell division So, now I’m gonna show you the movie So, here you see the constriction, and then the anaphase furrow Let me describe this process in a still So, we’re having the bud neck furrow, which separates the germ cell… future germ cell from the rest of the embryo And then we have an anaphase furrow, which leads to the division of the cell So, as a result, we have, from one bud, two primordial germ cells formed Now I’m gonna run the movie again So, now you can see the bud neck [unknown] and the anaphase So, two orthogonal furrows creating two primordial germ cells So, how does this happen? So, there’s independent control of these two furrows And the way we figured this out was by inhibiting the anaphase furrow So, this is the normal constriction which occurs at mitosis And it can be inhibited, because if you inhibit microtubules and you inhibit the spindle apparatus, then this constriction does not occur And what you can see here is… this is just the control, where we have a normal cell formed, but in the… when we inhibit it, the anaphase constriction, we make now two nuclei, because they’re in the same cell But what you can also see… we still have a bud neck constriction, just like in the control So, that means this process only affected one of the two aspects And so microtubules are apparently not required for the bud neck constriction On the other hand, a mutant that had been identified many years ago called germ cell-less just specifically affects the bud neck constriction, and not the anaphase constriction And so here’s the control, again You have the anaphase constriction starting And we’re watching this all by looking at anillin-GFP, which helps us to see these furrows And then we have the bud neck constricting and the anaphase furrow coming down And then in germ cell-less mutants, the anaphase furrow still comes down perfectly fine, but there’s no bud neck construction And indeed, Ryan could show that the amount of germ cell-less was directly proportional to how much that second furrow formed These types of cytokinesis-independent constrictions are not reserved just for making the germ cells in Drosophila, but they have been observed in other contexts So, for example, as a mechanism of polar lobe formation, where certain determinants for development of Illyanassa and other mollusks are temporarily separated from the rest of the embryo Or in C elegans, where the primordial germ cells form a lobe, which then is eaten by the endoderm Or also, during the asymmetric division of neuroblasts in Drosophila So, possibly, there are more of these examples We do not know if they’re all based on the same molecular mechanism, which I will be telling you about next So, what I will be focusing on now is to tell you about how germ cell-less is blocking somatic signaling pathway by degradation, and thereby allowing this furrow to happen and to close And then I will also tell you about polar granule component, another one of these localized RNAs, which blocks the transcription in the germ cells So, let’s focus with germ cell-less As I mentioned earlier, germ cell-less was identified quite a few years ago And it was identified as a localized RNA, where the RNA is localized at the posterior pole It’s then translated into a protein, and this protein beautifully decorates the inner nuclear membrane The phenotype of germ cell-less mutants… and keep in mind, these are maternal effect mutations So, the mother provided the germ cell-less RNA, so when I talk about a mutant embryo, I mean an embryo which came from a mutant mother And so, an embryo which comes from a mutant mother, which lacks completely the germ cell-less protein, in most of the cases forms no germ cells In some cases, they form a few cells And we can quantify this here by counting, actually, the primordial germ cells per embryo

And again, you can, here, much more quantitatively see that there are a large number of embryos which have no germ cells So, what is the mechanism by which germ cell-less promotes this spindle-independent constriction? And this brings us to the sequence of germ cell-less, and also to the biochemical and functional analysis of germ cell-less And also to another graduate student in my lab, Juhee Pae, who at this point ran a half-marathon at the beginning of her PhD Now, she just received her degree, and has also run a full marathon So, the marathon here is really to understand how this protein acts Germ cell-less is a BTB/BACK protein, and it has a few other domains which I will tell you about So, what are BTB/BACK domain proteins? They’re quite common in humans and in Drosophila, and they are often associated with cullins And cullins are involved in degradation And let me explain what cullins do This is one type of cullin, where a BTB domain protein interacts with cullin-3, and then cullin-3 brings ubiquitin to a substrate, and that can then lead to the degradation, or the labeling and marking, of that substrate And the BTB domain protein, in this case, leads… acts as an adapter for that substrate And so, degradation is a really fast mechanism And this is used for many different processes, from development to signaling, cell growth And so this is an important process, the degradation of proteins, during development and differentiation And so the question then is, if we say germ cell-less would act as a cullin-3 substrate adapter, then we have to ask the question, does germ cell-less actually interact with cullin-3? And what is the substrate? And so the first question was answered by two ways And here, it was pretty wonderful, because we had identified this specific allele of germ cell-less which had a mutant smack in the BTB domain And that mutation interferes with the interaction with cullin Here, you can see in an IP and then a western blot… immunoprecipitation followed by western blot here is the interaction between germ cell-less and cullin-3, and then in the mutation it is inhibited And functionally, we can also show that this is causing a germ cell defect, because there is it behaves just like the mutant: no germ cells form, or very few germ cells form So, this suggests that germ cell-less is indeed adapting to cullin-3 So, that was a really important first step But of course, what we really wanted to know… we wanted to know how germ cell-less does it And so… getting to the substrate So, what is the substrate? And so, what we… we used this mutation, which interrupts the cullin-3 interaction, to be more easily getting towards the substrate, because the substrate will now not be degraded, but still interact with germ cell-less And so, Juhee identified torso as the substrate for germ cell-less, or, at that point, as a substrate for germ cell-less And so, again, we could show mutations in the germ cell-less domain of germ cell-less were important for the interaction with torso What is torso? Torso is a receptor tyrosine kinase And it was identified on the basis of its effect on patterning, somatic patterning of the embryo And it actually patterns the very anterior head part and the very terminal part And this receptor tyrosine kinase is specifically activated at the embryo’s front and back And you can see, at the back, it is activated right next to where the germ cells are So it… this receptor tyrosine kinase acts through the normal recept known receptor tyrosine kinase pathway And we can actually visualize the activity at the two ends by looking at a particular phosphorylation outcome, which is requiring a kinase activity And you can see how it’s active at these two ends And that then leads to transcription at the ends of the embryo And so, these are the regions right next… and these are the somatic signals, for example, that set aside the hindgut of the embryo right next to the germ cells So, we could now logically think, if torso was actually the primary target of germ cell-less,

and that was really needed for germ cell formation, then we would imagine if we don’t have germ cell-less, then torso is active and that would be inhibiting primordial germ cell formation However, if we now also deleted torso, germ cell formation should be restored And indeed, that’s the case So, here’s the single germ cell-less mutant with no or very few germ cells And here is the double mutant, which now has germ cells restored Interestingly, for those of you who are thinking about the first part about the formation, that tells us that germ cell-less is not directly required for that constriction, and that’s actually something we still have to figure out So, another conundrum I told you that germ cell-less was at the nuclear membrane, but receptor tyrosine kinases are at the cell membrane So, that’s a conundrum But somehow that conundrum was kind of… a little clearer to us when we realized that there was a nuclear localization signal and myristoylation domain The nuclear localization signal would get germ cell-less into the nucleus, but the myristoylation signal could also potentially get it to the membrane And I won’t go through all the mutants, but what I will tell you is that Juhee found that germ cell-less and torso do actually meet at some point And that is when the nuclear envelope breaks down during mitosis And now, torso and germ cell-less are at the cell membrane together And remember that germ cell formation occurs during a mitosis So, what we have learned, now, is that germ cell-less is a ubiquitin lig… part of the ubiquitin ligase complex, and is an adapter protein which uses torso as its substrate And so germ cell-less leads to the degradation of torso And then, unrestricted torso can interfere with PGC formation However, we do also know that that is MAP kinase- and transcription-independent And so now we’re trying to find out, what is the mechanism of this action of this receptor tyrosine kinase for germ cell formation? And so, on the other hand, what this also reveals to us is the struggle between somatic and germline signaling, where we have in the soma, we have this whole MAP kinase pathway, which leads to transcriptional activation And then, on the germline part, we’re degrading the receptor tyrosine kinase so that that signaling pathway cannot occur This is not the only place where soma/germline distinctions require degradation Indeed, in C elegans it’s been shown that another, cullin-2… linking to another degradation process, leads to the degradation of the somatic proteins, maternal proteins, which are in the soma, or in the future soma They are degraded But they’re not degraded in the germ cells So, here’s the distinction where the… the degradation machinery actually acts in the somatic tissues to get rid of the maternal proteins And in the absence of this adapter protein, Zif1, now these maternal proteins persist in the somatic tissues So, I’ve told you about germ cell-less, which is one mechanism, by just degrading the maternal products, which could interfere with germ cell development, because they promote somatic signaling And now I’m going to talk about another mechanism, which is, you could say, almost more brute force, and that is to turn off all transcription So, no new transcription So, then, obviously, anything which could promote somatic development will be blocked And so, here I’m showing you one of these RNAs that I talked about earlier, that are absolutely essential for the cellularization of the soma of the embryo This is a newly transcribed RNA, which is made by the embryo And if we go closer in, you can see how the RNA is transcribed in the soma, but absolutely not in the germ cells And so, what we know… there’s a big distinction at that time between those two cells So, what we’re seeing here, in red, is the activity of polymerase II at this stage

And what you can see is that polymerase II is active in making transcripts in the soma, but it is not present… and you can see those empty nuclei, here it is not present in the germline So, how is this achieved? It is achieved by a gene called polar granule component, because it was initially found as an RNA which localized to the granules And so, in PGC mutants… and again, keep in mind these are RNAs which are made by the mother and deposited in the egg So, in a PGC mutant… this means where PGC… maternal PGC RNA is lacking, what you can see here is you can now see the active polymerase also in the germ cell nuclei So, the cells form in this case, but the nuclei are now transcribing… can transcribe And you can see also, here, where now slam RNA is transcribed in these nuclei And what happens is that the germ cells do not develop well, and they do express somatic genes, and sometimes even participate in some of the somatic tissues So, transcription is obviously an involved process, and this is just a brief summary But it is kind of curious that the step that is regulated is not the initiation of transcription, or the initial process of transcription, but it is really elongation that is inhibited in many cases of transcriptional silencing And this is… can lead, then, to pausing of transcription And in this case, the two kinases which deposit this particular phosphorylation site signal onto the C-terminal tail of Pol II, of the polymerase, absolutely required for elongation, is blocked The access of these kinases is blocked by PGC, and so they cannot deposit this activating signal And so, this is the critical stage And what is very interesting is it has been observed in other organisms Here is just an example from C elegans, where you see transcription in three of the nuclei of the early embryo but not in the fourth These three will give rise to somatic tissue This cell will continue to divide and eventually give rise to the germline tissue And again, here it is a different gene, called PIE-1, which is also somehow interfering with the phosphorylation of the serine-2 residues in the C-terminal tail of Pol II So, using an antibody to those residues — which I showed you, that was the red staining in the earlier slides — it’s now been shown that for a number of other organisms, like Xenopus and also in ascidians, that there’s… the same is occurring during early onset of germ cell specification So, this is a very crude mechanism, in a way, to defend the germ cells from developing according to any of the somatic signals And you can see also how this type of mechanism can work both in organisms where you have a germ plasm that’s inherited, or if there’s an inductive mechanism And so, what I showed you is that in the germ granules you have these effector RNAs And I told you about the germ cell program that requires these effector RNAs And I was telling you about two of these effector RNAs: PGC, which blocks transcription in the germ cells; and then germ cell-less, which degrades the specific protein which could affect somatic signaling in the germ cells And so, the take home from what I was telling you is that germ cells first form by an unconventional cell constriction mechanism in Drosophila That… and then that Drosophila has two mechanisms of avoiding somatic differentiation, and thereby protecting its totipotency And one is by degrading the somatic signaling pathway And the other one is by blocking transcription And degradation is a very fast mechanism to get rid of components And blocking transcription, of course, is a very efficient mechanism So, these are the two ways of how the germ cells counteract somatic signals And it is interesting to think about how germ cells are specified and what are

the special mechanisms of germ cell specification And we do know that, in a way, blocking somatic differentiation is probably the most important part for the early germ cells, that they have to do to maintain their totipotent fate And with that, I want to thank the people in my lab: Juhee Pae, who did much of the work that I was telling you about, about the biochemistry and the function of germ cell-less; Ryan, who told us how the germ cells actually form in flies; Alexey Arkov identified the germ cell-less alleles in our lab; and we had a great collaboration with Michele Pagano and Antonio Marzio, who study degradation pathways, to develop this story And I want to thank our microscopy and proteomics facility, and other people in the lab who have contributed greatly to our understanding, especially of the transcriptional control And obviously, our funding bodies And thank you so much for your attention