VIDEO: René Ketting on The Role of Germ Granules in Small RNA Pathways

René Ketting (03:21):
I think the system that I’ll talk about offers a lot of opportunities for that. For instance, what you see here, so this is one of our two model systems, C. elegans, at least a fraction of it. What you see here, this structure right here that goes around the corner is the germline of C. elegans. You see in that germline, you see these green blobs or granules and these are the P granules. Of course, I would assume that they are very well known by all of you that are attending this, is I would say one of the best studies structures in this field, I would say. But these are P granules. Actually, I will talk about a novel type of germ granule that we identified not too long ago, which is actually not visible here because it does not contain this protein that we’ve labeled here and is present in this little structure right here, which are the sperm cells.

René Ketting (04:22):
This organ that you see here is a spermatheca that stores the sperm and whenever an oocyte is… An oocyte is pressed through the spermatheca, it’s fertilized by one of the sperm, and of course, oogenesis starts. I’ll talk about a dedicated condensate in these sperm cells about the biological function. That will be the first half of the talk. And then in the second half, and that is largely published, the second half I would go into efforts of us to study these novel granules at a more biochemical level. But yeah, this is still early days and I hope maybe to also get some good suggestions from your side. Why does this keep happening? Okay. Do it like this.

René Ketting (05:16):
Okay. So just briefly, these small RNA pathways, they are characterized by proteins that you see here revolving. This is a 3D structure from the lab of Dinshaw Patel. This is an Argonaute protein with a small RNA bound. So the red stuff is the small RNA. Really, it’s this combination of the Argonaute protein with a small RNA that makes the biological relevance. The small RNA alone would be highly unstable. The Argonaute without a small RNA as far as we know now has hardly any function, but together they form a very potent gene regulatory module by which the small RNA recognizes in a sequence-specific manner a target and the Argonaute protein then triggers an effect. Such an effect could be translation, typically the inhibition of translation, but also RNA stability and it could even attract factors to chromatin to modify histone tails, for instance. I’m sorry, I don’t know why this happens.

René Ketting (06:23):
So, small RNA pathways are highly conserved. Therefore they’re found in all clades of life or all domains of life. Whenever they are there, they’re highly relevant for either viability or typically also the fertility of the animal. So in animals, there is this clade of PIWI proteins that is highly germ cell-specific. António briefly mentioned that. And also in C. elegans, C. elegans has a PIWI protein, but C. elegans also has its own evolved set of Argonaute proteins that are mostly, as far as we know now, either specifically expressed in the germline or at least highly enriched in the germline. It’s this class of proteins that I’ll be talking about. It’s this class of proteins that we are trying to use to really get the different functions of the RNAi pathways, but also of the condensates because of like here, it’s indicated there’s like 13 different ones and they show very interesting differences in their localization and we start to understand a little bit about their functions, their specific functions as well.

René Ketting (07:39):
So I want to talk a little bit about or a little bit I want to talk today about in inheritance of silencing, so of these small RNAs, and in particular through the male and just to show inheritance of silencing in C. elegans is heavily studied. It’s been proposed to play a role in many different processes. This is for instance a paper from the Murphy lab that has to do with inheritance of experiences of the parents encountering pathogens and then transmitting that information to offspring in the form of small RNAs. This is a paper from the lab of Oded Rechavi showing behavioral effects that can be sometimes translated across generations. But the problem with these and many studies is that the relevance, that there are effects but the molecular mechanisms that are behind it are really not well understood. And I think we really need to get to those molecular details before we can really be certain of how to interpret these studies.

René Ketting (08:56):
Now, this is a paper from the group of Andy Fire, one of the discoverers of RNAi, and they already noted very early on that RNAi can be inherited and it can actually be very effectively inherited through the male gametic line as it’s written here. This is remarkable because, I mean, the oocyte is, of course, big, has a lot of cytoplasm and a lot of material is inherited. Of course, we all know about maternal effects, proteins and RNA that come via the oocyte. The sperm typically doesn’t have that because a lot of the cytoplasm is kicked out of the spermatocytes during development. So you end up with these very small spermatids. A lot of the cytoplasm is kicked out into this residual body. This is not specific for C. elegans. This happens also in mammals and, I guess, virtually all systems that make sperm.

René Ketting (09:56):
One of the functions is assumed is to make the sperm as small as possible. So there’s a lot of the material is being kicked out of the sperm. And so, young PhDs students or now a postdoc in lab ask the question, is there a dedicated mechanism maybe that goes against this process of kicking stuff out of this developing spermatids so that you can actually inherit small RNAs via the male? This was initiated by studies that he did on one of these big clade of Argonaute proteins in C. elegans, the WAGO proteins. He studied WAGO-3. You see a very similar picture as I showed you in the beginning of the P granules. These are indeed P granules. These are all nicely overlapping with the P granule marker, PGL-1, but then you see the difference here. In this case, you see signal here in this region where the sperm is made and stored, whereas the PGL-1 marker is not expressed there.

René Ketting (11:00):
And so, we wondered, so how can WAGO-3 protein be… How is it specifically retained in that sperm? And so, when we looked at the process of spermatogenesis, so here you see a gonad of C. elegans that is actually doing spermatogenesis. I will not bore you with all the biological details there, but here’s spermatogenesis going on. This animal is expressing both PGL-1, which marks the P granules, as well as WAGO-3. You see here in the beginning, there is nice co-localizations between the two. And then here, you see the PGL-1 signal starts to separate away from the WAGO-3 signal. You got this distinct foci of WAGO-3 that end up in sperm. I would show you this diagram, this image that I showed you before. So this is the development of primary spermatocytes to mature sperm. Here, you have the stage where you have this residual body, where all this materials being kicked out, but the WAGO-3 granules clearly are retained within the developing sperm.

René Ketting (12:07):
And so, WAGO-3 is inherited, sorry, is loaded into sperm. So does it have an effect on inheritance? The short answer is yes, it affects inheritance of gene silencing via the sperm very effectively. I just want to show you one assay that we did that looks at this in an endogenous setting. So we are not triggering silencing here by double-stranded RNA, but this is all endogenous. And so, what you see here, this is a cross between two wild type animals, a male and a female or a hermaphrodite. Of course, the offspring is again wild type and is again fertile. The thing is we can take out many of the small RNA pathways and there are many defects, but typically the animals are still fertile. So here, the male is defective in these two pathways, the female also, the offspring of course also. There are problems, but it’s not so bad.

René Ketting (13:20):
The problem really starts when the parents do not have these small RNA pathways, but the offspring has. So in this case, the offspring needs to initiate the small RNA pathways. So it needs to initiate the pool of small RNAs that it’s using for silencing without any information if there is any from the parents. Apparently, there is a lot of information from the parents because if the parents don’t have this, the offspring don’t know what to do and they make small RNAs against all sorts of genes that actually are needed by the germ cells. And so, the animals develop into sterile animals. This is a phenotype we call mutator induced sterility. We use this to look at inheritance of endogenously required small RNAs.

René Ketting (14:09):
I’ll just show you these sort of diagrams every now and then, not too many. So this shows this sterility effect. There’s this blue bar here, 70% of sterility. Here, it just showed that WAGO-3 is specifically required in the male to prevent this sterility. So here in the WAGO-3 mutant male or WAGO-3 mutant father sires an offspring in this setting that is completely sterile. Whereas if the mother is WAGO-3 mutant, it doesn’t really care. So WAGO-3 is really… I mean, if this went above the genetics that… I mean, it’s a complicated setting. It shows that WAGO-3 is really important for the inheritance of small RNAs via the sperm.

René Ketting (14:52):
And so, the question we started to ask is how is WAGO-3 maintained in this mature sperm? We have seen these granules. The study was propelled forward by a finding that Jan did when he did an IP-mass spectrometry experiment, where we identified this PEI-1 protein, standing for paternal epigenetic inheritance. PEI-1 is a protein schematically like this. It has a BTB and BACK domain at its N-terminus. For the rest, it’s appears to be rather disordered. PEI-1 is highly specifically expressed just at the later stages of spermatogenesis. So here, again, you see a spermatogenesis arm of C. elegans going from left to right. So here you have early stages and here you have late stages. PEI-1 here is expressed only in these late stages and we haven’t seen any expression anywhere else. And it recruits WAGO-3 into these granules. So WAGO-3 and PEI-1 basically always co-localize. Whenever the PEI-1 is expressed, WAGO-3 is also expressed in the same granule.

René Ketting (16:08):
I told you, there’s this large family of Argonaute proteins. There’s about 13. A number of them are also expressed during spermatogenesis, such as WAGO-1 and ALG-3. They are quite like WAGO-3. WAGO-1 is actually highly homologous to WAGO-3, but these two proteins do not go into these granules. So they go into these residual bodies that you see here. So here, you’re indicated this is a residual body. This will be the sperm. So these granules are not only interesting because they are specifically going into sperm and recruit WAGO-3, but at the same time they’re also highly selective. For longer term studies, that’s what I would get to at the end. We want to understand how the selectivity is achieved in these condensates.

René Ketting (16:58):
Okay. So this is showing at another stage, or this is the budding spermatid stage, where you see here this residual body and here the spermatids. Just to show you a bit clearly, these granules are really specifically in sperm and not in these bodies. If you take away the PEI-1 protein, so WAGO-3 is still there, but PEI-1 is not there. This is the wild type that I’ve just shown you. In the mutants, WAGO-3 is basically completely lost into the residual body. So you see that WAGO-3 is not by default maintained there. It needs the PEI-1 protein. Presumably, it needs these granules. And the PEI-1 protein, just like WAGO-3, is required for inheritance of small RNA information also in genetic assays. That’s the one I just explained to you.

René Ketting (17:49):
I want to get a bit more mechanistic here in terms of how these granules’ RNA kept within the sperm. Why don’t they go into the residual bodies? For that, I need to briefly explain to you, introduce to you what is happening during spermatid formation. So during development, during this sperm meiotic division, this organelle forms. So this is called fibrous body-membranous organelle. It’s an organelle that derives from the Golgi. It has some membranes and it has some fibrous proteins that are getting in there. This organelle at this stage looks like this octopus. It has already been described years ago that indeed these organelles are transported selectively into the mature sperm and do not end up in this residual body. Later in the spermatids they have specific functions. Forget the rest, the right part of this graph. So the sorting of these FB-MOs, so this process right here that they’re not here, but they’re all in the sperm depends on myosin-6, which is in C. elegans called SPE-15.

René Ketting (19:08):
And so, we looked in SPE-15 mutants at how the PEI granules are behaving, and sure enough, we saw that in the wild type they are nicely out of the residual body. In these myosin mutants, they basically are all over the place as if they’re not sorted properly. So we started to think maybe they are attached to these FB-MOs in some way. One way to look at this, we collaborated with the microscopy facility at the EMBL doing CLEM. And so, here you see electron micrographs of early stages of spermatocytes and later stages all the way to single spermatids, labeled the different organelles that you can recognize here. So here, you may recognize this FB-MO organelle the same shape as here. If we overlay now to PEI-1 granules, they basically always are right on top or right next to these FB-MOs. And so, our thinking was and still is that these condensates have formed contacts with these FB-MOs, and because the FB-MOs are being transported into the mature sperm, the PEI-1 granules are also transported into mature sperm.

René Ketting (20:30):
So how are these condensates then, how may they be attached to these FB-MOs? I’ll show you some data that is we believe through this protein that we call PEI-2. So PEI-2 was identified in an IP-mass spec experiment. PEI-1 was highly enriched and PEI-2 was one of the top hits. PEI-2 has basically an identical expression pattern as PEI-1, only during spermatogenesis always in this granules, always in sperm and not in the residual bodies. If you now delete PEI-2, you still have granules. They’re fewer and we haven’t fully characterized their properties yet, but the striking thing that we observed is that they look like these myosin mutants. They’re basically all over the place. They don’t appear to get sorted. And so, we interpret this as that PEI-2 may be forming a link between the PEI granules and these FB-MOs. Sure enough, PEI-2 has the same phenotype as WAGO-3 and WAGO-1, so genetic incapability of paternal inheritance.

René Ketting (21:48):
So this is how we think about this schematic of these PEI granules. There’s these FB-MO membranes. We have the granules that are formed by PEI-1 and where also PEI-2 is present. They can selectively recruit WAGO-3, but they would not take up related Argonaute protein such as WAGO-1. And then, of course, the question also is how are these then connected to these membranes? We don’t have a full answer, but we have an indication as to what may be happening. For that, I want to show you the phenotype of a mutant, which is called SPE-10. SPE-10 is a palmitoyl transferase. So it transfers fatty acids onto proteins and it’s specifically found on these FB-MO membranes in sperm.

René Ketting (22:42):
SPE-10 mutants, these PEI granules, they start to form, but then they aggregate together in much larger structures that are no longer nice and round, and they end up in the residual bodies. So we think they are not attached anymore to the FB-MO. Indeed, we have evidence that PEI-2, so you see here western blot of PEI-2, this would be the unmodified size of PEI-2. This is a modified version that we see that we always detect of PEI-2, and this is basically… This high event totally depends on the presence of this palmitoyl transferase. So within the SPE-10 mutants, we don’t have this modified form, we have only the unmodified form. And so, we think that PEI-2 may be palmitoylated by this enzyme SPE-10 and that these fatty acid tails may provide interaction most with the membrane. Of course, this is no direct proof. Personally, I would think if such a link would be strong enough to act in such a way, but at least the genetic data seems to point in that direction.

René Ketting (24:06):
So this is for the published part. What is happening is that in naive germ cells, you have a lot of Argonaute proteins. One of them is WAGO-3. They are typically found in these P granules. I haven’t talked about this. People have also described other types of condensates that are typically very close to P granules. But in that respect, I would like to keep it simple. So they are in the P granules. During spermatogenesis, WAGO-3 is leaving the P granules and the P granules are also dissembling. We don’t know what is cause or consequence. It may simply be a disassembly of the P granules that makes WAGO-3 leave. At the same time, these PEI granules start to form and they form on the membranes of these FB-MOs. They take up WAGO-3, but not other proteins such as WAGO-1. And then this P-15 protein myosin-mediated transport makes sure that this complex is retained in sperm and is not discarded in the residual body.

René Ketting (25:19):
Then, of course, and then we get into a topic that we haven’t studied yet, but it’s also our prime interest, we are starting to look into that. So then you have a single spermatids that develops into a spermatozoon. Of course, that would fertilize an oocyte. To be active then, of course, the WAGO-3 protein would need to be released from these PEI granules. How this works, we don’t know, possibly simple dilution because of the big oocyte and the small sperm that could be effect, but we also are not excluding post-translational modifications to play a role in this process. And then the released WAGO-3 can act in the zygotes to do its thing.

René Ketting (26:06):
Okay. So, the questions that we are now trying to address and we are doing that not alone, we are also collaborating heavily on this project together with Lukas Stelzl here in Mainz is how do they form the PEI granules? What is the role of this BTB-BACK domain and the IDR? What do the different domains do, and how is the specificity achieved towards WAGO-3? And then in later stage, but I will not touch upon that, how do they then release WAGO-3 in the end in the oocyte? So, I’ve shown you this little diagram of the PEI-1 protein. It’s a pretty simple protein. BTB-BACK are two folded domains and a large IDR at the C-terminus, and a little IDR at the N-terminus.

René Ketting (27:05):
What we’ve done before, I didn’t go into details there, we have genetically deleted different domains of the PEI-1 protein and looked at the effects in vivo. What we saw is that… So this would be wild type, the left part here. The middle part here is when we delete the BTB and BACK domain, but the IDR is still there. What we observe is that the PEI granules, they’re much less intense. They’re typically very small, but they still contain WAGO-3. So WAGO-3 is still there. So we think that the BTB-BACK domains are driving the assembly of the granule, but are not involved in a WAGO-3 interaction. And the other way around, if you delete the IDR, the PEI granules still form. We see these, the purple granules, but it loses the propensity to recruit WAGO-3. So the BTB-BACK domain we think is a scaffold to polymerize maybe, whereas the IDR is not driving here the condensate, but seems at least not on its own, but it’s more recruiting the WAGO-3 protein.

René Ketting (28:28):
But this is still in vivo and we want to make the step in vitro studies to understand these condensate in a bit more detail to really be able to explain the observations here because of course, here, there’s a lot of factors that are still affecting these granules and every mutant that you make can have indirect effect. So in principle, a bit hard to interpret. So Diego, a postdoc in the lab, has been picking up this project and has been purifying PEI-1 and PEI-2 proteins, full length in first instance. He gets relatively nice proteins and tries to do phase separation assays. What we see is that if he only takes the PEI-1 protein, we get very little units formed. PEI-2 doesn’t do anything. It seems to be a very soluble protein. When we mix the two, we get larger condensates formed.

René Ketting (29:33):
If I just blow it up a bit for you, so we get these larger units that are not nicely rounds, right? I mean, you can clearly see that. I guess in the first guess that I would say, they’re maybe a bit more gel-like or so, but we don’t have material properties on this as yet. You may already see some sort of substructures here. This is something that Diego has seen. If he labels PEI-1 and PEI-2 with different dyes, we see that PEI-2 typically forms most of the granule and the PEI-1 proteins seems to form foci within these larger units.

René Ketting (30:17):
This is something that we have not… Well, we haven’t gone back yet in vivo to see whether substructures exist also in the PEI granule in C. elegans. That will be difficult also because the size bar here is 10 microns, and 10 microns is basically the size of the complete spermatid. So these structures here are clearly much larger than what we have in vivo, but it could still be an interesting… I think it’s an interesting lead for us to see whether this sort of formation of substructures, of sub-domains may apply to PEI granules in vivo as well, and of course, what’s the relevance of that is.

René Ketting (31:02):
Now, to the BTB-BACK domains, BTB and BACK domains have been described. And there’s this very nice paper from the lab of Tanja Mittag, she showed in yeast in this protein SPOP that the BTB and BACK domains form a polymer, where the BTB domains dimerize and the BACK domains can dimerize. So you can get this polymeric chains and different chains can form together to form larger condensates. So that is also applied to our proteins, and we are studying this now. This is what Diego is currently doing. He is doing a lot of different studies using either purified proteins, but also yeast two-hybrids to see which BTB domain could bind to which other domain and vice versa for the BACK domain.

René Ketting (32:02):
So far, just as a summary, without going into all the data points there, PEI-1 seems to be able to polymerize as I showed you for this case here. So the BTB domains can dimerize, as well as the BACK domains can dimerize. For PEI-2, it’s different. PEI-2 only seems to be able to dimerize by its BTB domains. So you get only very short units. I guess this would be consistent with what we see in the phase separation assays, where PEI-1 on its own can do something, but PEI-2 seems to be very soluble. And then testing all the combinations, it seems that PEI-2 can fit into this PEI-1 polymer chain very easily in many different combinations. Why this then drives the condensates to the extent that we see, we don’t know. Of course, this is something that Diego is now further following up on.

René Ketting (33:08):
One thing that we are trying to do there, also to go back in vivo and test, is this relevant at all for what we see in vivo? Now, using AlphaFold, we seem to be getting very nice predictions of the BTB:BTB interactions and can predict how PEI-2:PEI-2 interaction would look like and a PEI-1:PEI-2 interaction would look like. We are making mutations in the residues that appear to be involved. At least the yeast two-hybrids, these residues are involved, but we have not yet been able to test such mutants in vitro phase separation assays or in C. elegans. Then as a final part, and so this BTB-BACK domain story I think really forms the backbone of the PEI granules. It’s a polymerization domain, and PEI-1 and PEI-2 clearly co-evolved to be able to form mixed chain polymers in that sense.

René Ketting (34:19):
Then I want to briefly touch upon the specificity and what we have learned about how WAGO-3 could be recruited. I told you that the PEI-1 IDR is required, right? So if you don’t have the IDR on PEI-1, you lose the WAGO-3 into the residual body and WAGO-3 really does not accumulate in the PEI granules. Now, what about the side of WAGO-3, right? So WAGO-3 is put a larger part of folded protein indicated by this very simplistic ellipse and it has an N-terminal intrinsically disordered region, which is very protein-rich. So, one thing that we did is we took out the IDR, but at least in vivo, this creates a dominant sterile phenotype. Dominant sterile is not something that you can really work with because already the heterozygotes are dead or are sterile so we can’t really analyze it. And so, we are making inducible systems. A PhD student in the group is doing that and we can indeed now induce the expression of IDR-less WAGO-3. We induced that phenotype and we are going to study how WAGO-3 then behaves, but we are not far enough there to draw any conclusions.

René Ketting (35:45):
Evidence that whereby we think that the IDR of WAGO-3 is involved in recruitment to the PEI granules comes from a collaboration with Lukas Stelzl and his PhD student Kumar Gaurav. they’re doing dynamic simulations, and this is one where you see individual IDRs change of PEI-1. The simulations, in a course-grained simulation here, they do form condensates. So this is something that we may have expected, of course, but it was good to see. And then we proceeded to test does the WAGO-3 IDR like to be in these condensates, yes or no? How about the WAGO-1 IDR? The answer there was that clearly the WAGO-3 IDR likes to be in these condensates. Wheras the WAGO-1 IDR, let’s say, it doesn’t hate it, but it doesn’t like it as much as the WAGO-3 IDR. But this is still very vague term. This is very much work in process, but we think we may be onto something here that there is sequence characteristics within the IDR that make it go into the PEI granules for WAGO-3, but not for WAGO-1.

René Ketting (37:05):
You see this also in a simulation here where you see in green, the PEI-1 IDR forming condensates. In red, you see the IDR of WAGO-3 and you see it nicely joins these condensates. So we think that the WAGO-3 IDR has an important role in going to these granules. Actually, we think that the dominant sterility that we see is related to other condensates, where the IDR may also regulate where WAGO-3 goes, because what we think is that the WAGO-3 without the IDR is actually loaded with the wrong type of small RNAs. So it may be present in the wrong condensates and thereby getting the wrong small RNAs. So also for that, we really need to understand how the IDRs of the WAGO proteins are determining which condensates they are joining and which ones not.

René Ketting (38:01):
The final indication in this direction is a final set of experiments that I’d like to show you, we have seen that the IDRs of these WAGO proteins are processed. This paper here, we provided a little help to the lab of Helge Grosshans, who published a nice paper on this protein called DPF-3, which is a protease. It has an activity that cleaves two amino acids from the N-terminus provided that the second one is a proline or alanine. DPF-3 has very strong effects on WAGO-3 and also on WAGO-1. They could show that the IDR of WAGO-3 and WAGO-1 is actually processed by DPF-3.

René Ketting (38:51):
I want to briefly introduce you a second enzyme called APP-1, which is very similar to DPF-3, but it’s substrate specificity and its activity are slightly different as depicted here. This enzyme we identified in an inheritance paper, where we were looking at factors that were required for inheritance of RNAi. And so, both of these proteases, both DPF-3 and APP-1, they localized to PEI granules. So you see the PEI granules labeled by PEI-1, you see DPF-3 and here you see APP-1. Clearly, APP-1 is not fully recruited to this PEI granules, but that’s clear signal there. Genetically, I’m not going to show you the data. Genetically, both APP-1 and DPF-3 are required for inheritance of silencing via domain.

René Ketting (39:48):
So then we start to ask the question, so if the processing of N-terminal tail is important, we should start to look at WAGO-3 and other organelles, not with the GFP, not at the full N-terminus because it would block processing, but at more internal sites. And so, this is what Ida, a PhD student in the group, is doing. Here we have WAGO-3 with an internally tagged GFP. See, it localizes to the PEI granules. And then in absence of these two proteases, we completely lose the localization of WAGO-3 to these PEI granules. So processing of the IDR, shortening of the IDR greatly stimulates the participation of WAGO-3 in these granules. We currently try to understand how such trimming of the N-terminus could help explain this phenotype and how in vivo this type of processing may be used for specificity causes to determine what WAGO protein will go.

René Ketting (41:01):
So, finally, this is my last slide, how we think that the PEI granules are working is we have a backbone of BTB-BACK domain interactions of these two proteins called PEI-1 and PEI-2. The IDR of PEI-1, we think, forms a brush that allows interaction then with the IDR of WAGO-3, but only if the WAGO-3 IDR is properly processed and trimmed. So here you see a shorter version. A longer version does not like to interact with this IDR brush. At the same time, PEI-2 we think is palmitoylated and that provides an anchor to these FB-MOs. I didn’t really show you this, but we also have indications that this tethering to membrane actually brings post-translational modifications also to the PEI-1 protein. We just don’t know what these modifications are at the moment.

René Ketting (42:00):
And so, by trying to understand this one germ granule in much greater detail, and I’ve shown you what we are trying to do, we hope to use the small RNA pathways to really understand how condensates can be used to segregate out very specifically specific molecules and secure them in some place of the cell. I haven’t talked about this. This type of protein is conserved. You find it also in human genome. Not specifically, not very directly, but BTB-BACK domains followed by an IDR are protein that you clearly see, and so we think that this type of process may drive subcellular organization also in other systems than C. elegans spermatogenesis.

René Ketting (42:57):
And so, with that, I’ve come to the end and I would like to thank the people I have mentioned, Jan Schreier who has done the work. I mentioned Diego. Diego didn’t join us on this particular retreat. He’s supposed to be doing the in vitro work now. Ida and Liza are following up on the IDR biochemistry and genetics of WAGO-3. I’d like to highlight here our collaboration with Lukas Stelzl and Kumar Gaurav, which is leading to the identification of really interesting motifs that may or may not explain the observations that we are seeing in vivo. With that, I’d like to stop and take any questions or suggestions on how we can better understand our PEI granules. Thank you very much.

Jill Bouchard (43:54):
Thank you.

António Domingues (43:56):
Thank you. So there was a question in chat, René, from Shruti. So Shruti, would you mind asking the question?

Shruti Jha (44:03):
Yeah. Thanks for the talk. Really nice talk. I was just wondering, since as much as I understood these are RNA-binding protein and I was wondering if you have looked up about the… try to change the availability of the RNA or try to silence, remove the RNA that these proteins bind to and then try to see where they localized to and if their ability to form condensate is affected by presence or absence of these RNAs?

René Ketting (44:35):
Yeah. Sure. Sorry, I have a strong echo here. WAGO-3 is indeed an RNA-binding protein, right? It binds these small RNAs. Without the small RNAs, WAGO-3 is mostly very unstable. So it doesn’t accumulate at all to any appreciable level, and so it also doesn’t join the PEI granule. That said, the PEI granules themselves as judged by PEI-1 or PEI-2, they don’t rely on WAGO-3. So you can take out WAGO-3, but these foci that are formed by PEI-1 and PEI-2 that they are unaffected. So in that sense, the small RNAs don’t contribute to the condensates formation, which I guess is not very surprising. Whether there are larger RNAs in there at the moment, we just don’t know.

Shruti Jha (45:36):
If they’re unstable without small RNAs, then how can we say that they do not contribute in condensate formation? Does it not require to be stable in that sense to localize to the condensates? So you mean that the WAGO-3, even in absence of small RNA, they do participate in the condensate?

René Ketting (45:57):
No, because we can’t see the protein. So, I don’t know. I mean, the few proteins that are there, no, we just can’t see them because they’re just too few.

Shruti Jha (46:10):
Okay. Okay.

René Ketting (46:10):
That’s it. There is an RNA binding. So the IP-mass spec that we did on PEI-1 did identify protein with an RBM. It was very high up in the list. I think it was even most high up. So far, we just don’t know the function of the protein and whether it indeed binds RNA and if that contributes to PEI granules at all, so I don’t know.

Shruti Jha (46:37):
Okay. Thanks.

António Domingues (46:40):
I will also follow up with a question since we don’t have any more in chat. But I’m curious, you mentioned that there was a difference between the WAGO-1 and the WAGO-3 terminal, the IDR. Could you comment a little bit further? Because I believe the WAGO-3 was rich in prolines?

René Ketting (47:01):
They both are. They both have this very proline-rich sequence. The net charge, so if you look at the overall charge distribute, not the distribution, the overall charge of the WAGO-3 IDR is a bit different from the WAGO-1 IDR, and that difference gets stronger when you process like that IDR, when you make it shorter from the N-terminus. So it could be that that charge difference plays somewhat, but we have not looked at… We have not been able to identify stickers or any type of that sort of thing and how this interaction would take place. This is, I’m afraid, a bit too early in those stages. Yeah. So there are differences, but it’s nothing spectacular.

António Domingues (48:00):
I guess one experiment would be the reverse, so tag the tail of WAGO-1 and WAGO-3 and see what happens.

René Ketting (48:08):
Yeah. I mean, that’s something that we have considered, of course. Hybrid protein, I’m not the biggest fan of making such hybrids, but I think in this case it may actually be a good thing. Yeah. We just have not gotten to do that. Yeah.

António Domingues (48:35):
Jill, any more questions on your side?

Jill Bouchard (48:38):
I don’t know anybody at the kitchen table? Did I not see anybody? I mean, I can ask a question because I’m a big fan of BTB and BACK domain as I used to be in Tanya’s lab.

René Ketting (48:47):
Oh, I didn’t know.

Jill Bouchard (48:49):
I’ll hold mine. I see there’s Ashish with a hand up. Do you want to unmute yourself?

Ashish Bihani (48:56):
Yeah. So, one thing I wanted to ask was have you looked at the mobility of these protein maybe through FRAP recovery or something? Maybe that would show how these… Hello?

René Ketting (49:13):
Hello? Yeah. Yeah. Do you still hear me?

Ashish Bihani (49:20):
Yeah. Yeah. I can hear you.

René Ketting (49:21):
Oh, yes, we have done that. I mean, I haven’t shown that. So we have looked at FRAP of WAGO-3 and we have done that in the PEI granules, as well as in the P granules, right? So we did it in the body of the germline, as well as in the sperm. I don’t recall the exact recovery time, but at least the recovery in the P granules was significantly faster than in the PEI granules, which would be consistent with these PEI granules, which seemed to be doing transport of some sort, having a really a different substance maybe than the P granules, one being maybe more liquid and the other may be being more gel-like. I’m a bit cautious there because we just didn’t do any of these assays that I think that people advocate that you should do, and I think that are fair to do to really make statements on what these materials could look like, but WAGO-3 is definitely less mobile in the PEI granule compared to a P granule.

Ashish Bihani (50:29):
Right. Maybe the PTM subspecies could drive that kind of difference in mobility.

René Ketting (50:39):
Sorry, again? You were breaking up a bit.

Ashish Bihani (50:42):
Okay. Yeah, I was just saying that the PTM subspecies of these proteins could drive that kind of heterogeneity.

René Ketting (50:50):
Yeah.

Ashish Bihani (50:50):
Okay. Then another thing I wanted to ask is what about the molecular grammar? What kind of patches are increasing the preference of WAGO-3 and what are excluding that? For example, you have acidic patch or just polar residue patch that create that environment.

René Ketting (51:13):
Yeah. So we have played a little bit, but only a little bit with the IDR of PEI-1. We added a glycine and serine-rich region of PGL-1 and stuck it onto PEI-1, and that made WAGO-3 move quicker. So, FRAP times went down. But again, I don’t really know how to really interpret that other than when you change the IDR, you change its properties, which I think is not very surprising. So we have not been able to really make strong statements on some sort of combinations or certain spacings of amino acids.

René Ketting (52:02):
Within WAGO-3, really the only thing… I mean, there’s lots of serines and threonines, and so there’s ample opportunity for phosphorylation there.We have identified actually a kinase also in this whole process that seems to be acting there. So I think phosphorylation would definitely play a role. As António said, I mean, there’s tons of prolines. It’s very proline-rich IDR. That is really the first thing that you notice. When you open that sequence in WormBase, you’re like, “Oh, wow.” So that is really the one thing that stands out. I mean, we haven’t been able to find any other particular enrichments or depletions.

Ashish Bihani (52:47):
Okay. Yeah, these are actually very interesting observations that you told me about. Okay. Thank you.

Jill Bouchard (52:57):
Loving the discussion, guys. I guess I’ll ask mine now. So, I was really intrigued by the BTB-BACK interface and the PEI-1, 2, I don’t know. So I was interested that the BACK… Was it the BTB or the BACK that didn’t dimerize in PEI-2?

René Ketting (53:18):
In PEI-2, the BACK domain doesn’t dimerize. No.

Jill Bouchard (53:22):
Did you guys look bioinformatically and see why that is?

René Ketting (53:26):
So, we try to do that now with AlphaFold. Yeah. At the moment, we just don’t know. We just know that the BTB and the BACK domains of PEI-1, they seem to be pretty promiscuous. They can do it with other PEI-1s or with PEI-2. That can mix and match.

Jill Bouchard (53:50):
Are they structurally similar and that’s why?

René Ketting (53:52):
They’re very similar. I mean, if you look at the predictions, they’re very similar, but clearly not identical, right? I mean, the sequence is not identical. Also, if you look at the fold, at the predictive folds, these are not identical. And so, what I would like to do and what I hope that we can do is that if we can make nice monodispersed dimers by mutating some of these interfaces, we can really get structures of individual dimer interfaces and then maybe are in a better position to explain why and how exactly. But I think in the end, Kds and Kons and Koffs of the BACK-BACK domain interactions and the BTB-BTB domain interactions will determine to a large extent the content of these granules, so how much PEI-1 and PEI-2 is going in and how much they like to form like homotypic clusters or heterotypic. Yeah. It’s the sort of information that we currently still just don’t have.

Jill Bouchard (54:57):
Gotcha. Well, it’s a great example of how sometimes structured domains drive droplets more than IDRs too.

René Ketting (55:03):
I think so. Yeah. When we’re doing the first experiments, I was pretty simplistically thinking, yeah, we take out the IDR and the granule’s gone, but it was really around so it’s-

Jill Bouchard (55:15):
Yeah, but it turns out polymerization matters a lot too.

René Ketting (55:18):
Yeah, definitely.

Jill Bouchard (55:19):
Any other questions? I guess not, but it’s been a nice discussion with you, René.

René Ketting (55:27):
Thank you.

Jill Bouchard (55:31):
António, do you have anything else or should I just close it down?

António Domingues (55:34):
Nope, nothing else on this side except to thank René again and great to see this work again.

Jill Bouchard (55:42):
Yes, 100%.

René Ketting (55:42):
My pleasure.

Jill Bouchard (55:42):
Thank you so much for teaching us about the RNA and the granules and all the things. I know I learned a lot. So, thank you everyone else for joining on Zoom and at the table and at the other table. Great to see all the faces. We will be here in about a month again, so feel free to come back. Thanks again, René. Have a good one.

René Ketting (56:03):
All right. Thank you. Bye-bye.