TWiV 1361: Ancient Viral Sword and Shield

0:00This Week in Virology, the podcast about viruses, the kind that make you sick.

0:10From Microbe TV, this is TWIV, This Week in Virology, episode 1361, recorded on September 25th, 2026. I'm Vincent Racaniello, and you're listening to the podcast, All About Viruses. Joining me today from Western Massachusetts, Alan Dove. Good to be here. Here in Western Mass, it is currently broken to overcast clouds, 69 Fahrenheit, 21 C. And we're supposed to get part of this nor 'easter, but we're 50 miles from the coast, so it's not supposed to be as bad here as down on the shore. also joining me from Madison, New Jersey Brianne Barker hi, great to be here my weather is actually rather like Alan's we also are supposed to be getting this nor 'easter I think we will get a bit more rain and wind and in fact is already getting rather windy the weather has that little wind on it now and tomorrow today, tomorrow, Sunday Monday, Monday, it's supposed to rain.

1:24Great. Not that I don't like rain. It's just it makes a mess. And it could flood, and that's a problem. Especially when you're supposed to get three to five inches. Yeah. Here in Chelsea, it's 22C and cloudy. Well, anyway, let's go on with our business here. I'm like in a dark cave. I don't even see a few feet beyond the light. So it's good. I kind of disappear for a while. If you enjoy these programs, we'd love your support. You can go to microbe.tv slash contribute. A couple of news items. First of all, Bob Tesh has died. And he was 90 years old. A couple of obituaries to post. I heard this and I got a link, but it wasn't any of these links, but it's fine.

2:20Bob Tesh was a virologist at the University of Texas Medical Branch Galveston and a lot of work on arboviruses. In fact, he was on TWIV 583 along with Tom Kysak and Jim LaDuke. And, of course, Tom Kysak was on TWIV not too long ago. We were supposed to have C.J. Peters on that TWIV, but he hurt himself that day and couldn't make it, and he passed away recently also. Right. The other news item, this is the biggest one, of course, is that yesterday was the 18th anniversary of TWIV. It was first released on September 24th, 2008. TWIV is now old enough to vote. And I think it should get a vote. Yes, yes, absolutely.

3:10Well, at 21, when it's old enough to drink, then it'll be very interesting. That's right. But I think we do influence people's votes, probably, to some extent, by giving them good science. You should vote for people who are going to give you good science, because science is crucial for your health. If you think it's not, then you're not thinking properly. You're mistaken. So you should listen to us. Every week, we talk about science that is relevant to your health. Also, Kathy Spindler posted this in. So there's been a proposal out of the NIH, and the public can comment on that. There's a link.

3:53You can comment by October 13th. And I think this is a horrible idea, but it has to do with peer review. That is how grants are chosen or selected for payment once they're submitted to NIH. Now, what happens is they go to a study section of experts and they review it and they create a score. And the score determines largely whether it will be funded or not. And the score is passed on through the process all the way down to the person who wrote the grant. And now the proposal is no more numerical scores. They're just going to be classified into one of three categories. is most competitive, competitive, or not discussed.

4:43All discussed applications will fall into the either most competitive or competitive category based on the overall score. The scores will no longer be released to the principal investigators, to the applicant institution, to NIH program staff, ICO leadership, or NIH advisory councils. And there's also a Forbes article. You can read about it here. this is a way to hide what this administration wants to fund. You know, they have ideas about what science should be and what it shouldn't be. And this is not transparent whatsoever. So this is a bad idea, in my opinion. You can voice yours at the link that we'll have in the show notes.

5:26Yeah, I think that this showed up right after the sort of comment period ended from the previous terrible funding idea in the federal register that seemed to kind of get rather shut down or at least, you know, pretty well argued against in that point. This sort of seemed to be like a second attempt at this non-transparent way of funding what they want to fund instead of good science. Right. So by... By hiding the numerical scores, that would enable the NIH program officer or whoever is making the funding decision to pick any of the top 50 % of grants and say, oh, we'll fund that one. Yeah. Which may have been, you know, 49th percentile, whereas the 99th percentile grants didn't get funded.

6:25And I think – And that one might be because of other interests. Yes. Yes. For reasons that have nothing to do with science. I, this, this implementation is awful. I agree on that. There are elements of it that are, that are reminiscent of things that have been discussed in more intelligent ways. Yeah. Like this notion of identifying the top quarter or a third of grants and then, for example, you could do a lottery among those, which would be a fairer way than this top 2 % funding or whatever the pay lines that are now, which we all know are arbitrary. Because when you get above – when you get to less than 25%, you're really making finer and finer distinctions.

7:12And when you're trying to distinguish the top 5 % of grants, you really can't. Nobody can. So that is a problem that could be addressed with some portions of this, but the way this is implemented is just awful. The fact that the scores would not be disclosed to anybody outside of NIH leadership, I guess. Yeah, I was going to say, it's – I mean, not to the program staff, not to – there are so many people who it's not being released to that it just reeks of trying to not be transparent. Trying to hide information. I mean, in particular, why would you not tell the applicant what their score was so that they could try to improve their next grant?

7:56Right. That's just insane. You don't hide that information. No, but in the top 50%, there's a huge range of scores. Yeah. So you could really do better. And, you know, this is all so that they can make biased selections. It's all to hide the decisions that are actually – You can imagine that. Oh, Harvard, we're not going to give them a – Yeah. So – but also together with this, the administration is trying to – actually, it's the Office of Management of Budget is drafting an executive order for the president to sign, which would give him the final say on any funding decisions, which is also horrible.

8:33He should not have any say. The OMB should have no say. The OMB should write the checks and not be involved in anything. That's its role. And so another way that they're trying, I mean, the OMB wants it to actually to defund NIH completely. And even Bhattacharya, who we don't like as the head of NIH, has fought back against that. Well, without an NIH, you wouldn't have a job. Right. Anyway, we'll put a link in the show notes and you can go comment on that. But now we have some amazing science for you, as usual. These papers are so cool. A couple of people have written in about this one. The first set we're going to talk about.

9:16This is a couple of articles, two back-to-back articles in science. A non-contiguous code for RNA-guided DNA recognition at the origin of CRISPR-Cas. And authors on that one are two co-first authors, Peter Yoon and Kenneth Loy. Senior author is Jennifer Doudna. The team is from University of California, Berkeley, Lawrence Berkeley National Lab, and UC San Francisco, an all California group. And shall we just do title and authors for the second one because it's almost the same group? yeah so the second one is VIPR Viper RNA guided DNA recognition by non-contiguous geometric triplex formation we're going to explain what that is in a few minutes first author is also Peter Yoon co-first authors Trevor Doctor and Zeyuan Terry Zhang who are also on the first paper senior author once again Jennifer Doudna they are from the same places because it's basically the same group with some names moved around So the summary here is that you may know about CRISPR-Cas, which provides RNA-mediated immunity in archaea and bacteria, and which has been repurposed to do gene editing in eukaryotic cells.

10:39But, you know, how that arose isn't clear. So the authors in this paper went looking for an ancestor. They found it. They called it VIPER, Viral Interference Programmable Repeat System. They found it in bacteriophages, which consists of a protein which they call VIPER, which is ancestral to the earliest CRISPR-Cas proteins, and VIPER RNAs, which are the incredible part. They're made up of alternating GGYNN motifs, so those GG-perimidine slash NN-anybase. and they work totally differently from CRISPR-Cas. CRISPR-Cas pairs an RNA perfectly with a DNA and that guides the nuclease to cut the DNA, to cut both strands.

11:28This is kind of like an older brute force cousin where the protein moves in by this guide RNA that we'll tell you about. It's imperfectly base paired. There's no nuclease. It just wrestles the DNA and binds to it and stops transcription. And so they think this is the ancestor of CRISPR-Cas. But how they discovered it is just so cool. I mean, the whole idea here is that these are RNA-guided systems. And we see these everywhere, not just in CRISPR-Cas, but in other systems as well. and CRISPR-Casas are divided into two classes based on the effector proteins that are doing things. When we say effector, we mean the protein that's doing something.

12:15The class one systems are thought to have come around first. Their feature is these repeat-associated mysterious proteins or ramps. I loved that. I love that, yes. It's very good. I love it too. I've not heard it before. Have you heard it before? No. When I first was sort of looking and saw the ramp, I got all excited and I thought it was going to be related to innate immunity stuff that has AMPs. Yeah, yeah. But this is a pretty good acronym. I was not sad when I found out it was something else. Anyway, so these RAMP proteins oligomerize on CRISPR RNA and they form the RNA-guided effector complex.

12:55They're very conserved and on ancient features. And we don't know where they come from. So the RAMPs have little sequence similarity, so you can't just do homology searching to find related ancestors, for example. So they took advantage of the fact that there are structural features in RAMPs. They have an RNA recognition motif. They have a loop. They have a thumb domain. So they use these conserved features to do structural similarity searches against an alpha-fold database. If I could just back up for a second, the other type of CRISPR system is the one that you've probably heard about when you hear about gene editing, and that's the class two system, which uses a much more complex multiprotein machine to do the shutoff of the RNA.

13:50Go ahead. So this is the cool part here. First, they looked for related motifs in an alpha fold, which means that it's got the structures of these proteins there. And they got hits. Yes. And I will say the alpha fold, the other thing to point out about alpha fold is that alpha fold is basically an AI program. Yes, totally. So these are not actual structures that somebody made a crystal structure of. These are predicted structures coming from AI. And I know people, I know that there's a letter writer here who mentioned something about me having said something negative about AI. And AlphaFold is one of the things I really like about AI.

14:43No, actually, your paper was negative, right? Because it was not you. Yes. The paper was disturbing. So I've just been reading the history of AI. And this is a defining moment because people had tried to put information into computer systems, especially scientific information. It didn't work. And so then they came across this transformer idea with large language models where you just train computers to see what's going to be next. And so they had a contest. A consortium set up a contest to use this. And AlphaFold was the winner, basically. they came up with this and they blew everyone away that they could just give them all the known structures and train the large language LLM.

15:27What's the M again? Large language model. Model. Model on the protein structures. And now you give it as an amino acid sequence and it just says, what's the position of the next amino acid based on everything we know? And that's alpha fold and it bloody works. It's phenomenal. It predicts structures accurately with very, very high accuracy. Yeah. Highest. Yeah. So basically, even though they don't know the structures of any of these mysterious proteins, they can have AlphaFold predict them and they can find other mysterious proteins that AlphaFold predicts have the same structure. And so they can do this really cool search.

16:07So this search revealed a large and previously unreported family of RAMP homologues, 2 ,664, present in 0.07 % of prokaryotic genomes. And this is part of a system they named the Viral Interference Programmable Repeat or Viper. Now, I just want to point out here, these guys, now down in the lab, co-discovers of CRISPR, right, and modifying it. you get the Nobel prize with that with Charpentier, right? Was there a third person in there? Was Yang also? I think it was just the two of them. Two of them. Anyway, so they're thinking about the origins of CRISPR and they're thinking, how do we find the ancestor?

16:52And here it is. Well, we try searching the database, nothing comes up. And now let's, let's try this. So that's what science is now. So in the old days, people were looking at sequences and what are all these arrays of sequences repeated over and over again? They weren't looking for it. They stumbled on it. And that became CRISPR. But it happens less and less nowadays. We have huge databases of sequence. You just have to know the right question to ask. And here they had a question. What's the ancestor of CRISPR? And they went after it. Okay. So that's what led them to this. I mean, it's very smart.

17:28But on the other hand, they were also thinking about it all the time because that's what the lab does. Okay, so these Viper systems are different from CRISPR in two ways. So Vipers are minimal and have two protein-coding genes, the Viper gene, which is the RAMP homologue, and a Viper accessory protein called VAP. I thought it was a vape. I don't know. Then you can just think of a vaping Viper. But doesn't the vape have an E on the end? It should, yes. But we can call it either, I suppose. Like it's vapor. So the word vapor, water vapor. Right. If you take, I don't know. Okay. Okay, CRISPR-Cas has numerous cast genes, multiple ramps, and CRISPR arrays, the repeated sequences that originate from viruses or phage or plasmids.

18:22The VIPER systems don't have CRISPR arrays, but the gene is flanked by intergenic sequences that could encode non-coding RNAs. Okay. And then second, unlike CRISPR-Cas systems that occur mostly in cellular genomes, right, there are defenses against viruses. The VIPERs are found mostly in phages and sometimes in archaeal viruses. They find them in prokaryotic chromosomes, about a quarter of them, and plasmids, but mostly they're in the viruses. Yeah, and I wonder, you know, how many of those in the prokaryotic chromosome are remnants of provirus? Yeah, it's a good question. It would be interesting to look at those.

18:59But they're related, they say, and they suggest that VIPR represents the ancestral form and some of the data that we're going to generate. You know, VIPR proteins are simpler and they lack a lot of things. And so this absence of features that are present in CRISPR, right, is consistent with one being the ancestor of the other. And also these viper proteins are quite diverse. And so they say they've had a lot of time to diversify and spread. So they propose that the viper proteins are an ancestral ramp lineage that diverged before the CRISPR ramps. Okay, so the Vipr proteins contain G-loop and thumb domains that enable them to recognize guide RNAs, all right?

19:52And so this, I mean, this suggests that Vipr proteins have G-loops and thumb domains, which in CRISPR guide recognition of RNA. So it suggested to them that they would have the same function here. So they wanted to test this, a very simple experiment. They infect E. coli with a phage called SUSP1 that encodes a viper system, and they sequence the small RNAs that are made in the infected cell. Very simple experiment. And they find 400 nucleotide transcripts encoded within the intergenic region of the viper VAP operon. They call these viper RNAs, VR RNAs. I just want to note that it's a record. VR RNAs, right.

20:37But, you know, you could come to VR easily in any way, obviously. But I like that very much. It's cool. So, all right, VR RNAs. And these RNAs are fascinating, okay? You look at the sequence of them. They have 9 to 16 tandem repeats of Y, N, N, G. In the summary, I told you that. It's this conserved G, G, Y motif, right, where Y is pyrimidine, N is any base. and it alternates with NN, which means could be any two nucleotides. They had to use EVO2 to find this, to align them, because all other alignment methods failed because there's too much variability in this sequence, right? If you try and blast the sequence against the database, just using BLAST, which says what's a complementary thing, you can't get any hits because of the NNs.

21:29On the other hand, EVO2, it was trained. Remember, that's the nucleic acid language module that was trained to recognize what base comes next. Remember, it was used to make synthetic phages. And so they simply said, what is this? And it told them exactly what it was. Alignment-free approach, they call it. You know, what's the probability of the next base? And Evo2 got it because it was part of Evo2's training, right? We didn't know it, right? This is the other cool part. These things, these proteins and RNAs were in the database. They've been sitting there for years. Nobody knew how to find them.

22:08That's the key. That's the point. You have to know the right question to ask here. Okay, so you have this tandem repeat of these VR RNAs. Then there's a tail, which is an invariable three prime extension. And they folded this, and it looks like it adopts a structure, including a pseudonaut in this case. And so this parallels CRISPR RNAs, where you have a programmable spacer with a repeat handle in them, too. Okay, so they want to know if the similar architecture of these VR RNAs and CR RNAs plays a role in nucleic acid targeting, right? So they're hypothesizing that this G-G-Y position, which is conserved, they don't participate in base pairing, but rather the N-N bases, the N-N positions, specify the target.

23:02Right? So the idea is that this complex protein RNA is going to target a DNA and silence it. It's not going to cut it. Remember, there are no nucleases here like CRISPR. It's just going to silence it. So they think the N-Ns are important for that. Now, you may be wondering at this point, NN, two bases, right? If you do a search for a random two-base sequence, it will be everywhere. Yes. So how the heck this is going to work is still unclear, but right now they're just looking for, does it work like this? And we are going to get to how it works, and it's so cool.

23:45Yes, and you can imagine that these spacers, the G-G-Y spacers, are giving you some flexibility so that the NNs could base pair. Right. I mean, that's the whole thing. It's in the title, discontinuous. Yes. Right. That's the whole thing. So they test this model, right? They search the E. coli genome, the pan genome, all the E. coli strains using the inferred targets of these SUSP1 VR RNAs. So to do that, they concatenated the NNs with 0 to 3 intervening wildcard bases. So it would be NN, XNN, XXNN, or XXXNN, right, to model. Because they have to see adding which bases is going to make hits in the genome, right?

24:34So, you know, you could have zero bases, one, two, or three. Among the top matches that they got, most inferred one nucleotide skip rule targets that perfectly matched protein coding sequences. So if you put an extra base in and you do an array of these NNs, it will match protein coding sequences, right? And so the skip bases align to the third wobble position of the codons. And they could find high-quality match sites on both strands, and that's consistent with the fact that these VPR systems target both strands of the DNA. So they did experiments to confirm the validity of this one nucleotide skip rule.

25:18They reconstitute SUSP1. That's the phage viper RNP in vitro. So the RNP is the viper RNA and the protein mixed together. And they test its ability to bind double-stranded DNA targets designed with zero to three nucleotide skip lengths. So basically making synthetic targets. and the RNP bound all four substrates without a competitor, but in the presence of excess competitor DNA, only the one nucleotide skip substrate remained bound. All right, so this is all about, because there's gaps in the binding, the binding is not great unless you have a lot of them arrayed together. It's very different from CRISPR because CRISPR, you have a perfect match of what, 21 bases or something like that.

26:08Then in vivo, they corroborated this. They placed target sites with zero to three nucleotide skips upstream of a GFP, kind of reminiscent of two weeks ago, our Turin experiment. And they say specific binding will repress GFP transcription. And only the one nucleotide skip target sequence caused GFP repression. And if you delete VIPER or perturb the GGY motifs, you also abolish silencing. So it looks like their model is supported by these data. So the RNA, the VR RNAs dictate the target specificity. And they tested that. They tested different VR RNA, SUSP ones, the ones that they discovered. They confirmed that each will silence GFP expression when the target is upstream of the GFP reporter.

27:07And the specificity is cool. The specificity is reprogrammable. All you have to do is alter the NN, and you can program it to any gene, as long as the NNs match with the skip sequence. And are spaced appropriately. So you need that sequence, one base, that sequence, one base, and so forth. With the, yeah. And remember the G... G-G-Y. G-G-Y. Is bound to the protein. Right. So you can imagine that this protein has two bases sticking out and it's bound to the protein and there's two bases sticking out. And you have to add an extra base to match it up. Okay. So they could obviously predict natural targets.

27:51targets. And they did. And they say this reveals that it is actually, they say, an agent of interphage warfare. Yes. So, for example, in the SUSP1 Viper system group, 18 of the 20 vRNA targets converge on a different phage, a satellite phage that probably parasitizes SUSP1. and they so that's this is probably an offensive thing right because the sus-p is a lytic phage and they find some other examples of that they say some VR RNAs target other viper systems in the clay so that sometimes they can be aimed at one another in a defensive role so they conclude this is a cool viper thus operates as both a sword and shield in the interphage yes conflict it's pretty cool I like that you're being creative there Yeah, because the first sort of thing that you think about is, you know, earlier in the paper when they find these in phage genomes is sort of like, what the heck's the phage defending against?

28:57But in fact, the phage is actually in sort of this offense-defense battle with other phages. Yeah, it's not going to, the phage is not going to do anything to the host, right? Because it needs the host. It must be another phage, yeah. And that's interesting because if this is an ancestor, then it arose first in phages and then it was co-opted by cells, which we know happens all the time. And this neat, neat aspect that they point out about this discontinuous base matching strategy is that because you're matching two bases separated by one, two bases separated by one, and on down the line, that system will not target the locus that encodes it, which has G-G-Y, three bases, and G-G-Y.

29:49Yes, yes. So this constitutes a very simple and amazingly elegant way to distinguish self from non-self. Yeah, it's true. Like, wow. Very cool. And for the computer folks who are listening to this, this discontinuous matching is kind of like doing a regular expression search. Yeah, right. You know, find all the phone numbers in this document. Well, you could find one particular phone number by writing it out and searching for that string. But you could find all the phone numbers by saying, okay, everything that's got three digits separated by a character, then three more separated, and then four more should be a U.S.

30:32format code number or phone number. So you can do that kind of search. And this is a discontinuous search, which would give you that pattern. So next they did an experiment to see if the viper systems can defend against an invading phage. So they have E. coli with a Viper system, and they challenge it with—so they have a Viper system targeting a protein called Cro, which is a lytic regulator in a lytic lambda phage variant. And when they target crow with this viper RNA, it reduces placking efficiency by five orders of magnitude. So very, very efficient in a liquid culture. And a non-targeting guide, VR RNA, doesn't confer any protection.

31:31If they mutate any one of the 12 NN dinucleotides in the VR RNA, it greatly reduces protection. So there are 12 in this RNA. Second, they tested whether you could force a lysogen, a resident prophage, so an integrated copy of a virus, in this case lambda, which is silent. Can you force it into lysis by targeting a protein called C1, which is a regulator of lysogeny? So they're targeting the gene encoding that protein with a viper RNA. And of course, yes, it does trigger. So you need that protein made to keep lysogeny, to keep the genome silent. And if you target the DNA, yeah, it turns on the prophage.

32:21And so they conclude these are flexible systems for control over interphage conflict. So these systems are diverse. They're built around this tandem repeat code. And there are seven different types defined by the protein, by the architecture of the VR RNA, and by accessory genes. They vary in array architecture and additional gene function. Lots of details that we don't need to get into here. Some of these VIPER systems appear to be repurposed for host virus warfare. So types 3 and 5 are enriched in what are called defense islands of host chromosomes. And the vrRNAs target viral sequences, suggesting some kind of immune function here.

33:18And so these would be, again, vrRNAs in the host that are targeting the virus, which makes perfect sense. That's basically what CRISPR is doing, right? And these actually encode nucleases, unlike the other VIPER systems. And so the idea would be that they are degrading the virus. They didn't actually look at that, but that's a prediction. I'm sure we'll see more papers on that. Yeah. In that case, you'd almost have to have or find the phage that matches to be able to do that. And my impression from Jolene is that you can't always get the phage. Yeah, yeah. So they suggest that there's an ancient targeting function of the viper systems.

34:11This was transferred to prokaryotes, right? A virus carrying this infects a prokaryote and leaves it behind. I don't need that suitcase. Let's leave it behind there. It's left. The prokaryotes repurpose it to be antiphage, right? And they add new genes for, in this case, nuclease activity. And so they're saying that the ancestors of modern CRISPR may have followed a similar trajectory where it came from bacteria and then went to the host and then diversified. So that's the discovery of this really new system. Why is it important? because it's a punitive ancestor to CRISPR. It works without nucleases.

34:53It's ancient. And it doesn't cleave the virus virus one. It doesn't cleave the DNA, but rather just sticks to it. We're going to see how it sticks to it in the next paper and prevents it from being transcribed, essentially. Yeah. The other reason why it's important is that if we were to go back in time and read some of the early papers on CRISPR, we might say, why is this important? Basic biology of prokaryotic immune systems. But clearly this kind of DNA targeting system is something that we can potentially do stuff with. And so this perhaps is sort of the next CRISPR. And at the time, those early experiments were done.

35:49And in fact, there was a bit of an inversion there because the very first discovery of the CRISPR system was in an industrial lab. One of them, yeah. One of the very earliest. Well, they were a yogurt company looking at their yogurt and they found this system. They're like, hey, we found this weird system. And then it built from there. And as Brianne said, for years, this was just, oh, yeah, there's this weird thing bacteria do. And now it's used in every lab that does any kind of molecular biology, gene editing, and it's also the subject of FDA approvals for gene editing treatments. Well, and CRISPR is sort of the most famous example of this.

36:33But as I think about other cool nucleic acid related bacterial immune responses, the next most famous is probably restriction modification. We figured out some pretty important uses of that for biotechnology, too. So the idea that this could this sort of really cool basic research could become something interesting in a biotechnology way has a lot of precedent. Yeah. And also, for a more direct link here, this Viper system has some significant potential advantages over the current CRISPR-Cas gene editing system. It's a whole lot simpler. Very much. You've got two proteins involved, and you can do this discontinuous matching, so you can use very short sequences to target.

37:24This could end up being a very useful system for a lot of the things that CRISPR is currently used for. So, in fact, you only need one protein, which is small. That's right. Yes. Less than 200 amino acids and a single guide RNA, which is less than 100 nucleotides. Yep. And you can reprogram them so they're really minimal and could be, yeah, they suggest genome editing, DNA locus imaging, epigenetic modifications. I'm sure there are already patent applications in process. Of course. Of course. And they say, well, you know what? We found something, they say, a previously unidentified code for RNA-mediated sequence recognition.

38:07There's probably more out there to be found. So the way this works, and we're going to do the next manuscript, but just let me give you an overview. So you have these viper proteins. They oligomerize along the VR RNA by binding to these G-G-Y motifs. We don't know yet how, but we're going to find out. And they present the intervening NN bases for base pairing to the target. The viper protein binding distorts the GGY bases so that a single base on the target strand spans the three GGY bases. Right. Because then you get converts each GGYNN unit into an XNN target. So the NN is two bases, and then the GGY spans one base.

39:02So there's a lot of distortion going on. Right, which is that negative one thing that they were talking about in the paper. Yes, and that's part of how the DNA strand gets separated, as we'll see in the next paper. All right, so accompanying paper. You already heard the title, but I'll say it again. And viper RNA guided DNA recognition by non-contiguous geometric triplex formation. And I just want to mention, I don't think I mentioned this before, Peter Yoon, the first author on both papers, I looked him up and he was a graduate student in Doudna's lab. So grad students out there listening to the podcast, the bar has been raised.

39:48This guy just had two back-to-back science papers as a co-first author. I mean, the first paper is that you discover a whole new system. A whole new system and a whole new way of base matching. And you test it experimentally and you show fundamental aspects. We didn't answer all of the questions, by the way. There's one question here. Let me open this up again, which I forgot to go into. I wanted to. Hang on. No, it's not the one.

40:23the question is how are these vrna's assembled they're not it's not just an array like in crispr right um they actually go into this the mechanism of vrna acquisition is an open question because non-contiguous targeting cannot arise through wholesale spacer insertion. So in CRISPR, you know, if a virus comes into a bacteria, the bacteria chops up the genome and integrates pieces of it into the chromosome. And those are the CRISPR arrays that are then transcribed to make the CR RNAs that will target the next genome that comes in. So the origin of CRISPR is the virus genome or the plasmid DNA. Here, it can't be, right?

41:17Because it's got this G-G-Y in between every two bases. So there's some novel mechanism that is generating this, right? I can imagine some kind of redundant mechanism where some polymerase somewhere is reading and then inserting these G-G-Y. Well, we'll find out. That's the thing. That's one of the things we have to find out. It can be really interesting. You need to leave something for the next grad student. That's right. Oh, there's plenty, right? Yeah. Anyway, in this next paper, they wanted to look at the structure of this complex to say how VIPER actually does what it does in the first paper.

41:57So this is the mechanism. They solve 21 cryo-EM structures of various combinations of the protein alone, the protein plus the RNA, and plus the target. And these are real structures, not alpha-fold. Right. It's a real structure. I want to summarize the paper without going through each one individually. So first of all, the viper proteins form a right-handed helical filament along the VRRNA with about six subunits per turn. So each protein binds one repeat unit. Each protein binds one G-G-Y. Okay. Okay, adjacent proteins cooperatively displace the G-G-Y bases, so only the NN nucleotides are available for base pairing with DNA.

42:54And so when you look at this, actually it's a long tubular structure with the DNA in the middle, as you'll see in a moment. And probably what's happened is there's an initial viper protein bound to an RNA. This hits a target DNA, and then more protein RNAs assemble to form a long structure. I don't know how long it has to get, but it's longer than the initial one. And that's how you can trap the DNA on the inside. And that's also how you ensure that you don't go silencing everything that's got that NN, which, as we said, would be all over the genome, just a two-base sequence. So these protein RNA complexes could stick.

43:38One of them could stick somewhere on the genome that's not supposed to be shut down, but it'll just be that one sitting there. Yeah, and that's going to be too weak to stay there. Right. It's too weak to stay there until others come along and pair next to it. And then when you get up to some level, some quorum, they're going to say, oh, yeah, this is a legit match. now part of the what's what you see in the structures and part of the mechanism is the formation of what's called an r loop and so classically an r loop and this is what happens in crispr you have uh two strands you have the target dna strand and you have a guide rna and the other the other dna strand is displaced so this one and it's called r because in the electron microscope, when you look at this, it looks kind of like an R.

44:29But the VIPR-R loop is three strands. So you have the VIPR-RNA, which is the guide RNA, which is finding the target DNA. You have the target strand, which is base paired to the VR-RNA. And then the non-target strand DNA, the original complementary strand, is just off on its own. It's actually threaded through the center of the protein filament. So that gets put in there as the vipers are, viper RNA complexes are assembling and denaturing the DNA basically and storing that single strand in the center of this tube that's forming. And so this is the geometry that makes this work. So all three strands, it is very weird geometry, all three strands have the same helical pitch, about 66 Sangstrom's, but the VR RNA, 25 nucleotides per turn.

45:25The target DNA, 15 nucleotides per turn. Remember, because there's a mismatch between the VR RNA and the DNA, as we said before, and the non-target DNA strand is 15 nucleotides per turn. So that means the VR RNA has to stretch its path compared to the DNA strands to maintain the same pitch, it creates a strained structure. And why is that important? Why is a strained structure important? It doesn't unwind the DNA like CRISPR does by directly base pairing invasion. The CRISPR RNA comes in and together with the protein, it invades the duplex and base pairs perfectly. So that's just pure displacement, right?

46:11The viper filament has to grab the non-target strand and the geometric strain distorts and destabilizes the duplex. And then the target strand is handed off to pair with VR RNA. So the result is the duplex is peeled apart by this protein-driven mechanical force rather than by the base pairing and CRISPR, which would be thermodynamically favored, right? And the non-target strand is locked in place by direct protein contacts, not by base pairing. It makes it unique. So that's the triplex geometry that's happening here to facilitate this non-enzymatic. So the protein is just sticking to the DNA in this way.

46:57So it's very cool. And the structure shows all this, right? That's what you learn from the structures. And having that weird structure means that the DNA is probably not going to be usable or able to go through other sorts of processes. Certainly not going to be transcribed, right? Yeah, so this polymerized sequence of proteins has now gone like beads and sequestered one strand of the DNA, rendering it useless. Right. And the experiment they did in the first paper, They put a target for a viper RNA upstream of GFP, and it blocks GFP, presumably at the level of transcription. So I'm sure more work will go to understand that.

47:46So that's basically the summary of the structure paper, right? It's really lovely. The figures are beautiful. But essentially, you understand now the mechanism by which this is happening. So it's really quite an impressive set of discoveries, right? Yeah.

48:07And seminal because it's going to lead to a lot of other things. Yeah. So just a final comparison. So CRISPR-Cas9 cuts both DNA strands. The DNA is broken. The gene is disrupted. Disruptive mechanism. It's been repurposed to do gene editing, right? Viper has no nucleus activity except those ones perhaps in the cell chromosome. The filament blocks RNA polymerase from accessing the DNA, right? It's sequestered in the filament. The other strand is bound up to the viper RNA. And this geometric strain on the molecule that I discussed and the protein clamping prevent transcription, the DNA is intact.

48:48So presumably this could be reversible, right? I was just going to ask that. They didn't really talk about that. And actually, this is kind of reminiscent. There's a form of CRISPR called CRISPR interference. When you use a catalytically dead Cas9 that just sits on the DNA, it can also block transcription rather than cutting the DNA. So that you can manipulate gene expression and you can do it by turning this dead Cas9 on or off. so i like this very much i think on many levels it's cool and i think also that just the idea that you can if you have the right question you can find the answers in existing data right that's the beauty of having sequenced databases that are accessible to everybody yeah and that's the beauty of sort of science as we move throughout time is learning new ways to ask those questions and to address the sequence database that's out there.

49:52But it's just so cool to think about how much is out there. Yeah, nobody knew this was hiding in the corner. And I think it's really cool that this system evolved for interviral warfare and then got co-opted by some bacteria as a defense mechanism. And this is probably pretty close to the origin of immunity. Yeah. I mean, it's prokaryotes, and it appears to be an ancient system within prokaryotes. We're gotten way back in evolutionary time here. And this is a very simplified system that is capable of distinguishing self from non-self and stopping non-self. Got all the elements. All right, now we're going to really switch gears.

50:46Oh, boy. Yeah. It's still self versus non-self. It is, yes. And it's another - It's true. It's kind of an immune defense. It's another AMP. Yeah. It's Brianne's version of AMP. Now, the reason I picked this paper is because we recorded a twin this week, this week in neuroscience. And Jason Shepard came because he had just published a paper on ARC. So ARC is basically an endogenous retrovirus particle that is made in the brain, which he believes is important for neural activities. It's produced in neurons, and it can go from neuron to neuron. And this paper, he presents evidence that it's involved in dementias in some way.

51:35Okay. And it was, we were talking about the background and he, and I asked him, he looks at tau and beta amyloid. amyloid, yeah. And I asked him, so, you know, there's an increased rate of dementia is associated with herpes simplex infections, and the Shingrig's vaccine reduces that. I said, what's going on there? So he said, well, there's this paper just published recently by, it's October, December, 2025, by Rudy Tanzi. He showed that tau is induced by virus infection. Oh, we got to do that on Twitter. So this is Nature Neuroscience. Phosphorylated tau, T-A-U, exhibits antimicrobial activity capable of neutralizing herpes simplex virus 1 infectivity in human neurons.

52:32And the first author is William Imer, and the senior author, as Vincent said, is Rilf Tenzi. The group is from Mass General Institute for Neurodegenerative Disease in Charlestown, Mass. Harvard Medical School also in Charlestown. Ulm University and Ulm, Germany, and Wellesley College in Wellesley Hills, Massachusetts. Is that far from you? No, not far at all. Well, I mean, Ulm is pretty far from me. No, Wellesley, I mean. Yeah, Wellesley's not that far. Wellesley's closer to me than, they're still near Boston. It's probably about an hour. Yeah, Wellesley is sort of in the area of suburbs where a lot of folks from HMS And a lot of folks from Boston end up suddenly leaving.

53:21They're just outside Route 128, which is kind of the Boston Beltway, except it doesn't go all the way around because then there's the ocean there. Okay, so this is about, to some extent, about dementias and herpes simplex viruses. And these, you know, Alzheimer's disease involves deposition of amyloid beta protein and plaques and also hyperphosphorylation of the tau protein, which causes them to form neurofibrillary tangles. These are thought to be hallmarks of AD and actually play pathological roles in the pathogenesis of it. And in fact, to confirm a diagnosis of Alzheimer's, you need to show abundant levels of amyloid plaques and these neurofibrillary tangles or NFTs.

54:12You know, NFT means non-fungible. Yeah, no, it means neurofibrillary tangles. Let's keep it straight. So in this context, neurofibrillary tangles. There are six different isoforms of tau. And tau, these are generated by alternative splicings. And they have different microtubule binding domains. So tau is involved in binding to microtubules. And detachment of tau from microtubules is normally bound to microtubules. It's hyperphosphorylation and aggregation into NFTs. These are defining pathologies of what are called tauopathies, including AD and other dementias where the tau. So tauopathies. It's a tough word to pronounce.

55:00I always called it tauopathies. Tauopathies. Yes, that's better. What am I saying? To? No, tau. Tau. Also, I went to Towson State University, so maybe that's blasting me. In case anybody's not sure what microtubules are, that's part of the cell internal skeleton. Yeah. And microtubules are particularly important in neurons because neurons have to transport things a lot. And so they use that skeleton for a lot of transport. along the long-skinned neuron. And in fact, the tau protein helps the tubulin to polymerize on the microtubule. So really important for microtubule function in the neuron, right?

55:50So you have various types of tauopathy that are differentiated by the isoform of tau that is prone to pathological aggregation, and also factors that drive dissociation from microtubules, mutation, increased expression of the gene, increased kinase activity. So hyperphosphorylation of tau is a big one for detaching it from microtubules. Okay, so the reason this is in TWIV is because we think now that tau is part of an antimicrobial program. So we've talked before about small microbial binding peptides, right? The antimicrobial peptides. Or AMPs. AMPs that are important. And it's part of innate immunity, important for initial response to infection.

56:44These AMPs can bind to and neutralize invading pathogens and sometimes modulate inflammatory responses. And this group and others have shown that amyloid beta is an AMP, which can entrap and kill or neutralize bacterial, fungal, and viral infections, both in vitro and in vivo. So the antimicrobial protection hypothesis of Alzheimer's disease is based on these results. And what that says is that amyloid beta is not an aberrant product, but is rather a relevant antimicrobial peptide as part of the innate immune system in the brain, right? And so getting dementia is, I guess, an accident of this response.

57:31Getting dementia is just getting old and people are not evolutionary set up to survive that long. Well, but this is, I mean, this answers kind of a longstanding question in Alzheimer's disease. If this amyloid beta forms these plaques that damage the brain, why do we even have that protein? And the answer appears to be that's there and it has that ability to polymerize specifically to eliminate pathogens that get into the brain. because the brain is an immunologically privileged compartment, which means you do have the blood-brain barrier protecting it, so pathogens should not get into it. But if they do, they're kind of free of most of the immune system and they become very hard to eliminate.

58:19And so a system like this active in the brain would be there as a last resort, like form a plaque that'll kill the bacteria or kill the virus. So not everyone develops dementia, but the idea may be that, I mean, there are triggers. And maybe brain infections are one trigger. There may be other triggers. And so if you have multiple herpes simplex reactivations, that may predispose you because maybe it's inducing these antimicrobial peptides like amyloid beta and tau, as you'll see in this paper. And that's maybe why the shingles vaccine gives you some protection, because it reduces this response in the brain.

59:00Each time you have an infection or you have a reactivation, you're getting amyloid beta and hyperphosphorylated tau production. And I kind of almost imagine them as almost like little clumps that are trying to sequester the microbe. Right. And as time goes on, as you keep having this happen over and over again, you have more clump debris, more sequestration debris that can cause other problems. In fact, so tau has got a lot of similarities to amyloid beta, and so people have thought maybe tau has a similar role. And phosphorylation of tau is induced by amyloid beta amyloidosis, so it may be a tertiary role in defense, especially for pathogens that escape whatever amyloid beta is doing.

59:55In fact, herpes simplex is known to bypass these extracellular networks as it goes up by axonal transport along the microtubules, which are stabilized by tau. So in this paper, they're looking at the role of tau in HSV1 infections. We're going to see that tau phosphorylation increases in response to herpes simplex infection, and phosphorylated tau localizes to cellular regions of HSV1 infection. So remember, tau is intracellular, and amyloid beta is extracellular? Usually extracellular. Usually, yeah. So that's the idea that simplex will escape amyloid beta because it's going through, you know, the nerve is very long.

1:00:42You can come in at a peripheral site and write up the... Well, and it's sort of interesting that the simplex virus is using microtubules, which are stabilized by tau. And so you almost wonder, is this almost like an HSV-specific immune response, given the importance of herpes viruses? Yeah, ubiquity, right? Yeah. Everybody gets infected. Yeah. Could be. Yeah, a virus that we've had since before we were a species, it wouldn't be surprising to find a protein that we evolved to deal with that virus. Yeah. So you could look at that. You could look at phylogenetically. I think every mammal has tau probably, right?

1:01:23Yeah, I'm not sure. Somebody could answer that question. Yes. Yeah. All right, so look at the results. So first of all, they do some nice infections to test the antiviral properties. They pretreat a human neuronal cell with tau before infecting with HSV-1. So they pretreat for 24 hours. They have synthetic tau. And they have tau and phosphorylated tau. And then they look at the infected cells. They look at confocal microscopy and they do plaque assays as well. So if you pre-incubate with phosphorylated tau, you attenuate single cell infection and also plaque counts compared to untreated controls.

1:02:14The protection is concentration dependent. You get reduced plaque counts. You get smaller plaques. And non-phosphorylated tau has no effect on any of this. So it's quite clean. It's very nice. The phosphorylation is the key. Now, what is it about tau? How is it doing this? And they want to know if tau is interacting with the capsid. So the structure of herpes simplex virus is such. It's an envelope virus. Within the envelope is an icosahedral capsid, and inside of that is the viral double-stranded DNA. So they're assaying the binding of different isoforms of tau to the capsid, They isolate these capsids using a binding ELISA assay.

1:03:07All right. And they show that these tau isoforms will bind the capsid. They get increased binding affinity of phosphorylated tau with HSV capsids compared to the whole virions, right? Now, that makes sense because the whole envelope virion is not going to be inside the cell. It's going to lose the envelope at the plasma membrane. It's just a capsid that will travel, in the case of a neuron, all the way from the end, all the way on microtubules, all the way to the cell body. And they do this with multiple isoforms of tau, the concentration-dependent binding to the capsid. They have one form, 2N4R phosphorylated tau.

1:03:52It has the most microtubule binding repeats, has the highest binding affinity to the HSV1 capsid. It's almost like it's competing it off the microtubule. Yeah. Yeah. And they look for, I mean, that would be interesting to know what binds the capsid to microtubules and see if it overlaps with tau binding. They use antibodies to look for potential binding sites for tau on the capsids. They have antibodies to a variety. They have a capsid scaffolding protein. They have tegumind proteins, which are around the capsid, the capsid proteins themselves. And they incubate the capsids with these antibodies before they treat with tau.

1:04:37some of them decrease tau binding some of them have intermediate effects some of them don't bind at all they don't really provide a structural context for this like to deconstruct it for us they don't but that's the observation they also find that like oscillation of the capsid is important for binding So they conclude that the tau is preferentially targeting the viral capsid, so it's targeting the intracellular form of the capsid. All right, so these antimicrobial peptides often form net-like structures to trap microbes by targeting surface proteins, and this has been shown for amyloid beta. So they want to know if HSV capsids can seed tau fibro formation, right, neurofibrillary tangles.

1:05:39And they do this by imaging with transmission electron microscopy. They mix the capsid with tau, and then they look under the electron microscope. So if they have tau with purified capsids, they see what they call amorphous aggregated tau and fibril tau. And these agglutinate multiple capsids together. They also looked at infected neuronal cultures by EM, and they find tau associated with the surface of the capsid inside the nucleus of the cell, which is the final target for the herpes simplex virus. So they want to know the response of tau to herpes infection in a neuronal environment. So again, they infect these neuronal cultures and they're assaying for aggregation of tau.

1:06:36They stain for tau 24 hours after infection. They look by confocal microscopy. And just like we just saw that the virus drives aggregation of tau in these cells, they see what they call P-tau-positive dystrophic neurites and neuronal soma compared to uninfected cells. So basically confirming the presence of aggregated tau during infection. They then look at the aggregation in neuronal cultures infected using different methods to validate it. They have a dose-dependent increase in the insoluble-to-soluble p-tau ratio after HSV infection. And interesting, so no cytotoxic cell death at any HSV concentration, which suggests that the tau is protecting the cells, right?

1:07:34So after infection with HAV, they see a decrease in the levels of intracellular soluble tau. And in the absence of cell death, where's it going, right? So they're suggesting it's somehow being released or expelled. So they want to measure the movement of tau after infection. And they find P-tau in the media. So somehow it's getting from the cell and it's getting into the media. They use a microfluidic approach to isolate the infected cells from adjacent cells so that the virus can't spread from one cell to another. But they still see tau traveling from the infected cell to the uninfected cell in this microfluidic approach.

1:08:33So they suggest that an infected cell will make phosphorylated tau, and then somehow it gets to another cell and will protect it against infection. And they show this by doing infections and plaque assays under these conditions. So that would make sense, of course, that you could protect neighboring uninfected cells. Yeah, I kind of thought about it as, you know, a mediator of an antiviral-like state. Yeah. Yeah. Yeah. Okay, now they want to... So the question here is, is tau involved in cell-to-cell transfer as part of its antimicrobial activity? So they have a two-culture system where they have cells in a...

1:09:22So they're working in little wells in a plastic plate. And you can seed the bottom of the plate with cells. Then you can have a plastic insert, which you have cells seeded on insert. And the membrane prevents virus from traveling from the cells on the membrane into the bottom of the well. But tau protein can still go through. So this is a way of studying the migration of tau. So they can see, they label tau with azide so they can actually watch it. And they can see the virus, sorry, the tau traveling from the HSV infected cells. into the cells on the body. They can also localize it with antibodies to tau.

1:10:07They want to know what's promoting the increase in tau levels, phosphorylated tau levels in infected cells. So they have a kinase inhibitor that's the kinase responsible for phosphorylation of tau, and they see a dose-dependent increase in plaque counts when you inhibit the kinase that phosphorylates tau. So that suggests that that kinase, another showing that phosphorylation is important. Then they test whether interferon gamma signaling increases P-tau and the protection. So they add exogenous tau just to cells without infection. It doesn't elicit an inflammatory reaction. They measure 10 different pro-inflammatory cytokines.

1:10:55They can also reduce interferon gamma with antibodies. It doesn't make herpes simplex infection worse. But anti-interferon gamma antibody neutralizes the protective effects of synthetic tau pretreatment. So somehow interferon gamma is important for the antimicrobial action of p-tal, although on its own, it doesn't inhibit. What do you think of that, Brienne? Well, I... I'm confused. I'm confused. It is confusing. And this is a place where I'm sort of like, maybe there's a neuroimmunology piece that I'm missing here. But as I look at the inflammatory cytokines they've tested, they are not the ones I would have expected.

1:11:44And so I'm sort of like, why these cytokines, why not, say, interferon alpha? Or why not a type 1 interferon instead of a type 2 interferon? But maybe there's some neuroimmunology reason I'm not aware of. So I'm sort of confused about which cytokines they were testing here, and why not some others? Yeah, so the observation, just to repeat it. So you put antibodies to interferon gamma in these culture systems and it doesn't have an effect on herpes infection, but then the antibody to interferon neutralizes the protective effects of tau, pre-treatment. So maybe the interferon is helping cell-cell transfer?

1:12:33So maybe the interferon is helping make a pore? Oh, really? Okay. Maybe a gazdermin pore. Yeah, we just did an immune episode and we were talking about gazdermin pores very recently. So maybe it's something about how the transfer is happening, but I don't know. Okay. All right, last experiment, they want to see if microglia contribute to spread of tau, right? So you can imagine the microglia could endocytose aggregates and then, well, they might digest it, but maybe not. Maybe if they have indigestion, they bring it. If it's big and sticky, it's maybe hard to digest. Yeah, microglia are the resident immune cells in the brain or are some resident immune cells in the brain.

1:13:21So they do a culture. They have a neuronal astrocytic culture with microglia. They infect them with HSV-1. and then they look at tau and they find co-localization of phosphorylated tau and herpes simplex inside microglia cells. And I think the virus itself wouldn't necessarily infect the microglia. They're infecting the neurons and somehow the implication is the tau and the herpes is getting out and transferred. But they don't show that and they don't comment on it. So they say more work is done. So this is really an incomplete experiment. Yeah. I feel like doesn't HSV infect other immune cells?

1:14:10I don't know if it infects. Let's look it up. Doesn't it infect other myeloid cells?

1:14:27No. Yes. So I'm seeing that it can infect macrophages and dendritic cells. I don't know about microglia, though, right? Right. Well, but microglia are related to... Yeah. The thing is that that would be a trivial explanation for their findings, right? Right. Anyway. But the main point here is that that's the last experiment, that this tau, apparently Phosphorylated tau can protect against herpes simplex infection of neurons. And this is kind of similar to amyloid beta protecting against a lot of other pathogens, salmonella simplex, HHV6. So maybe tau is another AMP that has evolved to protect the brain against these infections.

1:15:18They notice that, they note that tau is not a typical. It says it doesn't comply with the attributes of typical AMP, as if the protein cares what it has to comply with. You're off code. Well, one thing I was thinking about that is, you know, something that's really challenging with these experiments is it looks like the culture system that they're using, the 3D REN cell VM cultures, which have neurons and as well as some microglia and some astrocytes and are pretty interesting kind of 3D cultures. Those seemed pretty new culture systems that may be somewhat more physiologically relevant than like a cell line.

1:16:02And so if you think about it, you have to do this experiment with human neurons. Yes. Because HSV is only going to infect human cells. And so you have to have a culture system for human neurons that is a pretty good culture system if you really want to test this in this kind of model. So this isn't something that, you know, could have been done super easily. I was thinking of different follow-up experiments and then I was like, oh wait, but I'd need another organism, herpes virus and another organism. Well, you could also use organoids, right? Right. Instead of just cells. And you could add immune cells to it.

1:16:43Yeah, and it looks like that's what these cultures are kind of organoid. Yeah, they did. And they added their astrocytes and microglia to it. But it shows that there's an involvement of tau in herpes infection. They say that maybe phosphorylated tau, we know that it destabilizes the microtubules, So that can also inhibit herpes infection because it needs microtubules to move from the periphery up to the cell nucleus. I liked your comment about it competing with the capsid for the microtubule. Yes. Yes. That's why it would be interesting to know what sites are on. That could be done. Yeah. So they say, you know, we consider phosphorylation of tau and spread of tau among neurons to be pathological.

1:17:27but maybe an antimicrobial feature. And if you have too much of that, then it becomes pathologic. But you think evolutionarily, you know, most of the evolution of humans occurred when they died in their 30s or 40s, right? And so there was no evolutionary pressure to live to... You had grandchildren by the time you were 30, so, you know, this... Yeah, so if you had buildup of this excess tau, it wasn't a big deal because - Right, right. As long as it protected you early, you're good. You didn't live long enough to have this buildup of excess tau. Yeah, and this could explain why Shingrix appears to be protective.

1:18:11Yes. Because you have then fewer reactivations of herpes virus, maybe, which prevents the system from getting triggered as much. And so maybe tau as well as amyloid beta would be, well, they're being considered as therapeutic targets for dementias. There's no question about that because obviously that makes sense, but the origin may not be as obscure as we thought it was. That's why I like this. And you know, a trial was done at Columbia, actually. It may be ongoing, of treating, taking patients with early onset Alzheimer's, not early, but beginning of Alzheimer's, not necessarily early onset, and then giving them acyclovir, which is an anti-herpes virus drug.

1:19:07And I mean, I was just at a meeting where they told me, they said the dose, it's like seven grams a day because it's not very efficiently getting into the CNS. and I don't think that had an effect on progression of dementia. Yeah, it would seem like if you're already at a point where you're presenting with symptoms. Maybe too late. Maybe too late. Yeah. Anyway, that's quite a mix of papers for you to digest, folks. I hope you made it to this one, which you should get a shingles vaccine. Yeah. Yeah, yeah. Maybe someday there will be a herpes vaccine, a herpes simplex vaccine, and this would be a good reason to get it, right?

1:19:54Sure. But I wonder what, there are other triggers that are not infectious, obviously. So, you know, MS, EBV is just accounting for a fraction of MS. There's other triggers as well. Let's do some, a couple of emails. Yeah, Brianne, you should take that first one. Okay. Blaine writes, hello, Dr. Racaniello et al. A few months ago during the pics section of the podcast, the team was discussing the individual experience of time due to Einstein's time dilation being in force even at low velocities or low gravitational gradients. You seem to question the real importance of time, and I have a remedy for that doubt, if there is any.

1:20:36Chemical kinetics. My physical chemistry professor quoted an unsung chemist as writing, if it were up to equilibrium, we would all be dead. Thank God for kinetics. Chemical reactions go at a pace, which means that time is of the essence. Indeed, without the miraculous enzymes that catalyze incredible chemistry in our cells at body temperature, animals would have the metabolism of trees and the life expectancy of mayflies. Reaction times may be minutely different for each of us in our own time zone, but it's close enough for all practical purposes. As a shout out to Brianne, I would paraphrase that chemist and say, if it were up to virologists, we would all be dead.

1:21:16Thank God for immunologists. Keep them in line, Brianne. Keep asking for evidence. Best regards, Blaine. Well, thank you, Blaine. And I like this a lot. But people, past listeners may also recall in the past, Vincent picked a book called Recursion by Blake Crouch, which talks a lot about individual experiences of time. And I actually just finished it last night. So this is striking me particularly that you are giving an interesting thing to think about with relation to what I was just reading and thinking on last night. Did you like Recursion? I did like recursion. It's pretty clever, I think.

1:21:59It was really good. Yeah, so chemical kinetics is, yeah, that's an enzyme. Right, that's true. Alan. Yeah, I may have to check out recursion. I read, what did I read of his? Dark matter. Yeah, I liked dark matter. Dark matter is good. And recursion was also really good. All right. Right, so Alan writes, during your September 9th, not me, During your September 9th, 2026 Office Hours podcast, at 49 minutes, you shared a story from a listener about a U.S. Army requirement for exercise immediately after vaccination. Here is a scientific justification for that practice and provides a link. It is not sadism to try and improve an immunological response to vaccination.

1:22:44And Alan is an MD in Las Cruces, New Mexico. He says, not New America. And I would say, why can't it be both? So, you know, if the army is doing it, they may be doing it for both the, you know, the reason that it improves immunological response and because, you know, it's pain. It is an interesting paper. Exercise after influenza or COVID vaccination increases serum antibody without an increase in side effects. That's cool. I want to read this paper. It's done in mice.

1:23:25how do the mice exercise in a wheel maybe treadmill or something they've got a mouse drill sergeant drop and give me 20 so it's interesting I think it could be both they call it a behavioral adjuvant so the lady on office hours she had been in the military and she said it happened to her and she said they were sadists. But it does help. Apparently, it does help immune response. Okay. Now, whether they're clinically relevant as Paul Offit would say, we don't know. Just because you increase antibodies doesn't mean that you're changing anything. So I don't know if they've done that study. You want to take another one, Alan?

1:24:12Sure. So Erzabet writes, Dear Dr. Twiv, hello. Thank you for being the most wonderful informative, caring people in the entire podcast universe. I appreciate every episode you produce. The passion and joy you have for science is inspiring. I miss Dixon's wonderful laugh. So say we all. My name is Erzabet. Gives a pronunciation. I am a science fiction author. My next novel is going to be about virologists disproving lab leak accusations. I have a BS in physics and love science, but I do not have a sense of smell, which leads me to my odd questions. First, what does a virology lab smell like? Does anything really stink or smell sweet?

1:24:58Let's take these one at a time. I'll start with mine. So, yeah, there are things that stink in many labs, including virology labs. You know, somebody's making a gel and they open a bottle of beta-mercaptoethanol and it's this intense. I would say it's a rotten egg smell, but it's so much fiercer than that. But that's like a one-off. Some reagents smell bad or weird or what have you. The smell I remember, I mean, yes, they're all chemical kinds of smells and there are lots. But the one I remember is when you autoclave agarose, when you microwave it, right, to pour a gel, and you smell that throughout the lab because it's hot and the vapors are going all over the place.

1:25:47And whenever someone is doing it, I can smell it. I can't give a description of that smell, but I would know it instantly if I walked into a room. It depends if you have just plain agarose or bacterial media. Yes. That was what I was going to say. was I'm sort of thinking about if your lab is doing any molecular work and you are making plates or making LB media, then there is the smell that we all know kind of smells like beer brewing. Yes, yes. But not. It's also got kind of a, I think that soup has gone bad. Yeah. You know, it's a weird. And then, of course, when you're culturing the bacteria, one smell that really sticks with me is walking into the warm room where it's 37C.

1:26:42You know, it's a hot day in there all the time. And you walk in and you're just hit with this smell of growing bacteria. Oh, yeah. And it's like, you know, yeah, this food has definitely gone bad. But depending on what the lab works on, certainly if it's a bacteriology lab and they're culturing Clostridia or what have you, you can get really nasty smells coming off. But in general, it's just a kind of a generally unpleasant, spoiled smell in the warm room. Yeah, the warm room is a really smelly place. The cell culture incubator has kind of a sweet smell, but not exactly appetizing. It's like a sickly sweet smell.

1:27:32Yeah, and it's not super strong unless you are using it a lot. Yeah. Yeah. So those are a few. Second question, what is something about lab work that you find is surprising to non-virologists but is commonplace for you? And I'll start with one. A lot of the movie depictions you see, there's always somebody pipetting, you know, colored liquids into something. Very, very few of the liquids we move around in a lab are colored. It's all these bottles of clear liquids lined up on shelves with handwritten labels. And you're pulling down various things and playing mixology with a, you know, micropipetter or with longer pipettes.

1:28:19and just it's like you're moving water from one place to another. People who walk into an actual lab after only seeing these stock photos may be confused about, gee, where are all the colorful fluids? There pretty much aren't any. Even bottles on the shelf are full of colorful fluids, right? Right, right. In their pictures. In the pictures. Whenever we have photographers come by to do photo shoots of us working in the lab, they ask me to get bottles of colorful liquid. And so the number of experiments we've taken photos of people doing with a bunch of things you would never use together all sort of around us is always funny.

1:28:59You know, it's interesting. If you ask AI to make a picture in a lab, they will have all colorful liquids. Yeah. Which is weird. Well, it's just remixing the internet, which thinks that it's all colorful fluids. Yeah. Do you have anything, Brianne? The only thing I could think about is the importance of an amount of time you spend on things like labeling and aliquoting and things like that. Although I do remember once my mom being at a lab where I worked for the first time ever, and she was very surprised that my workstation had a computer on it and not a microscope. So I have a story that illustrates.

1:29:43So basically people find it amazing that we never see what we're working on. Yes. Especially viruses. You know, most labs are not going to look at them in the EM, right? And that you can't see them. So once years ago, a CNN reporter came to the lab. There was a norovirus outbreak on a cruise ship. And she said, my producer wants you to film you inactivating norovirus. So I told the technician, spray bleach on the counter and wipe it down. And so she did that. And I said, that's inactivating norovirus. And she said, what? No, he wants you to see the virus breaking apart. I said, well, you can't see it.

1:30:27She was so disappointed. And she said, before she left, she said, well, I usually work the crime beat. This is not my thing. This is why it took like a million years of human history for people to realize that microbes were the causes of infectious diseases. Because you can't see them. And even with powerful microscopes, you can't see most of them until you get up to the electron microscope level. Yeah. All right, Elizabeth's third question. If there is one mistake to avoid when writing about virology, what would it be? Don't anthropomorphize viruses. Don't have viruses doing or thinking or wanting or evolving.

1:31:10Even evolving. Viruses don't evolve. Selection works to select out viruses with certain properties. So do not make viruses active. That's the worst thing you can do. That virologists are different from other people. You know, have like these terrible ulterior motives to doing virology. No, we either just think science is cool or actually want to help people. Yeah. Yeah. Yeah, those are good. That's true. That's true. finishing the letter if you answer this I'll be over the moon well there you go if any of you would like a free copy of any of my books let me know and it's yours after all you have given me it is the least I could do yours in vaccine positivity Erzabet cool Erzabet I'm currently on your website looking at what your novels are about yeah yeah look at this I write novels about smart women doing science in eerie places I'm going to buy one of your books but you don't have to send it to us I want to support you and I'm going to read it we can in the show notes when we include the letter we can just put a link to her site it's got gothic sci-fi books right yeah okay one more from John

1:32:31John writes your epitope 1359 answered a lot of my questions about why SSPE is so rare, takes so long to manifest, and yet can be correlated with prior measles infection. Am I correct in thinking that these hyperfusogenic variants with a FIX substitution can in a way be thought of as somatic mutations? They don't arise or at least don't gain any advantage until after the initial infection. I guess all of us who acquired our measles immunity the old-fashioned way, can be happy that such mutants aren't a part of the wild population, but why is that? Being hyperfusogenic is somehow incompatible with natural infectivity.

1:33:14Best regards from an overcast greater Braddock at 17C, where the awful humidity of late seems to have largely grounded out. So John, this is how I look at it, and this is what I asked Roberto, and he agreed. So in any population of measles, you have variants that are hyperfusogenic, But you don't need it because measles binds a receptor and effectively gets into cells, right? But in the brain, there are no receptors for measles virus. So in comes a population of measles viruses. And the ones that are hyperfusogenic are the ones that are going to get into cells and replicate and then move on to the next cell by fusing, right?

1:33:55So these are probably in all populations, but they don't matter until the virus goes into your olfactory bulb and gets into the brain. You agree with that, Brianne? Yeah, I think that that seems right. So the virus does not often get into the olfactory bulb and get into the brain. So that is a very good question that we didn't ask him. Is it that the virus often gets in there but doesn't get selected for hyperfusogenicity often, or does it not get there? I think it doesn't get in very much. Yeah. The one caveat I would say to the idea that you were just talking about is I don't know if the hyperfusogenic variants, are they transmitted at an equal rate?

1:34:47and so are you acquiring them at an equal rate or are they having to re-evolve after acquisition? Like is there a transmission bottleneck? Yeah, I don't know. I mean it could be that – I mean they must be present because when the virus gets into the brain, it can't enter cells to replicate. So it's got to be present to that initial population. But is it present in – In the inoculum. The inoculum or does it mutate after the inoculum? I would think it's present in the inoculum. It's present in everyone, and it just takes off in the brain. That's my view. Yeah, I think I'm kind of making an analogy to poliovirus, where the paralytic virus is there in every infection and only sometimes gets into the nervous system, right?

1:35:37Yeah, it could be that, yeah. But I'm hypothesizing I don't have data on that. Yeah, I could email you, Roberto. there. I'd like to know that. All right, let's do some pics of the week. Brianne, what do you have for us? I have an astronomy picture of the day from a couple of weeks ago that I enjoyed so much. This is from September 14th. And on September 14th, I was opening extra browser windows to look at this frequently. Ended up going and talking to my chemistry colleagues about it, all of whom also became very excited by it and look at it all the time. I know that this has come up before in some of our discussions of space-related things, but this pic is basically the periodic table of the elements with each element shown based on where that element originates.

1:36:32And so we can see hydrogen and helium originating in the Big Bang. We can see lithium coming from dying low-mass stars. And so on and on through every atom where that atom on the periodic table comes from. My chemistry colleagues really liked it because they thought it was a very nice way to see where the human made ones were right away without having to kind of think about it. You know, we see some are radioactive decay, some dying high mass stars, some comic ray collisions. But I just thought this was absolutely fascinating as we think about whatever element you are thinking of in whatever molecules you're looking at.

1:37:15This is where it came from. And so we are all space. Yes. We're all stardust. We are all stardust. Wow. All of us are just parts of us. Pretty much. Unless you've got one of those human made ones, you're pretty much a stardust. Most of those human-made ones would not be in you for very long. Exactly. So let's say potassium. All the potassium in me. Is from dying high-mass stars. Yep. And so bananas have a lot of potassium. They got it from a star, too. From the stars. From the dirt, right? Yeah. Yeah. Really? All of it? Yeah. Because you can't make potassium. Is that the thing? It's an element.

1:37:57Yeah. Nobody. Well, humans have made some elements. Well, right. But I guess they don't bother to make potassium. and ingest it. Right. I mean, it might be possible to make tiny quantities of some of these, but that's not going to— Yeah, I don't think anybody does. I mean, that's the idea of an element. It's never going to add up to any significance at all in the global perspective or the universal perspective, the amount of potassium that's there. Yes. This says this is the primary source on Earth. Yes. So the nitrogen that's in our atmosphere— Dying low mass stars. Yep. It's so cool. Yeah. It's very cool.

1:38:39I was so excited by this. I spent way too much time with it. When I saw this, I kind of had the same experience. I was like, wow. You know, science is really cool. Yeah. And everyone should love it. Not just people who do it. It's just so cool. I mean, look at that. Yeah. Helen, what do you have for us? I have a novel that I think Rich picked a while ago, and I'm picking it again because it's really, really good. This is Red Shirts by John Scalzi. And if you have ever watched an episode of Star Trek, if you have any contact with sci-fi at all, you really ought to read this. Um, so the concept behind the novel is as people who've watched Star Trek, no, um, uh, Spock, Kirk, and some random, you know, ensign in a red shirt beam down to the alien planet to see what's going on.

1:39:39Guess who's not going to make it back. Right. So there's this longstanding joke that the wearing a red shirt in Star Trek is, is a high mortality risk, which turns out statistically not to be quite true. But Scalzi took the concept and ran with it. It is so well done. So it's set aboard the – it's a Star Trek-like universe. It's not Star Trek fan fiction, but it's the Universal Union ship Intrepid goes out exploring across the galaxy. A new crew member gets assigned to the ship, and he's a brand new ensign. And he's very excited to be aboard the Intrepid, the flagship of the fleet, until he starts finding out that there's this thing called away missions that nobody wants to go on.

1:40:30Everybody hides when the science officer comes in to put together a team for an away mission. And he gradually learns that there's a high risk of death on these. And it turns out that things are not as they appear. and it's just it's really well done. The less you know about it going in the better. That's pretty cool. It's fun. It's on my list. It's a fairly short novel. Okay. It calls it a novel with three codas. The novel itself is almost a novella and then the three codas are the remainder of the book. So in real life the captain of a ship would never go down. Well, you would never have the captain and first officer going down together.

1:41:18Right. The captain would generally not go down, but yeah. But that's TV, right? Yes. You got to do something else. Okay. That's cool. So my pick is an art exhibition that I went to yesterday. So David Molesky painted the Ebola painting, which is behind me here at the incubator, which many of you may have seen. And he has an exhibition ongoing now at a thing called Spring Slash Break. It's down on 57th Street. And so he emailed and I was here. So I went and I was on the VIP list, so I didn't have to pay. It was cute. And so David is lately into painting bananas. OK. But but making them, you know, objects.

1:42:15He started with some Ebola themed bananas. So the paint I have a picture here for you guys to look at. He has three, four people in hazmat suits. And he was inspired by pictures of people in Ebola outbreaks, you know, carrying bodies, zipping them up into body bags. In fact, he has another painting of a banana being zipped up into a body bag. And then more recently, he reinterprets classical paintings with a banana in it. So he has one, you know, the myth of who stole fire from the gods? And then Prometheus was doomed to having his liver eaten out by a bird. So Prometheus is now a banana, right?

1:43:01and many others. But he's technically an outstanding artist, right? And so they're very interesting. So this exhibit was all of his banana-themed paintings, and I give you a link to his website where you can actually see them. But it was nice to see David again. He's really always thrilled that his painting is here. And I said, if you made a painting instead of Ebola virus, put the banana there, make it look like any bullet, then maybe I'd buy that one. It would be funny. I have a listener pick from Ann. Here's a gift link for lovely videos. I know some listeners enjoy these. Microscopic videos that show the invisible wonders around us.

1:43:46The winners of this year's Nikon Small World in Motion competition showcase the wiggly, feathery, undulating beauty of invisible life. That provides a New York Times link. Does it work? I'm just loading it now it looks like the initial video works and then it wants me to enter an email address to unlock the story someone unlocked this story for you but you have to log in to read it okay so it's not paywalled but it is email walled so it seems like it works for me these are some really cool they're videos not just photos right these are really cool Oh, actually, the top of the page is scrolling through a few of them, which are very cool, including my favorite from fourth grade, Volvox.

1:44:36Volvox? Yeah. Awesome. Well, thank you, Anne. That's TWIV 1361. You can find the show notes at microbe.tv slash TWIV. You can send comments, questions, picks of the week to TWIV at microbe.tv. and if you like what we do, we'd love your support, microbe.tv slash contribute. Now, I was walking back from the art exhibit yesterday. I'm waiting on the corner of 10th Avenue and something and I hear, hey, professor. And I turn around and he goes, nice podcast. He says, good to see you in the flesh. That was cool. Alan Dove is at alandove.com, turbidplack.com. Thank you, Alan. Thanks. It's always a pleasure.

1:45:29Breanne Barker is at Drew University Bioprof Barker on Blue Sky. Thank you, Breanne. Thanks. I learned a lot. I'm Vincent Racaniello. You can find me at microbe.tv. I'd like to thank the American Society for Virology and the American Society for Microbiology for their support of TWIV, Ronald Jenkins for the music, and Jolene Ramsey for the timestamps. You've been listening to This Week in Virology. Thanks for joining us. We'll be back next week. Another TWIV is viral.