Scientist Stories: Luciano Marraffini, Molecular Mechanism of CRISPR Immunity

Thank you, Steve, for that very nice introduction. Not sure I'm a hero of anything, but anyway, I thank you so much also for, for giving me this award. It's a tremendous honor, and I'm very happy also to share it with George. Just have to get off. All right. So I'm going to tell you today about the work that we do in the lab with CRISPR-Cas systems and how do we study this in the natural setup in bacteria. So to begin with, I put a series of different viruses just to show you that all organisms on Earth are infected by viruses and other pathogens, and in response to that, they need to develop immune systems. And my lab works with bacteria, and the main pathogen of bacteria is bacteriophages. These are viruses that infect bacteria and that are very deadly. They can be very deadly in a matter of hours. They can wipe out complete bacterial populations. And this year marks the 100-year anniversary of the discovery of phages by George Anderel. And since then, scientists have found many different resistant mechanisms that bacteria have to resist bacteriophage infection, but none of them is— but only one of them is actually something that is very close to an actual immune system. And that is encoded by CRISPR loci. CRISPR stands for Clustered Regularly Interspaced Short Palindromic Repeats., which is a very long acronym, but it represents very well the architecture of this loci. Here's a cartoon representation of CRISPR loci. And you will see here that in white we have a set of repeats. So that's the C and the R of the acronym. CRISPR is a cluster of repetitive sequences. The repeats are short and they are actually semi-palindromic. But the most important part of the acronym and of CRISPR system is that they're regularly interspaced. You see here that in between repeats, there are intervening sequences that we call spacers. And repeats are not adjacent to one another as in most repeat loci. So the spacer sequence, it was discovered first by informatics, by informatic analysis of CRISPR loci of bacteria, that the spacer sequences match the sequence of phages and plasmids that infect bacteria. And therefore it was proposed that this CRISPR loci will be involved in some sort of defense against these elements. And later on it was demonstrated experimentally by me and others that the— actually a bacterium that has a spacer sequence matching a phage or a plasmid is immune to infection by that phage or plasmid. In other words, the spacers provide sequence-specific immunity. And whereas the spacers dictate the targets of the CRISPR immune response, the Cas genes, which is a set of genes, CRISPR-associated genes, abbreviated Cas, that are usually flanking the cluster of repeats, these are the genes that are responsible for executing the immune response. So CRISPR immunity is adaptive. And if you have your culture of bacteria, you infect it with a bacteriophage of choice. I said that bacteriophages are very good at killing bacteria. In a matter of hours, you have your culture clear of bacteria. But not all bacteria die. There are always some bacteriophage-resistant mutants. And if the wild type— the CRISPR system of the original culture looked like this, when we isolate the bacteria that survive phage infection, the resistant mutants, and we look at the CRISPR array, we found an extension of that CRISPR array. A new, an extra repeat is introduced and a new spacer is also introduced. And on purpose here, I put the new spacer in red, the same color of the bacteriophage that was used to infect the cell because the new spacer has a sequence that is present in the bacteriophage that was used to infect the culture. And here, this is the first experimental demonstration of CRISPR immunity that shows exactly this. This is the wild-type strain. This is sensitivity to phage infection, very sensitive. This is the wild-type strain that doesn't have any spacer against the phage that was used in this experiment.. And the new resistant mutants where bacteria incorporated 2 or 1 new spacer against the phage, then the phage propagation is reduced by 6, by about 6 orders of magnitude. So it's very strong immunity that CRISPR provides. So how is that the CRISPR spacers provide immunity? They do that by being the template for a small antisense RNA that we call the CRISPR RNA. This CRISPR RNA can form a ribonucleoprotein complex with some of the Cas genes, more specifically with RNA-guided Cas nucleases. And that complex is able to scan the viral or plasmid DNA, and when a perfect match is found, then the viral DNA is cleaved, and then that's, that's the end of infection. The viral genome is destroyed. As Steve mentioned, as a postdoc in Eric Sontheimer's lab, I determined that the target of CRISPR immunity were DNA molecules and not RNA molecules. And at the time, CRISPR was mostly a bioinformatics field, and the bioinformaticians, they predicted that this was something like RNAi of bacteria. So the RNA guides will be targeting transcripts or RNA of viruses. But we did experiments that I don't have time to show you to determine that the that their DNA was the target of CRISPR immunity. And that was important for development of CRISPR technologies, also of course to understand the basic biology of CRISPR. And at the end of the paper, we made a note that it would be a very useful biotechnological application because it could lead to the destruction of specific sequences, DNA sequences, and that provided that CRISPR could be transported out of the bacterial system. And in the last couple of years, that simple comment that was just a comment at the time was possible— was actually— it became true that CRISPR could have very important biotechnological applications in what is called the CRISPR revolution. The work of many labs determined that CRISPR nucleases can be used for introducing site-specific mutations in many eukaryotic organisms. The CRISPR nucleus can be programmed by— simply programmed by CRISPR RNA to cut the gene of interest, and that double-stranded DNA cut that the CRISPR nucleus performs, that can be repaired in different ways to introduce mutations in eukaryotic organisms. I was a little bit part of that development of the technologies, mainly through collaboration with Feng Zhang at MIT, We were first— we were the first of two groups to— one of the two groups that first demonstrated that CRISPR can be used for making genome genetic editing of eukaryotic cells. We did also genome editing of bacteria with CRISPR, and then from then many other labs used it to make transgenic or genetic modified invertebrates such as C. elegans, Drosophila, fish, plants, mice, pigs, monkeys, and I stop there, and I hope also the community stops there too, because there's a possibility now of using CRISPR to modify the human germline, and that of course has many very serious ethical considerations that need to be thought about. But my lab did actually very little for the technology development per se. We always study the basic biology of CRISPR systems, and there's a lot of basic biology to cover because CRISPR systems are very diverse. There are about 50 different Cas gene families, and each bacterium or each CRISPR system has about from 5 to 10 Cas genes in the Cas locus, in the CRISPR locus. And depending on which Cas genes are present, The CRISPR systems can be classified into 5 different groups. And we've been very interested in studying Type 3 CRISPR systems, and in the next half of my talk, I will tell you what we learned from studying these rather complicated CRISPR systems. The simplest ones is the Type 2 system, which is the one that has Cas9 here. This is the one that is used for genome editing. So type 3 CRISPR systems are prevalent in staphylococci and in archaea. And they contain 9 of these Cas genes. Cas10 is the signature gene. And I studied a CRISPR system present in Staphylococcus epidermidis that has only 3 spacers. The first spacer matches a sequence on staphylococcal conjugated plasmids. And this is the spacer, or these are the experiments that we immunity against the conjugated plasmid was what I did as a postdoc and I told you about. When I started at Rockefeller, my own lab, I decided to look at the ability of this CRISPR to prevent phage infection, given that spacer 2 here matches a common staphylococcal phage. To tell— for you to understand the experiments that we did, I will introduce you very briefly to the life cycle of temperate phages. Temperate phages are— lambda is the most studied temperate phage. And temperate phages inject DNA into the cell, and after injection, they have two different possibilities, two different life cycles, either the lytic cycle or the lysogenic cycle. And in the lytic cycle, the phage circularizes in the cell and starts transcribing all the lytic genes, and that leads to the propagation of the phage and the killing of the host. In contrast, in the lysogenic life cycle, the phage becomes a provirus, a prophage. It integrates into the chromosome of bacteria and remains semi-dormant there. And at the molecular level, this choice of pathway is governed by a bidirectional promoter. If when the DNA is injected, there's some sort of a tug of war between these two promoters, and if the rightward promoter wins, then the lytic genes are expressed, and then that leads to the death of the host. If the leftward promoter wins, then the lysogen genes are expressed, and these genes promote the integration of the phage into the chromosome. And very importantly for this talk, The rightward promoter in the provirus is completely silenced because you probably will appreciate that the phage doesn't want, in a way, to express the lytic genes that will lead to the death of the host that is harboring it in its chromosome. So this is the staphylococcal phage, a temperate phage, phiNm1. This is the bidirectional promoter, lysogeny genes, lytic genes.. And the CRISPR system targets this region of the lytic area of the genome. This is the CRISPR RNA sequence. This is the DNA target sequence. And the first we wanted to know was whether CRISPR was able to have an effect not only in the lytic propagation of the virus, but also on the lysogenic cycle. And the first experiment was to measure the ability of the phage to propagate in the presence or absence of this CRISPR system. Here's the cells, the phage propagation in the absence of CRISPR immunity, and here in the presence of this CRISPR RNA. And you can see the 6, about 6 orders of magnitude of reduction of the phage propagation, as I showed you in that earlier slide. But the surprise came when we looked at whether the CRISPR system was able to prevent the lysogenization cycle. And we here in this plot, we have the total cells that were infected with phage in white, and the cells that became lysogens, that were the virus integrated, are in blue. And we didn't see any difference on whether we had or didn't have CRISPR immunity. And that was surprising because the point of showing you these sequences was to make a note that the same sequence is present both in the provirus as well as in the virus that is actively killing the cell. So this was the first indication that this CRISPR system, the Type 2 CRISPR system, cares about not only the sequence, the target sequence, something more, more about, more than the target sequence was important, that the presence of a target sequence was important to provide CRISPR immunity since this type 3 system was able to kill this lytic phage but was not able to touch the integration of the prophage. So the key experiment, or the breakthrough experiment, came from the analysis of CRISPR escapers. We were doing these experiments with a CRISPR system targeting this side of the phage, and I told you that the propagation is reduced by 6 orders of magnitude, but it's never fully reduced. There are always some phages that can make plaques escape the CRISPR immune system. And the most common way of phages to escape CRISPR immunity is by making mutations in the target sequence. When that happens, the CRISPR RNA guide cannot recognize properly the target, and therefore the phage escapes CRISPR cleavage. We found some of those, but many of those, many of the escapers, they actually had an intact target. But they contain instead mutations in the promoter, in this leftward promoter here. The mutation was in the minus 10 element, and this was a mutation that I don't have time to show you the data, but this prevented this promoter from functioning. So there was no transcription across the target. So this was the first suggestion or evidence that that this type 3 systems require transcription across the target for immunity. And we did some other experiments actually to demonstrate that that I don't have time to show you. So what happens, the reason why this type 3 CRISPR system prevents or attacks the lytic cycle but not the lysogenic cycle is because when the virus is in the provirus state, then there's no transcription across this region of the virus and this CRISPR system requires request transcription for targeting. So that's the explanation of this data. And this is beneficial for bacteria because in a way it's allowing to keep the provirus, which in many instances provirus have co-evolved with bacteria and they confer some benefits for the host. So to really show that— so those experiments determined that CRISPR immunity requires transcription. Type 3 CRISPR immunity requires transcription. And we wanted to know whether DNA cleavage by Type 3 systems also required transcription. And so we did the— we turned to biochemistry. We purified the nucleus of these RNA-guided nucleus of these systems, which is a complex actually of 5 different subunits: Cas10 and Csm2, 3, 4, and 5.. And then we did the experiments using a target DNA, which are two oligos annealed to each other, but we only see target cleavage when we use RNA polymerase in the test tube and when we add RNTPs. Here, if we generate these substrates that we call them elongation complexes that are stalled, there's no actual elongation without RNTPs. We incubate that with CRISPR, with the Cas10 complex, there's no cleavage. The moment we add our NTPs and polymerase transcribes across the target, then we see cleavage. And we found out that the active site of the DNA's active site of the Cas10 complex was in the Cas10 subunit itself, in something we call the PAM domain. A mutant in this domain does not do any cleavage. And then also something that was demonstrated by the Turns group in Georgia very early on is that the Cas10 complexes also can cleave RNA. If you incubate that Cas10 complex with a single-stranded RNA substrate that is complementary to the CRISPR RNA, that substrate gets cleaved. And we corroborated this in our studies with our own complex. And the active site for The RNA cleavage is in CSM3. A mutant in the active site does not do any RNA cleavage. So I'm about to finish. So these— the model that we have is that type 3 CRISPR immunity relies— the molecular mechanism relies on the co-transcriptional CRISPR RNA-guided cleavage of a target DNA and its transcripts.. And the complex cleaves not only the DNA but also cleaves the transcript. We did in vivo experiments to show that the cleavage of RNA is actually cleavage of the transcript of the target DNA. And of course the next question was, it's somewhat evident why DNA cleavage is required to eliminate the phage, but why is also required transcript cleavage? And I actually, we did the genetics experiments, there's only actually two RNases 3 and 6. And what we found is that when we mutate those RNases and we have a CRISPR system that is RNA minus, we see that immunity— the RNA activity is required for immunity against late targets. So these are 3 targets that are early expressed in the phage cycle, transcription cycle, and these are 3 targets that are expressed late.. And if we mutate the RNases, if you target the early genes, the cells grow very well after addition of phage. But if we target late genes, the cells die. So this late targeting requires the RNases. And the model that we have— we did other experiments that I don't have to show you— is that because of this transcription requirement, If it's an early gene that is the target, then transcription across the early gene happens very quickly after phage starts the lytic cycle and the DNA of the phage is cleared quickly. But if the target is in a late gene, then it can be 15, 20 minutes until the expression of the target. So the CRISPR system, the type 3 CRISPR system, doesn't touch the phage until about 20 minutes after infection.. And that is a problem because if you let a phage 20 minutes without any challenge, it will do many copies of itself and it will start to kill the cell. So our model at the moment is that the RNase activity of this CRISPR system degrades the transcript to prevent the completion of the phage lytic cycle and give time to the CRISPR system to clear all the DNA of the phage. So I mentioned that this is just an example of the interesting CRISPR biology. Just by looking at Type 3, we see this kind of interesting mechanisms of immunity, and we are very interested in looking at the rest of the biology of CRISPR system to see what they have to surprise us. With that, I'd like to thank the people that did the job, that I— the data that took the— collected the data I presented: Gregory with the transcription-dependent immunity, Pulami with the in vitro experiments, and Wenyan with the function of RNAases in the CRISPR system. Also the funding, Seattle Arita Allen, Hilma Hirschel, Simchaimer and Boris Wellcome Foundations, and NIH. And thank you again, ASBMB, for this tremendous award. I would be happy to take any questions if we have time.