It’s been a really good week for bivalves.
Why, just a week ago in the journal Science we found out that the shell hinge from the freshwater cockscomb mussel can be opened and closed 1,500,000 times in succession without suffering any fatigue damage (!), and we also learned the structural secret behind that. We’ve never had synthetic materials capable of this kind of feat, but now we’ve got a first demonstration of how to make them from researchers in Hefei, China.
So you’d think the mollusks would be satisfied with that for a little while. But not even a week later, the hard clam — or quahog (KOH-hog) in New England parlance — steps forward with a definitive “hold my beer”.
You’ve no doubt heard of CRISPR/Cas9, a system that bacteria use to snip the DNA of invading viruses to keep them at bay, and a system that has been adapted and customized all the way up to genome modification in humans to fight disease. Well, the quahog, it turns out, has its own DNA-editing system, an evolutionary cousin of CRISPR/Cas9. It’s not only biologically interesting to find this system in a higher organism (eukaryote), because we thought only bacteria could do this, but it should be a very pragmatic finding, too, because the DNA-cutting protein itself is half the size of Cas9, making it much easier to deliver to cells for therapeutic use.
This discovery occurred at the Broad Institute, a collaboration between Harvard and MIT, in the lab of Feng Zhang, who was instrumental in the original development of CRISPR/Cas9. We find out all about it in the June 28 issue of Nature.
The article itself is a rare Accelerated Article Preview, published as quickly as possible to get it out there, and as it happens, another group at MIT that includes Omar Abudayyeh and Jonathan Gootenberg published a very similar (though not yet peer-reviewed) finding in bioRxiv on June 14. We could be in for another protracted patent battle over the use of this system for biotechnology, like the one we still have over CRISPR/Cas9, but we’ll stick to the science here!
Without going into a long exposition of CRISPR/Cas9 — because you can find that in so many places now — we can say its job is to chop up invading DNA that gets into bacteria. The Cas9 protein is the actual tool that cuts DNA, but not just any old DNA; its specific instructions come from the CRISPR DNA array, which is a sort of Most Wanted List of snippets of invading DNA the bacterium has encountered and taken note of. RNA gets made from the CRISPR DNA array, and then that RNA physically attaches to Cas9 to guide it specifically to offending DNA from the Most Wanted List. You can’t just have a loose cannon DNA cutter running wild in the cell any more than you can have federal agents seizing people at random, so Cas9 has to stick to CRISPR’s instructions.
Here’s Cas9 cutting a piece of DNA on film!:
Less than two years ago, Feng Zhang’s group found that there are proteins called TnpBs out there in Bacteria World — with over a million different genes encoding them within known genomes — that look suspiciously like Cas9. And that just as Cas9 does, they can cut DNA and be guided by certain RNAs to know exactly where to go. But curiously, these bacterial TnpB proteins don’t do it for defense. They don’t have any Most Wanted List to refer to; that is, they aren’t associated with any CRISPR array.
Instead, genes encoding TnpB are always found next to just a single stretch of DNA that gets made into RNA but doesn’t continue on to be made into a protein. This is called the omega-RNA. Like the RNA that gets made from a CRISPR array, the omega-RNA is used to guide TnpB to go cut matching DNA. So TnpB doesn’t have a Most Wanted List, just a single Most Wanted Guy. But who is this Most Wanted Guy, and why does TnpB want him?
Comparison of the linear DNA arrangement of Cas12 (a close Cas9 relative) and TnpB. The “repeat/spacer” array next to Cas12 is the Most Wanted List, with the “spacer” portions encoding RNA that will guide Cas12 to foreign DNA to cut. TnpB is next to the “right end” (RE) of the transposon it is in, which contains one “omega-RNA” whose tail end serves as a similar guide. Note that Cas proteins like Cas12 are 1000+ amino acids (aa) long, while TnpB proteins are half that length or less
Well, it turns out that TnpB’s Most Wanted Guy isn’t foreign DNA at all, but rather some part of the bacterium’s own genome. Each TnpB is associated with a different omega-RNA that matches some part of its own genome, and we still don’t know exactly why.
But a big clue is that TnpBs are always found inside transposable elements (or transposons, for short), which are DNA sequences able to jump from place to place within a genome, or between different cells or even different species. In order for DNA to hop in and out of a genome, of course the DNA has to be cut, and in some shape or form, that is probably what TnpBs are involved with.
People like to pick on transposons as “selfish DNA”, which is true up to a point. But transposons can be extremely useful to their host organism. In fact, the reason jawed vertebrates like you and me can make a zillion different antibodies to protect ourselves is a DNA recombination mechanism that we evolved from transposons. For their part, these TnpBs are thought to be the evolutionary predecessors of Cas9 and its relatives, and surely they will keep evolving into other things.
OK, but bacteria are little weirdos anyway, right? Who cares what shenanigans are going on inside their genomes?
Well, don’t look now, but it seems TnpB has made its way into eukaryotic genomes, too, by horizontal gene transfer. TnpB-like genes in eukaryotes are called Fanzors. One lucky Fanzor recipient has been the quahog, and I’m focusing on that one here because it’s the “highest”, or most “animal-like” organism in which a Fanzor has been found and tested so far.
The Zhang group determined the 3-D structure of the Fanzor/omega-RNA-target DNA complex from the soil fungus Spizellomyces punctatus. Protein domains: REC = recognition (gray), WED = wedge (yellow), RuvC = endonuclease (cyan), NUC = nuclease (pink). ωRNA (omega-RNA) is colored in purple, DNA target strand (TS) is colored in red, and DNA non-target strand (NTS) is colored in blue.
Zhang’s group tested a few Fanzors from eukaryotes to find out first of all whether they could cut DNA, using omega-RNA as a guide to do it, as TnpBs do. But they also wanted to see if they could replace part of the omega-RNA to generate their own Most Wanted Guy and set that to work on human cells, with therapeutics in mind.
Most of the omega-RNA is there to grab onto the Fanzor protein and stabilize it (and you can see in the picture above how a lot of the purple part integrates into the complex), but a small and variable part dangling off the end is the “guide”, the sequence that matches the DNA that is to be cut. That’s the green part below:
Predicted secondary structure of the quahog omega-RNA. The guide portion in shown in green. “25 nt from stop” means that the omega-RNA starts 25 nucleotides after the Fanzor gene ends.
They tested four eukaryotic Fanzors and found that all four of them could indeed cut DNA using omega-RNA as a guide, though their patterns of doing so were a little different. They also found that each Fanzor had its own preference for a short sequence (the TAM, or target-adjacent motif, only about 4 nucleotides) that must be adjacent to the target DNA in order for cutting to happen. The TAM is a mini-anchor on the target DNA that the Fanzor protein has to grab onto to get situated and become active, and the TAM is always the same because it interacts directly with the constant Fanzor protein, not the variable guide RNA.
Then they swapped out the end of the omega-RNA (the guide part) with eight different pieces, to target eight different genes in human cells: B2M, CXCR4, VEGFA, CA2, KRAS, DYRK1A, HPRT1 and DMD. They found that three out of the four Fanzors (including the quahog’s!) were able to snip all eight genes and leave a small deletion behind. Looks like we have a whole pantry full of new gene editing tools for therapeutics.
The nice thing about Fanzors is that they are about half the size of Cas9, and they already work pretty well without having even been engineered. Plus there are lots of them, so we should have a whole lot more options for gene editing pretty soon. It’s possible that Fanzors, because they occur in eukaryotes, will be more compatible with eukaryotic systems, the most obvious example being therapeutic gene editing in human cells.
Zhang’s team also showed that Fanzors are well-behaved; that is, they only cut the DNA they’re supposed to. If there’s no guide RNA, there is no cutting. If the target DNA has no TAM sequence, there is no cutting. This good behavior is critical if Fanzors are to be used therapeutically in humans.
Buried within the Zhang paper is a list of eukaryotes in which they’d found at least one Fanzor gene. There are a lot of fungi and algae on that list, but the “higher” eukaryotes that contain Fanzor make up a pretty interesting little collection:
All of these organisms have a new tool for the next wave of evolution.
“Nature is amazing. There’s so much diversity,” Zhang told MIT News. “There are probably more RNA-programmable systems out there, and we’re continuing to explore and will hopefully discover more.”
Meanwhile, with all this horizontal transfer, the longer we higher organisms keep hanging out with bacteria…..
