Tuesday, September 29, 2015

More on CRISPRs


We have examined CRISPRs for the past few years since their introduction. Initially we had a CRISPR with a Cas9 molecule which managed to cut DNA at specific spots. The CRISPR was designed to match a specific sequence and the Cas9 was able to recognize the PAM sequences and using certain portions of the Cas9 it could then “break” both strands at opposite positions of the DNA, a specific set of base pairs from the end of the PAM.


This then becomes a useful tool in an ever growing tool-box for DNA modification. In bacteria this cut is applied to viral DNA or RNA and it is a “natural” immune system in the bacteria. In other cells, plants and animals, it enables precise and specific gene editing.

In a recent paper from Zhang’s Lab at Broad they have identified another protein which acts like Cas9. This new system is called CRISPR-Cpf1 and is identified as a class 2 CRISPR system[1]. Specifically Cpf1 is a CRISPR-associated two-component RNA-programmable DNA nuclease. It functions in a manner similar to Cas9 and targeted DNA is cleaved as a 5-nt staggered cut distal to a 5′ T-rich PAM. They have also identified two Cpf1 orthologs exhibit robust nuclease activity in human cells. In the paper in Cell they state:

The microbial adaptive immune system CRISPR mediates defense against foreign genetic elements through two classes of RNA-guided nuclease effectors. Class 1 effectors utilize multi-protein complexes, whereas class 2 effectors rely on single-component effector proteins such as the well-characterized Cas9. Here, we report characterization of Cpf1, a putative class 2 CRISPR effector. We demonstrate that Cpf1 mediates robust DNA interference with features distinct from Cas9. Cpf1 is a single RNA-guided endonuclease lacking tracrRNA, and it utilizes a T-rich protospacer-adjacent motif. Moreover, Cpf1 cleaves DNA via a staggered DNA double-stranded break. Out of 16 Cpf1-family proteins, we identified two candidate enzymes from Acidominococcus and Lachnospiraceae, with efficient genome-editing activity in human cells. Identifying this mechanism of interference broadens our understanding of CRISPR-Cas systems and advances their genome editing applications.

The figure below depicts their interpretation of its functioning.


It is worth comparing these two mechanisms. The Cas9 is a bit more rigid than Cpf1. As noted above and as discussed in the paper and elsewhere, this new protein complex does what Cas9 did but with many more attractive features.

In an MIT press release they state[2]:

The newly described Cpf1 system differs in several important ways from the previously described Cas9, with significant implications for research and therapeutics, as well as for business and intellectual property:

    First: In its natural form, the DNA-cutting enzyme Cas9 forms a complex with two small RNAs, both of which are required for the cutting activity. The Cpf1 system is simpler in that it requires only a single RNA. The Cpf1 enzyme is also smaller than the standard SpCas9, making it easier to deliver into cells and tissues.

    Second, and perhaps most significantly: Cpf1 cuts DNA in a different manner than Cas9. When the Cas9 complex cuts DNA, it cuts both strands at the same place, leaving “blunt ends” that often undergo mutations as they are rejoined. With the Cpf1 complex the cuts in the two strands are offset, leaving short overhangs on the exposed ends. This is expected to help with precise insertion, allowing researchers to integrate a piece of DNA more efficiently and accurately.

    Third: Cpf1 cuts far away from the recognition site, meaning that even if the targeted gene becomes mutated at the cut site, it can likely still be recut, allowing multiple opportunities for correct editing to occur.

    Fourth: The Cpf1 system provides new flexibility in choosing target sites. Like Cas9, the Cpf1 complex must first attach to a short sequence known as a PAM, and targets must be chosen that are adjacent to naturally occurring PAM sequences. The Cpf1 complex recognizes very different PAM sequences from those of Cas9. This could be an advantage in targeting some genomes, such as in the malaria parasite as well as in humans.

The above four properties are quite compelling and worthy of note. Cas9 did have the problem of cutting at opposite sites and trusting that a competent and non-aberrant re-fusion was made. This discovery, assumedly after hundreds of attempts, opens the door on another dimension of the CRISPR world.

As is noted in Xcomony they state[3]:

… the Cpf1 work is still in its infancy. It’s well behind CRISPR/Cas9—which researchers have used to make changes in the cells of all types of organisms, including humans. Several companies are working with CRISPR/Cas9 to create therapeutics for genetic disease. None have reached clinical trials yet.

The issue here is just how extensive is Cpf1 development and how readily available is the technology. The above presentation seems to imply an early stage. They continue:

But work with CRISPR/Cas9 to modify the human germline—eggs, sperm, and embryos—is also coming faster than expected, sparking ethical concerns. An international summit on the topic is scheduled for December in Washington, DC.

Meanwhile, researchers around the world are working to find new versions of Cas9, or new enzymes entirely, like Cpf1, to make the whole enterprise easier. “There is little doubt that… there are additional systems with distinctive characteristics that await exploration and could further enhance genome editing and other areas of biotechnology as well as shed light on the evolution of these defense systems,” Zhang (pictured above speaking at a 2014 Xconomy event) and his coauthors write in the Cell paper.

In other words, Cpf1 is the tip of the iceberg. I’ll outline three differences between Cpf1 and Cas9 that the paper’s authors have highlighted as potentially important for the field. First, for those unfamiliar with CRISPR and gene editing, it helps to think of these enzymes as molecular scissors. Bacteria use them in the wild to defend themselves against invading viruses, cutting up the viral RNA and storing the pieces in a kind of immune system memory bank.

It was only in recent years that the natural system has been modified and harnessed as a gene editing tool. The enzyme—a protein—and its guide—made from RNA—need to be sent into a cell (that’s one difficult trick) and hit the right spot (that’s another difficult trick).

The following is the Xconomy author’s description. It is a restatement of what was in the MIT release but rephrases the key differences:

Here’s why Zhang and his co-authors think Cpf1 could have advantages over Cas9:

—Cpf1 only uses one strand of RNA as a guide to reach its target gene. Cas9 uses two strands. A single-strand system might lead to simpler, cheaper designs and easier delivery of the enzyme-guide complex into cells.

—Once delivered into the cell’s nucleus, Cpf1 makes staggered double-stranded cuts in the target DNA, whereas Cas9 cuts both DNA strands in the same location. This could be important, Zhang and colleagues write, because the staggered ends make it easier to insert a new gene after the old one is removed. That could help get around one of the hurdles of Cas9: Scientists say using Cas9 to replace an old gene with a new one has proven far more difficult than simply cutting out a gene.

—When Cpf1 homes in on a gene, it actually makes the cut off to the side, relatively speaking—farther down the DNA strand. (Imagine your friend holding a string in the exact location that needs snipping. You don’t cut her finger; you cut off to the side.) Zhang and colleagues write that this could be a “potentially useful feature” because it preserves the target site for subsequent rounds of editing.

The off-setting of the splices is a significantly better method. It gives the “sticky” ends approach and tends to much fewer errors. This alone could make this much more attractive.

In a Nature discussion of these results they state[4]:


But now one of the technique's pioneers thinks that he has found a way to make CRISPR even simpler and more precise. In a paper published in Cell on 25 September, a team led by synthetic biologist Feng Zhang of the Broad Institute in Cambridge, Massachusetts, reports the discovery of a protein1 called Cpf1 that may overcome one of CRISPR-Cas9’s few limitations; although the system works well for disabling genes, it is often difficult to truly edit them by replacing one DNA sequence with another.

The CRISPR/Cas9 system evolved as a way for bacteria and archaea to defend themselves against invading viruses. It is found in a wide range of these organisms, and uses an enzyme called Cas9 to cut DNA at a site specified by 'guide' strands of RNA. Researchers have turned CRISPR/Cas9 into a molecular-biology powerhouse that can be used in other organisms. The cuts made by the enzyme are repaired by the cell’s natural DNA-repair processes.
Good, better, best?

CRISPR is much simpler than previous gene-editing methods, but Zhang thought there was still room for improvement.

So he and his colleagues searched the bacterial kingdom to find an alternative to the Cas9 enzyme commonly used in laboratories. In April, they reported that they had discovered a smaller version of Cas9 in the bacterium Staphylococcus aureus2. The small size makes the enzyme easier to shuttle into mature cells — a crucial destination for some potential therapies.

The team was also intrigued by Cpf1, a protein that looks very different from Cas9, but is present in some bacteria with CRISPR. The scientists evaluated Cpf1 enzymes from 16 different bacteria, eventually finding two that could cut human DNA.

They also uncovered some curious differences between how Cpf1 and Cas9 work. Cas9 requires two RNA molecules to cut DNA; Cpf1 needs only one. The proteins also cut DNA at different places, offering researchers more options when selecting a site to edit. “This opens up a lot of possibilities for all the things we could not target before,” says epigeneticist Luca Magnani of Imperial College London.

Cpf1 also cuts DNA in a different way. Cas9 cuts both strands in a DNA molecule at the same position, leaving behind what molecular biologists call ‘blunt’ ends. But Cpf1 leaves one strand longer than the other, creating a 'sticky' end. Blunt ends are not as easy to work with: a DNA sequence could be inserted in either end, for example, whereas a sticky end will only pair with a complementary sticky end.

“The sticky ends carry information that can direct the insertion of the DNA,” says Zhang. “It makes the insertion much more controllable.”

Zhang’s team is now working to use these sticky ends to improve the frequency with which researchers can replace a natural DNA sequence. Cuts left by Cas9 tend to be repaired by sticking the two ends back together, in a relatively sloppy repair process that can leave errors. Although it is possible that the cell will instead insert a designated, new sequence at that site, that kind of repair occurs at a much lower frequency. Zhang hopes that the unique properties of how Cpf1 cuts may be harnessed to make such insertions more frequent.

In contrast we also have an article in The Economist which states[5]:

CRISPR-Cpf1 may also be better than CRISPR-Cas9 in other ways. Cpf1 is a smaller and simpler enzyme (known technically as an endonuclease) than Cas9, which means it will be easier to deliver to the cells whose genes need modifying. And its slightly offset cuts to double-stranded DNA will help researchers to insert genetic patches more efficiently and accurately.

Its discovery also raises the question of how many other endonuclease-based systems are out there in the world’s bacteria. Viral infection is a serious threat to these microbes, and the natural job of both CRISPR-Cas9 and CRISPR-Cpf1 is to recognise viral genes and chop them up before they can do any harm. Conversely, viruses are constantly evolving to escape the antiviral systems’ attentions, meaning bacteria need to generate new ones. The chances are good, therefore, that CRISPR-Cas9 and CRISPR-Cpf1 are not alone. …

The tools to carry out that exploration now exist. CRISPR-Cpf1, for instance, was found not by searching in bacteria directly, but by scrutinising a published database of bacterial genetic sequences, which yielded two species that contain it. Further searches might be equally rewarding—and the more gene-editing systems are discovered, the harder it will be to monopolise their use.

Despite the optimism of those who think the new techniques may calm qualms about genetic engineering, however, some people are bound to have ethical worries—certainly when it comes to applying them to human beings. Earlier this year, for example, when Chinese scientists used CRISPR-Cas9 gene editing on a human embryo (albeit one that was unviable, and could not therefore have developed into a person) there was much brouhaha and several calls for a moratorium on this line of inquiry.

There may not only be ethical worries but as we have discussed previously there is a weaponization approach also readily available.
In a report in Nature World they state[6]:

The CRISPR/Cas9 system evolved as a way for bacteria and archaea to defend themselves against invading viruses. It is found in a wide range of these organisms, and uses an enzyme called Cas9 to cut DNA at a site specified by ‘guide’ strands of RNA. Researchers have turned CRISPR/Cas9 into a molecular-biology powerhouse that can be used in other organisms. The cuts made by the enzyme are repaired by the cell’s natural DNA-repair processes…

The newly described Cpf1 system differs in several important ways from the previously described Cas9, with significant implications for research and therapeutics, as well as for business and intellectual property.

In its natural form, the DNA-cutting enzyme Cas9 forms a complex with two small RNAs, both of which are required for the cutting activity. The Cpf1 system is simpler in that it requires only a single RNA.

Cpf1 cuts DNA in a different manner than Cas9. When the Cas9 complex cuts DNA, it cuts both strands at the same place, leaving ‘blunt ends’ that often undergo mutations as they are rejoined. With the Cpf1 complex the cuts in the two strands are offset, leaving short overhangs on the exposed ends.

Cpf1 cuts far away from the recognition site, meaning that even if the targeted gene becomes mutated at the cut site, it can likely still be re-cut, allowing multiple opportunities for correct editing to occur.

The Cpf1 system provides new flexibility in choosing target sites. Like Cas9, the Cpf1 complex must first attach to a short sequence known as a PAM, and targets must be chosen that are adjacent to naturally occurring PAM sequences.

Finally, in a discussion in Wired the reporting is as follows[7]:

The discovery comes at a time when CRISPR/Cas9 is sweeping through biology labs. So revolutionary is this new genome editing technique that rival groups, who each claim to have been first to the tech, are bitterly fighting over the CRISPR/Cas9 patent. This new gene-editing protein called Cpf1—and maybe even others yet to be discovered—means that one patent may not be so powerful after all…

Many different proteins are associated with CRISPR. But in the early 2010s, Emmanuelle Charpentier, who was studying the flesh-eating bacteria Streptococcus pyogenes, stumbled onto one with special powers. Her bacteria happen to carry Cas9 proteins, which have the remarkable ability to precisely cut DNA based on a RNA guide sequence. In 2012, Charpentier and UC Berkeley biologist Jennifer Doudna published a paper describing the CRISPR/Cas9 system and speculated about its genome editing capabilities. And they filed a patent application. Much more on that patent later.

The patent issue is something we spoke about when the PTO pushed the Broad version through in less than six months, an unheard of process time.

While Cas9 has driven thousands of lab experiments and millions of dollars in funding for startups trying to capitalize on the technology, Cpf1 has remained relatively obscure. This study drags Cpf1 into the limelight. “It’s a very comparable to Cas9 and it has a few different features which could be quite useful,” says Dana Carroll, a biochemist at the University of Utah.

That’s because Cas9 isn’t perfect, despite its hype as a laser-precise genome editing tool. Cpf1 offers some slight advantages. For example, when it cuts double-stranded DNA, it snips the two strands in slightly different locations, resulting in overhang that molecular biologists call “sticky ends.” Sticky ends can make it easier to insert a snippet of new DNA—say, a different version of a gene—though the Cell paper does not actually show data directly comparing Cas9 and Cpf1 when inserting DNA.

Cpf1 is also physically a smaller protein, so it may be easier to put into human cells. It requires only one RNA molecule instead of two, with Cas9. But it’s not a rival so much as a complementary tool: The two proteins favor binding to different locations in the genome, so together, they might allow more flexibility in where scientist want to cut.

The writer then returns to the patent issues:

Not long after Doudna and UC Berkeley filed a patent, the Broad Institute and MIT filed their own patent on behalf of Zhang for the CRISPR/Cas9 system. Zhang had been working on actually showing that CRISPR/Cas9 can edit mammalian genomes in mammalian cells, an application he published in 2013 and says he came up with independently. The Broad’s and MIT’s attorney paid a fee to accelerate their application. Ultimately, the US Patent and Trademark Office awarded the patent to Zhang, MIT, and the Broad Institute. The University of California, obviously unhappy with the decision, filed an application for an interference proceeding to get the USPTO to reconsider. That process is ongoing.

But biotech companies have raced ahead to develop therapeutics and techniques with the system. Feng and Doudna have since licensed their technology to rival companies, Editas and Caribou. Charpentier also cofounded Crispr Therapeutics in Switzerland. Whoever wins the patent dispute will have a monopoly on CRISPR/Cas9 technology, the hottest new thing in biotech.

But with Cfp1, the stakes of that specific patent dispute go down. A lab or company could use Cfp1 without infringing on the CRISPR/Cas9 patent. “It takes power away from whoever the winner is going to be,” says Jacob Sherkow, a NYU law professor. (Zhang has indicated the rights to Cpf1 may not necessarily go to the company he cofounded, Editas.) Whether a CRISPR/Cfp1 system is patentable as a separate invention—Sherkow says it probably is—perhaps isn’t even relevant because its very existence means Cas9 is no longer the only game in town.

This latter observation is of significant value. Namely Cpf1 if it is truly better makes Cas9 battles of less value.

It is of continuing interest to follow the dimensions of this new “tool box” available to those of us working on gene changes.

References

1.     Zetsche, B., et al, Cpf1 Is a Single RNA-Guided Endonuclease of a Class 2 CRISPR-Cas System, Cell DOI: http://dx.doi.org/10.1016/j.cell.2015.09.038

2.     Ledford, H., Alternative CRISPR system could improve genome editing, Nature News, 25 September 2015