In CRISPR circles, the original Cas9 still commands the lion’s share of the attention. But researchers including Banfield and Koonin are mining the astonishing diversity of microbial life on planet earth to identify unknown organisms and catalogue the diversity of CRISPR immune systems and Cas genes therein. Scientists have wasted no time in adapting some of these systems for new research and diagnostic purposes.
One early and profound addition to the toolbox was to take the molecular scissors and immediately blunt the blades to mute Cas9’s DNA cutting function. That may sound counterintuitive, but the RNA programmability of Cas9 serves a multitude of purposes beyond cutting DNA. Cas9 can be used to ferry many kinds of molecules to a specific spot in the genome to modulate gene expression up or down (CRISPR activation or interference). This non-cutting Cas9, described by Stanley Qi and colleagues, is called “dead Cas9.”17 Its applications include base editing,18 a new riff on CRISPR genome editing in which Cas9 is used not to cut DNA but to position different enzymes to nick the DNA and perform pinpoint chemistry on a specific base. (We’ll return to this in chapter 22.)
The original Cas9 protein from S. pyogenes (SpCas9) is made up of 1,366 amino-acid building blocks, which makes for a tight squeeze when packing this molecule into the limited cargo space of the most popular gene therapy vector, the adeno-associated virus (AAV). There are many flavors of Cas9 derived from other microbes, some significantly smaller than SpCas9, others that recognize a different PAM site. Many of these provide new tools, increasing the options for the CRISPR engineer. Doudna’s team also found a way to put Cas9 on pause, holding it in a locked formation with the molecular equivalent of a plastic zip tie, which can be snipped as required to release the nuclease.19 This affords researchers more control in exactly where—or when—they unleash Cas9, reducing the chance of undesirable off-target effects.
In 2016, almost a decade after they first talked CRISPR over coffee, Doudna and Banfield unearthed a trove of new CRISPR tools from a metagenomic sampling of microbes that have not yet been cultured in the laboratory.20 Among the highlights were two diminutive Cas nucleases, named CasX and CasY. CasX is only 60 percent the size of SpCas9. It cuts DNA in much the same manner, even though it bears no sequence similarity to SpCas9, suggesting it evolved quite independently. And unlike SpCas9, it is derived from a bacterial species that does not naturally infect humans, so in principle it would not have any of the potential immunogenicity concerns.
Another interesting Cas scalpel is Cas12 (formerly known as Cpf1), in particular Cas12a, discovered by the Zhang lab.21 Unlike Cas9, this enzyme produces a staggered cut—slicing the two strands of the helix in different places, rather than a clean cut. It is a small protein and doesn’t require a tracrRNA. While independently investigating the properties of Cas12a, members of the Zhang and Doudna labs were shocked to find that Cas12a did something quite different to single-stranded DNA—it didn’t so much cut DNA as shred it.22 Another new addition to the toolbox, Cas13, did much the same thing to RNA. If the detection of a specific DNA or RNA molecule could be coupled to some sort of chemical signal, the groups would have a simple diagnostics platform.
That’s exactly what the two groups did. Two of Doudna’s students, Janice Chen and Lucas Harrington, helped create Mammoth Biosciences to commercialize their diagnostics system, dubbed DETECTR. (Chen’s brother is world champion figure skater Nathan Chen.) Meanwhile, two of Zhang’s protégés, Omar Abudayyeh and Jonathan Gootenberg, joined Zhang, Pardis Sabeti, and the cofounders of Sherlock Biosciences; you don’t need to be a pipe-smoking detective to know what their system is called.VI
Here’s an example: let’s say we want to program Cas12 to detect the SARS-CoV-2 virus responsible for the COVID-19 pandemic. A series of guide RNAs are designed to recognize certain sequences that have been amplified from the coronavirus genome. But once Cas12 recognizes that sequence, a new enzymatic property is switched on, such that it will cut (and keep cutting) any single-stranded DNA molecules in the vicinity. By adding reporter molecules that light up when cut, the presence of even trace amounts of the virus can be detected using a simple color assay on a paper strip.23
Similarly, the Cas13 family can be used to detect infections such as flu, dengue, and Zika, and of course COVID-19.24 Once activated, Cas13 exhibits what Zhang calls “collateral RNase activity”—it keeps cutting RNA. By supplying a suitable quantity of chemically tagged RNA reporter molecules, his team has the basis of a simple, portable detection system that can work on urine, blood, or saliva. The presence of virus will switch on Cas13, cutting the RNA reporters and releasing a fluorescent marker that can be read on a simple paper strip much like a pregnancy test.
The reliability of simple, cheap, one-stop diagnostic tests have a bad reputation following the debacle of Theranos, the theatrically overhyped Silicon Valley unicorn launched by Elizabeth Holmes that crashed from a $9 billion valuation to bankruptcy following great investigative reporting by John Carreyrou.25 Unlike Theranos, which only belatedly published a single peer-reviewed study,26 the Doudna and Zhang teams have already laid out the science and technology behind their diagnostics discoveries in a series of top-tier publications.
The portability of Mammoth’s and Sherlock’s kits could find huge markets—at home for the flu, in hospitals for antibiotic resistance, and in the field where outbreaks of coronavirus and other viruses emerge. Chen’s group has shown DETECTR can accurately detect HPV samples in a fraction of the time of a conventional test. Led by Harvard’s Sabeti, the SHERLOCK test has already shown promising results in detecting cases of Lassa fever in Nigeria, dengue in Senegal, and Zika virus in Honduras. Both companies are actively adapting their platforms to detect the COVID-19 virus. Beyond diagnostics, applications beckon in areas from food security and agriculture to bioterrorism. The Zhang lab has already applied SHERLOCK to gene detection in plants, pointing to an array of applications in detecting pathogens or pests.27
In Toronto, Joseph Bondy-Denomy “found something amazing that we never expected,” said his PhD supervisor Alan Davidson.28 He discovered anti-CRISPRs, a growing family of viral proteins that are able to disarm or neutralize bacterial CRISPR defenses—the rocks and paper to CRISPR’s scissors. Erik Sontheimer described a means to use anti-CRISPRs to restrict genome editing to a tissue of choice. Harvard Medical School’s Amit Choudhary is identifying small chemicals that can fine tune Cas enzyme activity. DARPA has launched a program called Safe Genes to fund research into anti-CRISPRs, and Bondy-Denomy wasted little time in cofounding a company, Acrigen Biosciences, to make gene editing safer and more efficient.
Other toolbox additions include Cas3, a DNA shredder to generate large deletions; a system called EvolvR to introduce mutations and evolve a specific target region; and systems that engineer programmed DNA insertions at a target site. Similar ideas were developed in parallel by Samuel Sternberg’s group at Columbia University, and the Zhang lab. Sternberg adapted a CRISPR system from Vibrio cholera to develop a programmable system based on transposons (parasitic jumping genes) to insert a custom DNA sequence at a specific site in the genome.29 The system offers an appealing alternative to genome engineering without breaking DNA or triggering the cellular DNA damage response.
Scientists are just starting to appreciate some of these new tools, but these are just the tip of the iceberg. Zhang says about 150,000 microbial genomes have been sequenced, but we only understand something about the defense systems in about one third of them. There are so many more secrets yet to be revealed from the sequences of our microbial ancestors, which have had a mere billion years to innovate and evolve.