From Bacterial Immunity to Genome Editing: The 2014 Dr. Paul Janssen Award Symposium

Journal Articles

Emmanuelle Charpentier

Charpentier E, Marraffini LA. Harnessing CRISPR-Cas9 immunity for genetic engineering. Curr Opin Microbiol. 2014;19:114-9.

Chylinski K, Le Rhun A, Charpentier E. The tracrRNA and Cas9 families of type II CRISPR-Cas immunity systems. RNA Biol. 2013;10(5):726-37.

Chylinski K, Makarova KS, Charpentier E, Koonin EV. Classification and evolution of type II CRISPR-Cas systems. Nucleic Acids Res. 2014;42(10):6091-105.

Deltcheva E, Chylinski K, Sharma CM, et al. CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III. Nature. 2011;471(7340):602-7.

Jinek M, Chylinski K, Fonfara I, et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science. 2012;337(6096):816-21.

Jennifer Doudna

Hochstrasser ML, Taylor DW, Bhat P, et al. CasA mediates Cas3-catalyzed target degradation during CRISPR RNA-guided interference. Proc Natl Acad Sci U S A. 2014;111(18):6618-23.

Jinek M, Jiang F, Taylor DW, et al. Structures of Cas9 endonucleases reveal RNA-mediated conformational activation. Science. 2014;343(6176):1247997.

Nuñez JK, Kranzusch PJ, Noeske J, et al. Cas1-Cas2 complex formation mediates spacer acquisition during CRISPR-Cas adaptive immunity. Nat Struct Mol Biol. 2014;21(6):528-34.

Charles A. Gersbach

Gersbach CA. Genome engineering: the next genomic revolution. Nat Methods. 2014;11(10):1009-11.

High K, Gregory PD, Gersbach C. CRISPR technology for gene therapy. Nat Med. 2014;20(5):476-7.

Kabadi AM, Gersbach CA. Engineering synthetic TALE and CRISPR/Cas9 transcription factors for regulating gene expression. Methods. 2014;69(2):188-97.

Kabadi AM, Ousterout DG, Hilton IB, Gersbach CA. Multiplex CRISPR/Cas9-based genome engineering from a single lentiviral vector. Nucleic Acids Res. 2014. [Epub ahead of print]

Luciano A. Marraffini

Barrangou R, Marraffini LA. CRISPR-Cas systems: prokaryotes upgrade to adaptive immunity. Mol Cell. 2014;54(2):234-44.

Goldberg GW, Jiang W, Bikard D, Marraffini LA. Conditional tolerance of temperate phages via transcription-dependent CRISPR-Cas targeting. Nature. 2014. [Epub ahead of print]

Hatoum-Aslan A, Marraffini LA. Impact of CRISPR immunity on the emergence and virulence of bacterial pathogens. Curr Opin Microbiol. 2014;17:82-90.

Heler R, Marraffini LA, Bikard D. Adapting to new threats: the generation of memory by CRISPR-Cas immune systems. Mol Microbiol. 2014;93(1):1-9.

William R. Strohl

Kinder M, Greenplate AR, Grugan KD, et al. Engineered protease-resistant antibodies with selectable cell-killing functions. J Biol Chem. 2013;288(43):30843-54.

Raju TS, Strohl WR. Potential therapeutic roles for antibody mixtures. Expert Opin Biol Ther. 2013;13(10):1347-52.

Strohl WR. Antibody discovery: sourcing of monoclonal antibody variable domains. Curr Drug Discov Technol. 2014;11(1):3-19.

Strohl WR. Modern therapeutic antibody drug discovery technologies. Curr Drug Discov Technol. 2014;11(1):1-2.

Sponsors

This symposium was made possible with support from

Featured Speakers:
Emmanuelle Charpentier, Helmholtz Centre for Infection Research, Hannover Medical School, Germany; Umeå University, Sweden
Jennifer Doudna, University of California, Berkeley; Howard Hughes Medical Institute

Highlights

• The bacterial CRISPR/Cas system evolved as a form of adaptive immunity against mobile genetic elements such as bacteriophages.
• Of the three major types of CRISPR/Cas systems, type II is the easiest to adapt for genome editing.
• CRISPR/Cas-based techniques are faster and easier to perform than previous genome editing methods.
• The Cas9 enzyme uses target-specific RNA transcripts to identify a DNA sequence, which it then cleaves.
• Engineered CRISPR/Cas9 systems allow researchers to make arbitrary modifications to virtually any piece of DNA in a cell.
• By altering Cas9 or the targeting RNA, the system can also be used for reversible gene regulation and RNA quantitation.

Introduction

In the public imagination, basic science involves intellectuals studying arcane problems out of pure curiosity, while applied research is a practical exercise that solves real-world problems. Predictably, funding tends to go to applied projects, despite scientists' insistence on the importance of basic research.

It is hard to imagine a more comprehensive rebuttal to this false dichotomy than the story of CRISPR and Cas. The discovery of an odd genetic pattern in bacteria called clustered regularly interspaced short palindromic repeats (CRISPR) and a set of CRISPR-associated (Cas) genes stemmed from curiosity-driven research in bacteriology. However, two investigators soon found they could adapt the CRISPR/Cas system to edit DNA—not just in bacteria but also in eukaryotes. Two years later, laboratories around the world are using this technique to study human diseases such as cancer and HIV, and biopharmaceutical companies are designing CRISPR/Cas-based strategies for gene therapy and drug discovery.

Basic research on bacterial immunity led to a revolutionary genome-editing tool.

Emmanuelle Charpentier described how work on small regulatory RNA molecules in Streptococcus pyogenes led her team to the CRISPR loci in the bacterial genome, a series of repeated sequences interspersed with other segments, some copied from bacteriophages. Certain Cas gene products can use RNA transcribed from the CRISPR locus to identify foreign genomes and target them for destruction. Bacteria have evolved three types of CRISPR/Cas systems, but Charpentier found the CRISPR/Cas9 machinery from S. pyogenes easiest to adapt for gene editing in other types of cells.

Jennifer Doudna initially focused on a more complex CRISPR/Cas system, which her lab was studying for its role in bacterial immunity. After she and Charpentier met at a conference in 2011, they began collaborating on the simpler CRISPR/Cas9 mechanism. Doudna described their work to determine how CRISPR/Cas9 identifies foreign DNA. They then quickly adapted the mechanism for a variety of uses in prokaryotic and eukaryotic cells. CRISPR/Cas9-based techniques can now measure RNA levels in a cell, turn off transcription of specific genes, or make highly targeted deletions, mutations, and insertions in the genome.

Three researchers working in the field followed with presentations describing the implications of this novel technology for basic science, medical research, drug development, and human health. The meeting closed with a panel discussion led by Nobel laureate Craig C. Mello, who joined the other speakers in a wide-ranging exploration of the role of basic research, the evolution of bacterial immunity, and the potential benefits, pitfalls, and limitations of editing the human genome.

Rewrite: CRISPR/Cas genome editing

In 1970, molecular biology pioneer Jacques Monod assured the world that science would not give us the ability to tinker with our own DNA. "The genome's microscopic proportions," he asserted, "today and probably forever rule out manipulation of this sort." He would doubtless be shocked to see how wrong his prediction was.

Scientists have now developed three major techniques for editing the genome directly. The first, zinc finger nucleases, allows targeted deletion and insertion of genes in many types of cells, but each project requires extensive design and debugging. A more recent strategy, using transcription activator-like effector nucleases (TALEN), is simpler but still unreliable. CRISPR/Cas genome editing, based on the bacterial system, offers simplicity, efficiency, and specificity.

Emmanuelle Charpentier of Helmholtz Center for Infection Research and Umeå University described how CRISPR/Cas genome editing has enabled an explosion of experiments and discoveries across disciplines. "The system offers a number of advantages. It's cheap and easy; it's a democratic tool," she said.

Genome editing with sequence-specific nucleases. (Image presented by Emmanuelle Charpentier courtesy of Douglas B. Menke. [Menke DB. Genesis. 2013;51(9):605-18.])

CRISPR/Cas systems have two key components: a guide RNA (a short RNA sequence complementary to a specific DNA target) and a nuclease, an enzyme that cleaves DNA. The nuclease uses the RNA sequence to bind both strands of DNA at the target site, then cuts the DNA. By engineering the guide RNA sequence, researchers can use the system to delete, replace, or add arbitrary pieces of DNA virtually anywhere in a prokaryotic, eukaryotic, or archaeal genome.

Charpentier and her colleagues did not set out to develop a revolutionary genome editing tool. Instead, her team was interested in understanding the functions of small regulatory RNA molecules in S. pyogenes. The bacterium uses small RNAs to regulate gene expression through a mechanism analogous to the short interfering RNA (siRNA) system in eukaryotes. Charpentier's group focused on a small RNA called tracrRNA, which is transcribed from a locus next to a CRISPR-associated (Cas) gene.

The Cas operon yields several gene products for an adaptive immune system in bacteria. When a bacteriophage (phage) or other foreign DNA enters the cell, Cas enzymes recognize it as foreign, cut it into pieces, and insert some of the pieces into CRISPR loci in the genome. Other Cas enzymes then use transcripts from CRISPR loci to recognize and destroy the phage the next time it infects the cell. "This will function as a memory device, so that on second infection with the same phage the system can recognize the phage and destroy it," Charpentier said.

Snippets of phage sequences in bacterial CRISPR loci encode memories of past infections.

Bacteria have evolved three types of CRISPR/Cas systems. Types I and III are relatively complex, using multiple enzymes to recognize and destroy invading phages. Type II relies on a single enzyme, Cas9, which simultaneously recognizes a specific phage sequence from CRISPR loci and makes a double-stranded break in the phage's DNA.

After purifying and characterizing Cas9 proteins in vitro, Charpentier and her colleagues found Cas9 binds tracrRNA and a target-specific RNA sequence (guide RNA) transcribed from CRISPR loci. When this protein–RNA complex encounters a piece of DNA that carries both the CRISPR-derived sequence and a short three-base sequence pattern called a protospacer adjacent motif (PAM), Cas9 cuts the DNA. In the S. pyogenes type II CRISPR system, the PAM sequence is two guanine nucleotides plus any other nucleotide (NGG). PAM sequences differ according to which species the Cas9 is derived from. The PAM may act as a safety mechanism, preventing Cas9 from destroying the cell's own CRISPR loci, which do not contain the sequence.

The more the scientific community has studied and modified CRISPR/Cas9, the more useful the mechanism has become as a laboratory tool. Sequence database searches have revealed hundreds of different Cas9 genes in different bacteria. Evolution has conserved CRISPR/Cas9 systems throughout bacterial species, but the Cas9 enzymes from distantly related species have largely diverged in sequence and are not interchangeable with regard to their respective dual-RNAs. "This leads to really endless possibilities in the CRISPR/Cas9 toolbox for multiplex gene editing," Charpentier said. The S. pyogenes Cas9 seems to work well in human cells; the main impediment to using the system clinically is the difficulty delivering it to tissues throughout an organism.

Other researchers have also modified Cas9 to introduce offset breaks in DNA, leaving "sticky" ends that make it easier to insert new sequences into a genome. Fusing other types of enzymes to the DNA-recognition domain of Cas9 enables researchers to block the elongation of gene transcripts from a target location in the genome.

Basic science to biotech: an evolutionary gift that keeps on giving

Although Charpentier and Doudna turned the CRISPR/Cas system into a practical laboratory tool, some of the basic research behind the system was done by scientists in industry. While trying to understand the biology of yogurt cultures, these investigators demonstrated how CRISPR/Cas works as an adaptive immune system. Their work inspired Charpentier and Doudna, who independently started studying different aspects of CRISPR/Cas immunity.

While Charpentier's group explored the relatively simple type II CRISPR/Cas system, Jennifer Doudna of the University of California, Berkeley, and her colleagues were working on the more complex groups of Cas enzymes. "Those systems require multiple proteins to come together," she explained. "From the perspective of thinking about this as a technology, it seemed rather daunting ... to assemble many proteins in a foreign cell and get it to function."

After Doudna and Charpentier began collaborating in 2011, with a focus on the type II CRISPR/Cas system, they began looking for ways to make the simple system even simpler by combining the two small RNA molecules that bind Cas9—the tracrRNA and the target-specific RNA. "Wouldn't it be great if we could turn this three-component system into a two-component system?" Doudna asked. Adding a short linker sequence between the RNAs did the trick; now, scientists can design a single transcript consisting of a tracrRNA sequence linked to a targeting sequence. Cas9 will bind the engineered transcript, then find and cut the target DNA. There's a catch, but it's a small one: the target sequence must be adjacent to a PAM sequence.

Programming Cas9 with single-guide RNAs. (Image courtesy of Jennifer Doudna)

Doudna's group also characterized the mechanism of Cas9 binding. The enzyme binds and releases DNA repeatedly until it finds a PAM, which holds the enzyme a moment longer than other sequences. By mutating different bases of the targeting sequence, the investigators found that the bases closest to the PAM are the most important for subsequent recognition of the target. "It's very consistent with this idea of PAM binding and then a sequential unwinding of the duplex adjacent to PAM, leading to full engagement of the complex," Doudna said. Once the whole sequence has matched its target, Cas9 cuts the DNA.

Normally, the process only cuts double-stranded DNA. However, the scientists can trick it into recognizing single-stranded DNA or RNA by providing a single-stranded oligonucleotide matching a target sequence, with an added PAM. This PAM-mer binds the single-stranded target and creates a double-stranded molecule that Cas9 can recognize.

This technology enabled another application of the CRISPR/Cas system, targeting RNA to control gene expression. By choosing a transcribed DNA sequence that lacked a PAM and designing a CRISPR/Cas9-targeting oligonucleotide with a PAM-mer to bind it, the researchers could cut single-stranded targets with the same sequence—including the transcribed RNA. Cas9 does not cut the DNA lacking a PAM, but it can recognize and cut the single-stranded target with the PAM-mer sequence. Destroying RNA shuts off gene expression, and removing Cas9 activity reactivates it, allowing the researchers to turn the gene on and off at will.

CRISPR/Cas9-based tools can now edit DNA, regulate RNA transcription, or measure RNA levels.

The team modified the system further to build a tagless RNA recognition system. Current techniques to identify particular RNA molecules rely on attaching chemical tags that might interfere with the molecules' natural functions. Using the PAM-mer system, Doudna and her colleagues can tag target RNA indirectly by binding it to a catalytically inactive (tagged) Cas9. They can use this approach to analyze transcription levels and provide a visual readout of RNA expression in living cells. A similar approach might highlight viral transcripts in a cell or identify which copy of a gene is being expressed.

During the Q&A period, Doudna discussed the problem of chromatin, which packs much of the cellular DNA into a form that is normally inaccessible to enzymes. So far, researchers have found little effect of chromatin on CRISPR/Cas9 DNA targeting, suggesting that the enzyme can unwind even tightly packed, transcriptionally inactive DNA to find and cleave its target. Though limitations to the technology may arise, CRISPR/Cas9 currently looks like a genetic engineer's wish come true.

Speakers:
Luciano A. Marraffini, The Rockefeller University
Charles A. Gersbach, Duke Center for Genomic and Computational Biology
William R. Strohl, Janssen Research & Development

Highlights

• The type III CRISPR/Cas system tolerates lysogenic phages but destroys phages that activate lytic genes, which destroy the cell.
• Immune tolerance by CRISPR/Cas allows bacteria to keep useful foreign DNA elements while eliminating harmful ones.
• Genome editing could correct genetic defects, such as the dystrophin gene defect that causes Duchenne muscular dystrophy (DMD). Using CRISPR/Cas9, researchers can repair the dystrophin gene in cells from DMD patients.
• Fusing Cas9's DNA-binding domain to a transcriptional activator allows precise induction of genes.
• New genome editing strategies are reviving interest in gene therapy. If genome editing works in the clinic, the ability to correct underlying genetic causes could cure many diseases.
• Delivering CRISPR/Cas9 or genetically edited cells to the right tissues is now the major challenge in clinical genome editing.

How to train your phage

The simplicity of the type II CRISPR/Cas system, with its single Cas9 DNA-cleaving enzyme, has made it the darling of molecular biologists; but bacteria also carry two other CRISPR/Cas mechanisms that appear to operate independently. Different species carry distinct subtypes of all three systems, indicating that CRISPR/Cas has continued to diverge and evolve in multiple lineages. This redundancy and complexity puzzled Luciano A. Marraffini of Rockefeller University. "Cas9 is a fantastic enzyme to cleave double-stranded DNA, so why," he wondered, "[have bacteria] evolved many different CRISPR/Cas systems that have slightly different mechanisms?"

How does CRISPR/Cas distinguish beneficial DNA elements from harmful ones?

Redundant mechanisms abound in biology, but CRISPR/Cas takes the concept to new heights. About 20% of the spacer sequences in CRISPRs match other CRISPRs, suggesting that the three types of CRISPR/Cas systems may be at war not only with invading bacteriophages, but also with each other. Unchecked, these systems would make a bacterial cell a hostile environment for all types of foreign DNA, yet bacteria routinely trade plasmids and helpful phages. There must be some way for a bacterium to distinguish threatening from benign sequences—some form of immune tolerance.

CRISPR immunity is mediated by RNA-guided nucleases. (Image courtesy of Luciano A. Marraffini)

To test how such tolerance might work, Marraffini and his colleagues studied viral reproduction, which can proceed via either the lytic or the lysogenic cycle. In the lytic cycle, the phage rapidly kills its host cell and releases a new generation of progeny phage particles. In the lysogenic cycle, however, lytic genes are silenced and the phage integrates into the host genome, where it may contribute genes that aid the bacterium's survival. "The integration of phages is important for bacteria, and it would be nice if CRISPR could distinguish between phages that want to kill the bacterium versus phages that want to just integrate and express something that is good for the bacterium," Marraffini said. An integrated phage retains the ability to induce its lytic genes, so the benefits of harboring such a phage come with the constant danger that it could kill the cell.

In the bacterium Staphylococcus aureus, the investigators found that the type III CRISPR/Cas system encodes an RNA sequence complementary to a sequence from a lytic gene in the phage phiNM1. Carrying this sequence in the CRISPR locus provides strong protection against lytic infection with phiNM1, but has no effect on the phage's ability to lysogenize the bacteria. However, the phage's DNA sequence is the same whether the phage is carrying out its lytic or its lysogenic cycle, so the researchers wondered how the CRISPR/Cas system senses which program the phage is following.

Lysis vs. lysogeny. (Image courtesy of Luciano A. Marraffini)

Mutant phiNM1 phages that escape CRISPR/Cas destruction provided some clues. These phages carry mutations either in the target sequence CRISPR/Cas recognizes or in a transcriptional promoter for lysogenic genes. Marraffini's team determined that if the target sequence is not transcribed, it remains invisible to the type III CRISPR/Cas system. Because the target is in a lytic gene that is not transcribed during lysogeny, targeting it with this system provides an elegant strategy for tolerance. Marraffini refers to this system as "domesticating the phage." He explained, "the bacterium can keep the phage when it's nice and bringing something good to it, but then the moment the phage goes wild and tries to kill the bacterium, then the CRISPR system will cut it."

So far, this type of tolerance appears to be limited to type III CRISPR/Cas systems. Marraffini and his colleagues found that type II CRISPR/Cas systems attack any DNA sequence found in their CRISPR locus, and other researchers saw similar results with type I CRISPRs.

In the panel discussion at the end of the meeting, Marraffini also discussed using CRISPR/Cas systems to measure bacteriophage diversity. Only 5%–10% of the foreign sequences found so far in CRISPRs correspond to any known phage, suggesting that a vast number of phages remain to be discovered.

Strong medicine: human genome editing

While gene editing has proven its worth in the laboratory, medical researchers are still waiting for a version of the technology that will work in humans. Most gene therapy trials have relied on viral vectors, sometimes with troubling side effects, and even the most recent clinical techniques have been limited to adding new genes. For diseases that involve a defective gene, that approach often fails; the defective gene product continues to cause damage even with a good copy of the gene available.

Duchenne muscular dystrophy (DMD) is one such disease. In about one in 3500 male births, a defective dystrophin gene causes progressive muscle wasting, crippling the boy by age 10 and killing him by about age 30. Experimental treatments using antisense RNA to remove the defective exon from the dystrophin RNA transcript have shown some promise, but require large, repeated doses of RNA and only help in about 13% of DMD cases. "We saw this as an opportunity for genome editing," said Charles A. Gersbach of Duke Center for Genomic and Computational Biology. "Rather than removing the exon from the mRNA, perhaps we could go in using genome editing, remove it from the genome ... and the gene could be permanently corrected."

Genome editing has the potential to cure genetic diseases permanently.

Gersbach is not new to genome editing. When he first tried to edit genes using zinc finger nucleases, it took a year or two to get each system working. The newer TALEN system shortened the time to a month or two. With the CRISPR/Cas9 system developed by Charpentier and Doudna, however, his lab can go from thinking about a gene edit to characterizing the edited cells in less than two weeks. This speed allows the team to work quickly and simultaneously on several strategies to treat DMD.

Using CRISPR/Cas9, Gersbach's lab recently targeted the defective exon in muscle cells from a DMD patient. The experiment worked, precisely deleting the exon. Differentiating the cells into muscle tissue revealed that they now produced functional dystrophin. Next, the investigators modified the system to remove all the exons affected in the majority of DMD patients, again yielding functional dystrophin. "It just shows the power of the CRISPR/Cas system for genome editing, in that you have the ability to go in and really modify the genome however you want," Gersbach said.

Restoring dystrophin expression. (Image courtesy of Charles A. Gersbach)

The researchers are now trying multiple strategies to get the approach into the clinic. Modified cells from patients can engraft in immune-suppressed mice, where they grow and produce dystrophin, suggesting that cell-based therapies might work in humans as well. Viral vectors carrying the Cas9 nuclease gene and appropriate targeting sequences can correct dystrophin gene defects directly in a mouse model of DMD. That system produces only a modest effect, but it could be enough to maintain muscle mass in DMD patients.

Gersbach and his colleagues are also trying to use CRISPR/Cas9 to regulate other muscle-specific genes. They found that fusing Cas9's DNA-binding domain to a transcriptional activator can induce cells to express MyoD, a master regulator of muscle growth and repair. The induction is highly specific, in contrast to other techniques. "We were quite surprised by the specificity here. I think that it's rare that you find some type of treatment that you can give cells and one gene and one gene only across the genome is activated," he said. Reprogramming the cells this way seems to produce more stable MyoD levels than techniques that introduce a MyoD gene on an expression vector. The CRISPR/Cas9 system gives the team a reliable way to turn stem cells into muscle tissue, which could be used to regenerate damaged tissues in DMD patients.

Tools for a new era

When biologists develop a new tool for manipulating genes or proteins, the pharmaceutical industry takes note. Though biotechnology is infamous for overpromising what new techniques can do, most of the major breakthroughs in molecular biology eventually have led to useful—and lucrative—new treatments. Therapeutic proteins made with recombinant DNA techniques are now a $45 billion a year business, and monoclonal antibodies generate $76 billion annually.

As William R. Strohl of Janssen Research & Development explained, cell- and genome-based therapies have now reached the stage of development that monoclonal antibodies had in the 1990s. Though only one gene therapy and ten cell therapy products have been approved for clinical use, companies are conducting thousands of clinical trials in these fields.

Approved therapeutic and prophylactic molecules. (Image courtesy of William R. Strohl)

Drug discovery labs already use genome editing tools, especially CRISPR/Cas9, as research tools to find new drug targets and to understand disease mechanisms. "Of course what we're [most] interested in are the therapeutic uses," Strohl said. Indeed, CRISPR/Cas9 genome editing might be a crucial lifeline for the long-struggling field of human gene therapy.

In contrast to traditional small-molecule, protein, and antibody treatments, gene editing has the potential to cure chronic diseases. "Often times you take a small-molecule [drug] once a day or twice a day and you're treating the disease, but you're not curing it. You're not curing the underlying mechanisms of that disease," Strohl said. The potential to generate real cures has driven gene therapy research for thirty years, but the field has encountered several obstacles.

In 1985, scientists used a retroviral vector to correct an adenosine deaminase (ADA) gene defect in human cells, and five years later a child with severe immunodeficiency caused by the defect received the first successful gene therapy. The field progressed slowly until 1999, when a patient died in a gene therapy trial at the University of Pennsylvania. Further setbacks in other clinical trials stalled gene therapy efforts for several more years. CRISPR/Cas9, with its simplicity and apparent reliability, could not have come at a better time for the field.

Genome editing can stop numerous pathogenic processes, at least in cells and animals.

Strohl described several recent clinical and preclinical studies on genome editing. In one project, scientists used the CRISPR/Cas9 system to alter the CCR5 gene in CD4⁺ T cells of HIV patients. CCR5 is a co-receptor for HIV, and the edited cells became resistant to the virus. Another team used the zinc finger nuclease technique to edit a blood clotting gene in mice, curing the animals of hemophilia. Other efforts have shown promising results against diverse diseases, including hypercholesterolemia and malaria.

Despite the impressive preclinical results, Strohl warned that the field still faces some daunting obstacles, particularly, "delivery not only to the cells that you're looking to deliver to but also to the organs," he said. As a result, he predicts that the first successful genome editing therapies will be against diseases of the eye, gastrointestinal tract, lung, and heart—organs that are relatively easy to reach.

Companies might also use CRISPR/Cas9 editing to improve established cell-based therapies. These treatments often involve taking cells from a patient, treating defects in a laboratory, then returning the cells to the same patient. Because each treatment is custom-made, cell-based therapies are expensive and time-consuming. Different patients cannot be treated with the same cells, because their immune systems would recognize and reject the cells as foreign. Through genome editing, however, scientists may be able to generate cells that lack the genes responsible for transplant rejection, enabling companies to produce standard, healthy cells for delivery to any patient.

Regardless of which strategies finally work in the clinic, CRISPR/Cas9 has already been a boon for both basic and applied research. "This field ... in two and a half years has come an awfully long way," Strohl concluded.

Moderator:
Craig C. Mello, University of Massachusetts Medical School; Howard Hughes Medical Institute

The meeting concluded with a wide-ranging panel discussion. Introducing the session, Nobel laureate and previous Janssen Award recipient Craig Mello explained the impact of genome editing on biology and biotechnology. "We have all the genome sequences now, and now we can do more than just stare at them," he said, "we can actually do things to rearrange the information and ask what specific genes do." He added that besides the possible medical applications, technologies such as CRISPR/Cas9 will enable rapid production of new types of genetically modified crops to make agriculture more sustainable.

CRISPR/Cas9 also elucidates the complex relationships among bacteria, phages, and plasmids. "The microbiome is something really interesting, and the impact on human health we are just starting to understand, but then the next step will be understanding the role of phages," Marraffini explained. Similarly, variations among CRISPR systems have highlighted the extreme diversity of bacteria, archaea, and their viruses. "I think actually what's emerging is that different types of bacteria use the CRISPR system differently," Doudna said. Some laboratory strains have downregulated or even jettisoned components of their CRISPR/Cas machinery, making them more amenable to genetic modification. Meanwhile, phages have evolved anti-CRISPR systems to evade this bacterial immune system.

An audience member asked the panelists where the limits of genome editing might lie; for example, could CRISPR/Cas9 provide the key to stopping human aging? "I'd be lying if I said we hadn't thought about it," Gersbach replied. "It's theoretically possible ... but there are major technical hurdles," he added. "My impression," Mello concurred, "is the sky's the limit if you look far enough down the road ... but obviously there's a lot of work to do before anything like that is safe."

Indeed, CRISPR systems themselves provide a vivid reminder of the risks of making arbitrary genetic modifications. Bacteria appear to have evolved this system as a means of intercepting potentially harmful DNA that might enter the cell, while still integrating useful plasmids and phages. "It's kind of analogous to when you're sitting at your computer and you just downloaded a new application, and your computer says 'are you sure you want to open that?'" Mello explained.

Concerns about safety also explain why CRISPR/Cas9 techniques have moved slowly toward clinical applications, even as researchers have adopted the system wholesale. "Most of the excitement about CRISPR/Cas is how easy, fast, and cheap it is to use. That's great for scientists, but for a therapy you only need one [product] and it's going to take years of development anyway, so whether it takes six months to make [the initial product] or six days, in the context of a cure for HIV, I don't think that's going to be ... a defining factor," Gersbach said. As a result, drug developers who have developed a genome editing treatment based on zinc finger or TALEN nucleases are unlikely to switch to using CRISPR/Cas.

Regardless of the specific technologies they use, pharmaceutical scientists also have to worry about potential side effects of therapeutics. Genome editing looks promising now, but the history of biotechnologies suggests that tinkering with biological systems could have unintended consequences. "I think that's also why we'll probably see the first forays into using these technologies in gene therapy in localized-type approaches," Strohl said, explaining that treatments for eye or colon diseases would allow researchers to take a cautious peek at side effects without exposing patients to potentially risky systemic therapies.

The discovery of CRISPR/Cas, even after decades of research in molecular microbiology, also highlights the potential for finding more tools hidden inside cells. "It's really an exciting era that we're into," Mello said. "I think there's still plenty of room for there to be interesting undiscovered mechanisms."

Do animals harbor CRISPR/Cas-like mechanisms to protect their cells against foreign DNA?

What delivery systems will work best to use CRISPR/Cas9 in patients for new genome editing therapies?

To what extent do phage anti-CRISPR mechanisms affect the composition of microbial communities?

How do different CRISPR systems establish tolerance for useful DNA, and what determines which CRISPR system responds to a particular segment of incoming DNA?

Is CRISPR/Cas9 as specific in humans as it is in laboratory models, or does it make off-target changes in the genome?

What protects organisms that do not use CRISPR/Cas for viral immunity from dangerous DNA?