Biology Top Ten: A Crisper View of CRISPR

January 2015

No apologies for returning to the topic of CRISPR, a technology that underpins some of the hottest Research Fronts identified by Thomson Reuters Essential Science Indicators and which we wrote about in October. Three of the papers that define the core group of CRISPR papers are still young enough to dominate the Top Ten (which is based on papers published in the past 24 months and cited over a recent two-month period) at #3, #4 and #9, while new papers from the four research groups that essentially defined CRISPR are nipping at their heels. But first, a recap.

CRISPR/Cas allows molecular biologists to edit genes with unprecedented accuracy, offering the promise not only of greater understanding of the genetic machinery but also, it is hoped, the ability to fix that machinery much more effectively when it goes wrong. The technology is derived from a bacterial immune system that protects its owners from infection by foreign DNA. There is a nuclease, Cas9, which an activating RNA sequence guides to break the invading DNA at a site identified by another short piece of RNA cut from a longer sequence that features Clustered Regularly Interspaced Short Palindromic Repeats. In the lab the two pieces of RNA—the activating tracrRNA and the aiming crRNA—can be joined into a single guide sgRNA to mimic the natural hybrid formed in bacterial cells.

The four new papers, published back to back in a single issue of Nature Biotechnology, all pursue one of the current drawbacks of CRISPR: it isn’t quite as accurate as it ought to be. The pattern matching that guides the crRNA to the DNA target seems to be relatively tolerant to mismatches, and that means that in addition to homing in on the specific gene sequences researchers are interested in, the tool can also go off and inflict random damage elsewhere in the genome. How much collateral damage, and how to reduce it, is the focus of the four papers.

What’s Hot in Biology

Rank Paper Citations This Period (Jul-Aug ‘14) Rank Last Period (May-Jun ‘14)
1 ENCODE Project Consortium, “An integrated encyclopedia of DNA elements in the human genome,” Nature, 489(7414): 57-74, 6 September 2012. [85 institutions worldwide] 110 1
2 1000 Genomes Project Consortium, “An integrated map of genetic variation from 1,092 human genomes,” Nature, 491(7422): 56-65, 1 November 2012. [115 institutions worldwide] 87 2
3 L. Cong, et al., “Multiplex genome engineering using CRISPR/Cas systems,” Science, 339(6121): 819-23, 15 February 2013. [10 US and China institutions] 83 3
4 M. Jinek, et al., “A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity,” Science, 337(6096): 816-21, 17 August 2012. [U. Calif., Berkeley; U. Vienna, Austria; Umea U., Sweden] 48 6
5 S. Djebali, et al., “Landscape of transcription in human cells,” Nature, 489(7414): 101-8, 6 September 2012. [21 institutions worldwide] 38 7
6 Y.F. Fu, et al., “High-frequency off-target mutagenesis induced by CRISPR-Cas nucleases in human cells,” Nature Biotechnology, 31(9): 822, September 2013. [Massachusetts Gen. Hosp., Charlestown, MA; Harvard U. Sch. Med., Boston, MA]  37 +
7 L.B. Alexandrov, et al., “Signatures of mutational processes in human cancer,” Nature, 500(7463): 415, 22 August 2013. [55 institutions worldwide] 34 +
8 J. Seok, et al., “Genomic responses in mouse models poorly mimic human inflammatory diseases,” PNAS, 110(9): 3507-12, 26 February 2013. [21 North American institutions] 33 10
9 H.Y. Wang, et al., “One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering,” Cell, 153(4): 910-8, 9 May 2013.  [Whitehead Inst., Cambridge, MA; MIT, Cambridge, MA; Broad Inst., Cambridge, MA] 33 8
10 J.A. Tennessen, et al., “Evolution and functional impact of rare coding variation from deep sequencing of human exomes,” Science, 337(6090): 64-9, 6 July 2012. [10 US institutions] 31 +

SOURCE: Thomson Reuters Web of Science

NB. Only papers indexed by Thomson Reuters since September 2012 are tracked. The “+” sign indicates that the paper was not ranked in the Top Ten during the last period. In the event that two or more papers collected the same number of citations in the most recent bimonthly period, total citations to date determine the rankings

Two of the papers adopted almost identical approaches. J. Keith Joung’s group at Massachusetts General Hospital and Harvard, at #6, and Feng Zhang’s group at the Broad Institute of MIT and Harvard, just off the Top Ten at #12, systematically investigated the effect of introducing deliberate mismatches to the target-recognition sequence (P.D. Hsu, et al., Nature Biotechnology, 31[9]: 827, September 2013; 29 citations this period). Some of the changes were single substitutions. Others involved two or more, adjacent or separated to varying extents. They then looked to see how the changes affected both the target gene and other potential target sequences, and the results are both complex and illuminating. Off-target effects depend on the number of mismatches, their position, and their distribution along the target sequence.

Hsu et al. also looked at ways to improve the specificity of CRISPR-mediated mutagenesis. One starting point was the observation that shorter crRNA sequences sometimes failed to break the DNA at sites that were broken when the crRNA was hybridized to the tracrRNA. Zhang’s group created sgRNAs that differed in the length of the tail on the tracrRNA, and found that the longer tails could result in five-fold higher mutation rates at the target. Other changes they investigated included lengthening the target sequence and examining the whole construct to ensure that there were no sequences that could accidentally stop transcription. Finally, they looked at how the amount of Cas9 and sgRNA affects specificity by measuring the ratio of on-target to off-target DNA breaks. A roughly ten-fold dilution improves specificity four times, and a further ten-fold dilution improves specificity even more. The problem is that at these very low dilutions, the actual number of on-target breaks is also substantially reduced.

FIVE SIMPLE RULES

Having examined the effect of all these different variables on performance, Zhang’s group came up with five relatively simple rules to optimize the efficiency of CRISPR/Cas. Other researchers can make use of them thanks to an online CRISPR design tool to make their system as good as it can be.

Joung’s group, having produced very similar results, puts the emphasis on how bad things might be. They found that mutation rates at some off-target sites were even greater than at the on-target site, and point out that because some CRISPR/Cas constructs can mutate even off-target sites differing by up to 5 bases quite efficiently, choosing targets by counting mismatches is unlikely to be effective because thousands of such sites will exist across the genome. They also disagree over the effect of reducing the concentration of the reagents, which “did not appreciably change the relative rates of off-target mutations.” The two dilution studies differed in their details, however, and these discrepancies may yet be resolved.

For Zhang, it seems, the specificity glass may be half full, while Joung sees it as half empty. For pure research, the first may be more relevant, while for the therapy, the second may be more appropriate. And pessimism may explain why Joung’s paper, at this writing, has 206 citations to Zhang’s 161. It does not, however, explain why the other two papers published at the same time and in the same journal are (relatively) neglected.

STAYING ON TARGET

With 126 citations to date, another of the four Nature Biotechnology papers, by George Church’s group at Harvard Medical School, uses a very different method to assess in detail the off-target specificity of CRISPR/Cas, comparing it favorably to TAL-effectors, another relatively new gene editing system (P. Mali, et al., p. 835; 21 citations during this bimonthly period). And the Church group’s approach to minimizing off-target activity is to build constructs that make use of several parts that work together cooperatively to ensure that only the desired target receives the system’s full attention.

Rounding out the Nature Biotechnology quartet is a paper by Jennifer Doudna, of the University of California, Berkeley, one of the pioneers in CRISPR, and David Liu of the Howard Hughes Medical Institute at Harvard, and their colleagues (V. Pattanayak, et al., p. 833; 14 citations this period, 100 overall). This report, like Fu et al. and Hsu et al., used a combinatorial approach to see how changes in the target-recognition sequence affected off-target activity, with very similar results. And, like Zhang’s group, they find that higher concentrations of the Cas9-sgRNA complex will cleave off-target sites that are not attacked when concentrations are lower, reducing specificity. They also show that a shorter guide sequence is more specific but less active, which could offer another avenue to lower off-target effects.

All four papers, while they may differ in some of their particulars, contribute to a substantially improved understanding of CRISPR/Cas and point the way to substantially improved efficiency of the technology. So, do the different citation rates of the four papers have any deeper meaning? Probably not. The field is very young and expanding very rapidly, and it is probably all a researcher can do to keep up, let alone make sure they’ve cited all the required precedents.


Dr. Jeremy Cherfas is a science writer based in Rome, Italy.

The data and citation records included in this report are from Thomson Reuters Web of ScienceTM. Web of ScienceTM is a registered trademark of Thomson Reuters. All rights reserved.