Another CRISPR paper's just come out, demonstrating a way to deliver the game-changing genome editing system into cells by tissue type - a key development for clinical treatments, so-called 'personalised medicine'.
As Nature News point out in their coverage of the report, while it's ethically far simpler than germline (gamete/embryonic) research, where there's a small number of cells, medicating somatic cells requires considerable targeting effort. Not only have cells already been assembled into more elaborate 3D architectures than the clump of an embryo, there will be only specific cells within tissues with pathologic gene expression of interest to the genetic engineer.
Gene-therapy researchers often harness a virus called AAV [adeno-associated virus] to shuttle foreign genes into mature human cells. However, most laboratories use a gene encoding the Cas9 protein that is too large to fit in the snug confines of the AAV genome alongside the extra sequences necessary for Cas9 function.
Feng Zhang of the Broad Institute of MIT and Harvard in Cambridge, Massachusetts, and his colleagues decided to raid bacterial genomes for a solution, because the CRISPR system is derived from a process that bacteria use to snip unwanted DNA sequences out of their genomes. Zhang’s team analysed genes encoding more than 600 Cas9 enzymes from hundreds of bacteria in search of a smaller version that could be packaged in AAV and delivered to mature cells.
As I mentioned in my previous post on CRISPR/Cas9, Zhang's lab pioneered the technique in 2011.
Staphylococcus aureus, "Staph", well known as a source of food poisoning, features a mini Cas9 (SaCas9) - 1,000 nucleotides smaller than the commonly used Streptococcus pyogenes orthologue (SpCas9).
Speaking to Nature News, biomedical engineer Charles Gersbach of Duke University in Durham, North Carolina, was apparently eager to use SaCas9 in mice as a potential Duchenne muscular dystrophy therapy, a severe X-linked recessive human disease that leads to withering of cytosolic structural protein dystrophin from a defective gene carried by 1 in 3,500 boys.
Perhaps this will be the method that carries CRISPR into the clinic, he says, but it is too soon to tell. “It’s a rapidly developing field,” he says. “There are a lot of things that just haven’t been tried yet.”
The Gersbach lab recently published a really remarkable paper on light-inducible CRISPR-based gene regulation (in Nature Chemical Biology, described in Biotechniques here)
“This method allows targeted, [user]-defined gene activation by light and could contribute to many biomedical applications, such as exploring gene functions and utilizing them to control cellular functions in vitro and in vivo”
Most of these systems use light-inducible heterodimerizing proteins from plants... These systems enable control over expression of any gene in a reversible, tunable and spatially defined manner. However… [targeting] new sequences can be laborious and require specialized expertise. This is particularly a concern for gene activation with systems that must target several sequences in a gene promoter to synergistically achieve robust activation.
To address these limitations, we adapted the CRISPR-Cas9 activator system for optogenetic control. First, as was done previously with TALEs, we fused the light-inducible heterodimerizing proteins CRY2 and CIB1 from Arabidopsis thaliana to the VP64 transactivation domain and either the N- or C- terminus of dCas9, the catalytically inactive form of Cas9 (D10A, H840A). We used the N-terminal fragment of CIB1 (CIBN) and either full-length CRY2 (CRY2FL) or a truncated CRY2 (CRY2PHR) with higher levels of activity. We co-transfected the plasmids encoding one of the two VP64 fusions (CRY2FL-VP64 or CRY2PHR-VP64) and one of the two dCas9 fusions (dCas9-CIBN or CIBN-dCas9) with four gRNAs targeting the human IL1RN promoter7 into HEK293T cells. We used four gRNAs together on the basis of a previous observation that multiple gRNAs are necessary to robustly activate gene expression with constitutive dCas9-VP64 transcription factors. We expected that the gRNAs would direct the binding of dCas9 to the IL1RN promoter in both the light and dark but that VP64 would colocalize with dCas9 via CRY2-CIBN interactions and induce transcription only in the presence of blue light. All of the fusion proteins were expressed well, as determined by western blotting for the FLAG epitope tag. Beginning at 24 h post-transfection, we incubated cells in the dark or under blue light for 3 d and used quantitative reverse transcription PCR (qRT-PCR) to determine expression of IL1RN. The combination of CRY2FL-VP64 with dCas9-CIBN or CIBN-dCas9 showed detectable levels of gene activation in the light above negligible background levels of IL1RN in the dark (P < 0.05). However, the levels of activation were orders of magnitude lower than those for the constitutively active dCas9-VP64 fusion (Fig. 1b), which is typical of systems in which a direct fusion protein is divided into inducible dimerizing fragments.
(a) When expressed by cells, CIBN-dCas9-CIBN localizes to the DNA sequence targeted by the gRNA. In the presence of blue light, CRY2 undergoes a conformational change that enables heterodimerization with CIBN, which causes translocation of CRY2FL-VP64 to the targeted DNA sequence and transcriptional activation of the downstream gene. (b) IL1RN activation in HEK293T cells using four targeting gRNAs; either CIBN-dCas9,dCas9-CIBN or CIBN-dCas9-CIBN; and either CRY2FL-VP64 or CRY2PHR-VP64. Red bars, dark-incubated cells. Blue bars, cells illuminated with pulsing blue light (1-s pulses at 0.067 Hz). (c) Activation of multiple endogenous gene targets was achieved using different groups of gRNAs targeted to HBG1/2, IL1RN or ASCL1 in HEK293T cells. Activation levels in illuminated cells that contained the light-activated CRISPR-Cas9 effector system were statistically similar to those in cells that expressed dCas9-VP64 and the same gRNAs when HBG1/2 or IL1RN were targeted. Conditions not marked by the same letter are significantly different.
"Oncogenic risk" with CRISPR isn't something I recall reading of, but the adenoviruses commandeered by Zhang and co. apparently mitigate the dangers, which are said to arise from host-genome integration events.
The lab ventured out into the known microbial genomes, "in search of smaller Cas9 enzymes for efficient in vivo delivery by AAV... to interrogate and discover additional Cas9 enzymes that are small, efficient and broadly targeting". After getting six candidates together, they tested each orthologue in mammalian cells, and whether they made insertions/deletions in the genome (indels) as directed.
only the one from Staphylococcus aureus (SaCas9) produced indels with efficiencies comparable to those of SpCas9, suggesting that DNA-cleavage activity in cell-free assays does not necessarily predict activity in mammalian cells. These observations prompted us to focus on harnessing SaCas9 and its sgRNA for in vivo applications.
The optimal length of guide RNA was found to be 21-23 nucleotides long in the larger SaCas9 enzyme (probably totally coincidental, but that happens to be the exact same length as miRNAs). For SpCas9
the natural 20-nucleotide guide length can be truncated to 17 nucleotides without significantly compromising nuclease activity, while increasing specificity. Additionally, replacing the first base of the guide with guanine further improved SaCas9 activity.
After aligning 'reads' from protocol that sequenced the broken ends of double-stranded genomic DNA (DSBs) (christened BLESS in Crosetto 2013) to known genomic coordinates, they ran a nearest-neighbour clustering algorithm to identify groups of DSBs (DSB clusters), excavating from this data the Cas9-induced DSB clusters, from a 'background' of DSB clusters (from "low frequency biological processes and technical artefacts, as well as high frequency telomeric and centromeric DSB hotspots").
From the on-target and a small subset of verified off-target sites (predicted by sequence similarity using a previously established method and sequenced to detect indels), we found that reads in Cas9-induced DSB clusters mapped to characteristic, well-defined genomic positions compared to the more diffuse alignment pattern at background DSB clusters. To distinguish between the two types of DSB clusters, we calculated in each cluster the distance between all possible pairs of forward and reverse-oriented reads (corresponding to 3′ and 5′ ends of DSBs), and filtered out the background DSB clusters based on the distinctive pairwise-distance distribution of these clusters.
Third, the DSB score for a given locus was calculated by comparing the count of DSBs in the experimental and negative control samples using a maximum-likelihood estimate22 (Supplementary Discussion). This analysis identified the on-target loci for both SaCas9 and SpCas9 guides as the top scoring sites, and revealed additional sites with high DSB scores.
Extending the analysis of the scored results, they found new off-target sites that would be affected by the genome-editing technique (i.e. side-effects, which need to be characterised and minimised for clinical therapy).
These new off-target sites include not only those containing Watson–Crick base-pairing mismatches to the guide, but also the recently reported insertion and deletion mismatches in the guide:target heteroduplex.
Together, these results highlight the need for more precise understanding of rules governing Cas9 nuclease activity, a requisite step towards improving the predictive power of computational guide design programs.
They went on to test it in vivo, with an hepatocyte-targeting AAV conduit harbouring mouse apolipoprotein (Apob)-targeting guide RNAs in its mini SaCas9 casette.
Seeing the liver lipids rise, they hurried on to knocking down a further gene: proprotein convertase subtilisin/kexin type 9 (Pcsk9), the target of a promising class of heart-protecting inhibitors.
Fearing their efforts may make the animals sick, they checked for signs of toxicity or protests from the immune system — none were found.
We observed a slight trend in aspartate transaminase (AST) increase across all cohorts atfour weeks, including the uninjected animals. The elevated levels did not exceed the upper limit of normal and is not indicative of hepatocellular injury in animals (Fig. 5b). However, a larger cohort study should be conducted to further evaluate the potential side-effects of Cas9-mediated in vivo genome editing. In addition, the differences between mouse and human immune responses need to be better elucidated before considering this approach for therapeutic applications.
For more CRISPR goodness, check out my previous post on its background, the mutagenic chain reaction, and rumblings of a woolly mammoth de-extinction
