Ever wanted to cut a bad thing out of your life? Well, perhaps the secret to success doesn’t involve any cutting (of your DNA). In some people, a mutant PCSK9 gene interferes with the liver’s job to get rid of LDL, aka “bad” cholesterol. Now, researchers have developed an all-in-one “evolved engineered transcriptional repressor” (EvoETR) that reduced circulating levels of Pcsk9 in mice for weeks without the potential safety issues associated with cutting DNA with CRISPR-Cas9.
The PCSK9 protein controls the number of LDL receptors that are on the surface of liver cells by degrading some of them. But over-ambitious, gain-of-function PCSK9 mutant proteins in hypercholesterolemia patients don’t let enough receptors reach the surfaces of liver cells, leaving too much LDL in the blood.
So, how do those with a PCSK9 mutation cut their bad cholesterol levels?, CRISPR-Cas9 can calm down mutant PCSK9, but because it cuts DNA, it could also lead to mutations and off-target effects. Another option is to perform base editing, changing one base as Verve Therapeutics is testing in clinical trials. It doesn’t cut DNA, but it requires a guide RNA, an mRNA, and a modified Cas.
Some epigenome editors, like the hit-and-run and CRISPRon, CRISPRoff, use a dead Cas9 (dCas9) to bind—but not snip—DNA. Instead, the editors change epigenetic modifications, such as methylation. All-in-one editors with an effector domain made up of transcriptional repressors and a DNA-binding domain work well in vitro, at least in the short term, but they haven’t been tested in vivo.
So, Angelo Lombardo’s team (IRCCS San Raffaele Scientific Institute) designed EvoETRs as a one-stop solution to calm down Pcsk9 long-term in mice with only a transient expression via lipid nanoparticle delivery. But first, they had to optimize the parts. Here’s what they did:
- For their ETRs, they chose KRAB, the catalytic domain of DNA-methyltransferase A (cdDNMT3A), and its cofactor DNMT3-like (DNMT3L) as the effectors
- In a mouse hepatoma line, zinc-finger proteins (ZFPs) worked better with an effector than dCas9 and transcription activator-like effectors (TALEs) as the DNA-binding domain when silencing Pcsk9
- Epi-silencing lasted as along as CRISPR-Cas9 silencing up through day 28—the last day of one experiment – making this a hit-and-run method
- ZFP-ETRs reduced Pcsk9 expression by about 30-fold—similar to Pcsk9 disrupted with CRISPR-Cas9
- ZFP-ETRs methylated DNA at the target site CpGs and downregulated Pcsk9, but had some effects at genes far away—they downregulated eight additional genes and resulted in 18 other differentially methylated regions, aside from the three in Pcsk9
- The team found that the ETR effects were stable and heritable after they removed part of the liver, which prompted the liver to regenerate and make new cells
Finally, to simply things, they made all-in-one EvoETRs by putting a ZFP between KRAB and the DNMT3 pieces, connected with linkers. Here’s what they found when they tested the constructs:
- The most specific ZFP they tested was ZFP-8, and it only downregulated four genes, including Pcsk9, and Pcsk9 was the most significantly affected gene in mouse cells
- When delivered into mice, EvoETR-8 reduced Pcsk9 levels by 75% through day 43, similar to the effects of CRISPR-Cas9, though the triple ETR (which was not an all-in-one construct) reduced levels only by half
- EvoETR-8 also methylated more CpGs than the triple ETR
- In addition, LDL and total cholesterol levels decreased significantly
- Whereas CRISPR-Cas9 induced mutations at the Pcsk9 site, there were no such changes in mice treated with EvoETR-8
Overall, the team says that epi-silencing with EvoETRs can be durable and heritable in vivo, even after one year of follow-up, without worrying about causing DNA mutations. They say epi-silencing could also be reversed, so that treatment could be stopped if there are unintended effects.
Check out this hit-and-run epigenome editing method at Nature, February 2024.