In a previous article, we discussed how CRISPR is not almighty and that organisms can evolve resistance to the changes made to their genome. The simplest method for them to accomplish this is just to mutate the changed gene back to what it was before or just stop its function altogether. If the change made using CRISPR negatively affects the fitness of a species, then natural selection will also naturally lead to it being selected against.
Driving Genes
A solution to this was found, however, in the discovery of gene drives. The basis behind this idea is that there are a number of genes that are more highly conserved and not mutated due to their importance to the organism. This includes many regulatory genes and other systems that run the body. If the inserted genes are placed with these conserved gene clusters, then there is a significantly higher likelihood of them not being modified as well.
Up until recently, the difficulty was being able to target and place the inserted genes in the right spot. And, as you can likely guess by now, all of this was solved by CRISPR and its high precision. Which only becomes better as new versions and more research is put into developing it.
Channeling Mendel
Another primary part to gene drives and CRISPR is how it is being inserted. In the cases where a certain gene is desired to be enhanced or minimized in its effects, there are two copies in general that have to be worried about, relating back to the classical Mendelian idea of inheritance and whether something has two of the same gene copy, homozygous, or two differing copies, heterozygous. While this is generally focused on outward phenotypic traits, it still applies to the genes behind those traits as well.
Thus, to make a proper gene drive, the method is to insert the desired gene and then use CRISPR to disrupt the additional wild-type version. This makes the organism homozygous for only the gene drive gene and thus makes it the only copy that can be passed onto offspring. This essentially guarantees that all offspring will have one copy of the desired gene, at a greater than 90% success rate rather than just the 50% chance in normal cases. This is sometimes referred to as “super Mendelian inheritance”.
The reason why this chance isn’t 100% from the CRISPR modified genome is that, in most cases, the desired gene is meant to have some negative impact on the organism’s fitness. Any loss of this at all has some chance of the organism instead obtaining two copies of the gene from the unaffected parent, becoming homozygous for the wild-type gene. This only happens occasionally though for low fitness impacts, thus the greater than 90% success rate.
There are exceptions, however. If the fitness loss is over 25% (in regards to a population modeling system and the “selection coefficient”) or results in a loss of over 50% to fertility, then the backlash to the gene becomes exponentially stronger. Any gene, by mutation or otherwise, that results in suppressing the CRISPR gene drive gene will be heavily selected for, since it mitigates so much fitness loss. This results in the gene not being spread through a population whatsoever.
Modeling Alleles and Inbreeding
What was unknown was whether this was isolated to just lab experiments or if it also held true out in the wild if CRISPR gene drives were put into effect. Researchers at Indiana University decided to put this to the test by taking genome data from four separate populations around the world of the agricultural pest Tribolium castaneum, also known as the flour beetle.
Their model focused on the effects of inbreeding and standing genetic variation on the population and the gene drive’s success rate. The latter effect is when there are two different alleles within a population located on the same locus in the genome. It has not gone to “fixation” in the population yet where they all have the same gene at that spot.
This is to be expected in basically every case involving gene drives, since we would be deliberately introducing a modified organism with a new gene on the locus into the population that still contains the wild type gene. What is desired is for the gene drive gene to be the one that goes to fixation.
Defensive Repercussions
What the researchers discovered was that basically any sort of mutation in the protospacer adjacent motif (PAM) sequence required for targeting the Cas complex or in the seed region next to the PAM that is responsible for directing cleavage will cause the CRISPR system as a whole to become largely ineffective. This means that the gene drive will fail if such a mutation occurs. Variants that obtain such a mutation are referred to as PAM immune-to-drive (ITD) variants.
Unfortunately, the model also showed that even a fairly mild ITD mutation will drastically reduce the spread of the gene drive through the population. And, even worse, if a population does successfully rid itself of a gene drive, ITD mutations are then more common throughout the population as a whole, making the species more resistant to future gene drive attempts.
Furthermore, even without ITDs existing, inbreeding seems to put a major curb into the spread of a gene drive by forcing homozygous wild type offspring. It has even been suggested that gene drives cause genes related to causing inbreeding to occur more strongly, meaning that the effects shown in the model may be even milder than they would be in reality. The only upside was that the gene drive at least remained active for several generations in the population with only inbreeding involved, before eventually being removed.
Of course, a case with only inbreeding and no ITDs happening is unlikely indeed. The combined effect of them puts a damper on most gene drives having even a chance of being propagated through a population. But that doesn’t mean there is no chance.
Hope For CRISPR Variants
The researchers suggested that a thorough look into the genetic diversity of the population for the genetic locus of interest is conducted before using any gene drives. This is essentially a required step to make sure that the region has low variability. If the variability is high, then a different locus would be preferable, saving a lot of time and effort in direct testing first.
Additionally, new PAM sequences from other version of CRISPR may have a higher success rate. Some have been found to have a different PAM sequence and specificity other than NGG (the main version of PAM used originally from Cas9). A highly likely candidate for use in this regard is Cas12 (previously Cpf1) that also has a different PAM sequence. If the specificity can be matched for whatever target species is being used, then it could bypass all of the problems above by not being subject to mutation issues.
But, at least for current efforts with Cas9 gene drive systems, scientists will have to tread carefully and make sure to triple-check everything before committing to actual usage, since failure could doom any future attempts with the Cas9 gene drive system for that species.
Photo CCs: Tribolium castaneum87-300 from Wikimedia Commons
