Australasian Science: Australia's authority on science since 1938

Gene Drives: A Way to Genetically Engineer Populations

Credit: Mopic/Adobe

Credit: Mopic/Adobe

By Charles Robin

Gene drives occur when a bias in the mechanism of inheritance spreads particular genetic variants through a population. Developments in gene-editing technology now make it possible to construct gene drives that address problems in health, agriculture and conservation.

The concept of gene drives has been around for decades. They occur in natural systems, and scientists have imagined how they might be put to use. Recent advances in gene-editing technology means that “synthetic gene drives” can now be created, and the scope of applications is broad.

It’s time to consider the hazards as well as the opportunities this technology brings, and weigh up the benefits versus the risks. There are also ethical concerns, and scientists must be careful not to leave the community behind.

This issue of Australasian Science presents five articles covering these issues. We will start by explaining what a gene drive is.

In essence, gene drives result from tinkering with the mechanisms of inheritance so that particular gene elements spread through populations. They are a form of genetic engineering but, rather than engineering a single individual, breed or plant variety, they modify free-breeding populations.

There are many good reasons why we would want to engineer populations. Jack Scanlan (p.17) outlines the recent developments of gene drives aimed at altering mosquito populations so that they spread less disease. Alexandre Fournier-Level (p.23) describes how gene drives may be used in an agricultural setting to control weeds and insect pests. Pests can also threaten bio­diversity, and Ella Kelly (p.20) explores the ways in which gene drives might be used for conservation benefits.

But just how do gene drives work? First, let’s consider natural gene drives before turning to the technological developments that enable the construction of synthetic gene drives.

Natural Gene Drives

Under the Mendelian view of inheritance, the organisms of interest have two copies of each gene, one derived from each parent. The two versions are called alleles. These organisms are sexually reproducing, and when the males produce sperm (or pollen) or the females produce eggs (or ovules), Mendel’s law of segregation says that the two alleles will separate into different cells.

Now focus on a female animal and call the two versions of a gene that she carries “allele 1” and “allele 2”. The expectation is that half her egg cells will carry allele 1 and half will carry allele 2. When a fertilisation event occurs, the chance that allele 1 is sampled is 50% – just like flipping a coin. If she has many progeny, approximately half will bear allele 1 and approximately half allele 2.

There is some variation around our expectation of a 50:50 inheritance, the same as if you would not be surprised if you flipped a coin 20 times and got nine tails and 11 heads. However, if the coin was not fair because one side had been weighted, then one outcome would be systematically favoured.

The same can happen in biology. Sometimes inheritance shows non-Mendelian patterns, instead exhibiting “segregation distortion” so that one allele is favoured over others. Gene drives rely on such an inheritance bias.

A classic example of segregation distortion occurs in mice, where the “t-haplotype” allele is preferentially passed on. In this particular example there is a twist to the story because the t-haplotype allele is maladaptive. Mice that carry two t-haplotype alleles have major developmental defects or are sterile. Despite this, a gene drive favours the t-haplotypes when it is first introduced into a population.

Thus these “selfish genes” cheat the inheritance system in some way. In the case of these mice, sperm carrying the t-haplotype allele somehow kill the sperm that do not have that haplotype. Other natural systems tell us that there are various ways to cheat the inheritance system, although many are not completely understood.

Studies of natural systems acting in anomalous ways spark the imagination of scientists and make them ask: “What if?”. They also help us to understand the bounds of what is possible. What if we could deliberately bias inheritance patterns? Could we use such manipulations to drive genes encoding specific functions through populations? Could they be used to deal with pests that spread human disease or wreak havoc in agricultural systems? This last question, at least, motivated the development of theoretical models of gene drives that manipulate the genetic constitution of populations and the abundance of pest populations.

Gene Editing: A Disruptive Technology

Until recently the theory of the spread of gene drive elements through the population was well ahead of what was actually possible. However, two recent advances make synthetic gene drives a realistic prospect.

The first advance is a revolutionary gene-editing technology known as CRISPR/Cas9, which enables targeted alterations to the DNA of complex organisms. Essentially the CRISPR/Cas9 gene-editing technology allows the DNA of an organism to be cut at a determined site while it’s still in the living cells of that organism. The cell will repair the cut DNA. If another piece of DNA similar to that which has just been cut is provided, then that piece of DNA can be patched-in during the repair process. The patched DNA can encode a single nucleotide alteration or several genes.

The reason why the CRISPR/Cas9 technology is so revolutionary is the ease with which the cutting sites can be specified and the simplicity of the necessary components. Only two components are required to create the “scissors” that cut DNA. The first is the Cas9 endonuclease protein, which is an enzyme that uses the second component: a “small guide RNA” (sgRNA) to determine where the cut should occur. The specificity is determined by making use of the ability of the RNA to bind to DNA with the complementary code.

To edit an animal genome, Cas9 protein and sgRNA can be injected into eggs. As this does not have a 100% success rate, many individuals that develop from injected eggs, or more likely the progeny of those individuals, are then screened to see if the desired gene has been altered.

The second advance was the realisation that it is possible to design gene-editing constructs that give themselves biased inheritance. Such constructs are also elegantly simple. They encode the instructions to cut DNA at a place in the genome that will be repaired by themselves.

Specifically, a construct would consist of:

  • a sequence of the target gene to aid the stitching in of the patched sequence;
  • DNA encoding the Cas9 endonuclease;
  • DNA encoding sgRNA that complements the target gene; and
  • more of the target gene to be used as template for repair.

The constructs are also designed so that the Cas9 is active during the development of eggs or sperm.

The first people to design such a construct were not motivated by the prospect of population engineering. Instead they were excited about the prospect of converting heterozygote individuals (who have two different alleles) into homozygotes (both alleles become the same). If the maternally inherited allele could convert the paternally inherited allele then it would, for instance, be much easier to determine gene function, because both alleles would be equally affected.

If this allelic conversion could happen with 100% efficiency then all progeny in the next generation would receive the engineered construct. This would happen again in the next generation, and the next, until the synthetic gene drive spreads through the entire population.

There is one further component that varies depending on the gene drive application. The idea is that the synthetic gene drive construct will also carry “cargo”: a useful gene that the genetic engineer wants to spread through a population of the target organism. For example, a gene that blocks Zika virus replication could be spread through mosquito populations.

It’s important to note that gene drives only work in sexually reproducing organisms, so they will not work on clonally reproducing organisms (e.g. some aphids, many plants). They also work on a time span of generations. Unlike viruses that spread “horizontally” between potentially unrelated individuals in a population, gene drives are only spread down the generations, from parent to child. So gene drives will not work quickly on organisms that take years to be reproductively mature like elephants, humans or trees.

Of the many organisms and contexts where gene drives could conceivably be used, are there hazards that we should consider? Is it ethical to tinker with natural populations in this way?

The rest of this set of articles in this edition of Australasian Science considers these and other questions.


Dr Charles Robin is a Senior Lecturer at The University of Melbourne’s School of Biosciences, and Guest Editor of this series of articles on gene drives.