Gene drives have the potential to solve intractable problems in diverse areas of public health, agriculture and conservation but also present a range of social and environmental hazards. It is vital that the use of technology is open and peer reviewed, with research intentions made clearly transparent to the public. The Academy recommends scientists adhere to best scientific practices and follow the responsible conduct of research when investigating gene drive modified organisms.2 Ethical consideration of both social and environmental consequences should be considered prior to commencing any research. The National Framework of Ethical Principles in Gene Technology 2012 provides guidance on values and ethical principles in relation to gene technologies.
Such considerations should include a thorough and quantitative investigation of alternative methods to address the experimental problem. Not all problems that can be addressed by a gene drive modified organism should be: if there is an alternative available that will achieve the same outcome while presenting fewer hazards then it should be prioritised over new technologies. On the other hand, if a synthetic gene drive is the best solution it should be considered to prevent the consequences of inaction or ineffective action.
Multiple stringent confinement strategies should also be used to avoid the unintentional release of a gene drive modified organism while in development (Akbari et al., 2015; Oye et al., 2014). Molecular and physical confinement measures are described below in addition to possible safeguards that may be prepared in advance of a gene drive release.
There are a number of options which can be considered during the design of a gene drive construct that can act as a molecular confinement measure. These include:
using synthetic target sequences that are not in natural populations and therefore could not spread to wild organisms
targeting unique sequences which are very specific to the target organism to avoid a gene drive spreading to closely related species. For example targeting the toxin genes of cane toads which are not found in other amphibians
choosing a gene drive mechanism which has a low ability to spread, known colloquially as high threshold drives—these help confine the spread of a gene drive to a local breeding population. If the threshold is not exceeded, the drive system is lost from a population. This concept is illustrated by the loss of Wolbachia from natural populations (Nguyen et al., 2015)
designing a gene drive which is not self-sufficient by physically separating the elements. In the case of CRISPR/Cas9 drive technology the Cas9 and guide RNA would be separated, known as a split gene drive system. This has been tested in yeast (DiCarlo et al., 2015)
designing a gene drive that would stop after a few generations. This would limit the capacity of the gene drive to spread. Figure 3 demonstrates this ‘daisy chain’ gene drive where each genetic element drives the next (Noble et al., 2016).
Appropriate training of researchers in best practice and using precautions to limit human errors are very important. Other physical measures which can be implemented include:
following the specific guidelines for work on mosquitoes as outlined within The guidance framework for testing genetically modified organisms (WHO, 2014)
avoid transferring gene drive modified organisms between laboratories. Instead DNA constructs or information sufficient to reconstruct the gene drive should be sent, if required
ensuring that all work takes place in suitably confined premises as currently defined by Physical Containment levels PC23 or PC34 (Office of the Gene Technology Regulator) or Biosecurity Insectary Containment levels BIC25 or BIC36 (Department of Agriculture and Water Resources).
In addition to the containment measures described above, a strategy to mitigate potential ecological and environmental consequences from the accidental release of a gene drive or from unanticipated impacts of an intentional release is highly recommended. Options include:
an immunisation gene drive to block the spread of unwanted gene drives by pre-emptively altering the target sequence thereby preventing the gene drive from spreading (Esvelt et al., 2014)
a reversal gene drive designed in parallel with any gene drive experiment to overwrite any unwanted changes of a gene drive (DiCarlo et al., 2015)
trialling a gene drive using a benign change to enable the effectiveness of a gene drive spread to be studied prior to a release
ecological modelling to help predict the potential consequences resulting from a gene drive release (for example, see Unckless et al., 2017).
Wherever possible, the likely effectiveness of safeguards should be assessed in a quantitative way based on current knowledge.