Week 11: Evolution, CRISPR Gene Drives, and Ecological Engineering


In this class, Kevin Esvelt from the Wyss Institute, (now faculty at MIT Media Lab) talked about CRISPR genome editing, how it enables us to build gene drives, and raised the questions of whether, when, and how we should develop synthetic gene drive systems to address real-world ecological problems.

Abstract: Evolution is the central paradigm of biology, but one of the least appreciated in bioengineering. Electrical circuits do not evolve; genetic circuits do - and usually break. Human-designed technologies minimize interactions to enable modularity; genetic circuits evolved to promote a complex web of interactions that inhibit modular design. These factors cause problems for bioengineers in the laboratory, but even more so in the wild.

Because wild organisms have been selected for efficient reproduction in their ancestral habitat, altering them almost always decreases reproduction. Hence, releasing engineered or selectively bred organisms into the wild has little if any lasting impact because natural selection weeds out the human-made changes.

Not all genes are so thoroughly tied to their host organism. Gene drive occurs when a DNA sequence ensures that is inherited more often than normal. Any sequence that causes gene drive can spread itself - and nearby DNA sequences - through populations even if their collective impact makes each organism less likely to reproduce.

Fig. 1: Kevin Esvelt at EMW StreetBio (top) and proposed gene drive system (bottom)!

Creative Homework Assignment

Question: What features would you want to see in an online discussion platform devoted to guiding the development of gene drives? Assume that the researchers involved in the project are interested in soliciting public feedback before and during experiments so that they can better identify problems and redesign the technology. Please give specific examples of already-existing elements – if you want a discussion forum, should it be more like reddit, Quora, or something else? What should moderation be like? What would it take to get you to regularly participate in such a community?

Gene drives is a technology that could potentially alter whole ecosystems and affect directly human life. While the underlying principle of making a successful inheritable gene drive system is (almost) straightforward, it needs multi-modal biological and ecological analyses to achieve long term efficacy without catastrophic implications to the affected environment. In order, though, to have substantial discussions on the subject at the level of the society, we have to take into account that the average individual is not able to understand the complex scientific notions addressed by the experts in life sciences. Thus, the number one feature I want to see in such a platform would be the facilitation of interactions between scientists and the general public, in the form of disseminating fears and answering questions on the workings of a gene drive.

Imagining this online platform, I think of an amalgam of various current social networking platforms. I can see two types of profiles, maybe distinguished by different colors in the profil picture frames, one for the scientists and one for the non-scientists. The scientist profiles, should have a complete CV, with publications and current academic status, such that whatever they write can be backed by their experience. There could be also a point system, where a scientist gets points for answering questions -- 1 points if it was from another scientist, 2 points if a non-scientist asked it.

to be continued

Design Homework Assignment

Question: Identify a problem that could be addressed using a CRISPR gene drive. Which organism would you target and how would you alter it? Why is a gene drive a good solution relative to other options? What could go wrong? Don't go into detail, but list several possibilities. Who should be involved in the discussion of whether to consider this application? Design a basic but evolutionarily stable gene drive that should function in your organism.
Your goal is to design a proof-of-principle experiment that will determine whether CRISPR gene drives can function efficiently in the target organism. Your drive system should not cause population suppression, carry any 'cargo' genes, or change the sequence of any protein produced by the organism. It should only spread itself.

The Gypsy Moth (Lymantria dispar) is an important economic pest that causes large-scale damage to agricultural crops, forests, and humans worldwide. That's why it has drawn the attention of changing its olfactory receptors as a form of pest management. Thus a potential application for gene drives is an olfactory receptor gene suppression that would alter the species odorant binding affinities at a population level to no longer be able to detect its target crop. Therefore, the gypsy moth can exist in its natural ecosystem carrying out its ecological niche without acting as a pest to agricultural crops.

However, many things could go wrong without careful design as the Gypsy moth uses its olfactory receptors also for mating, which means that altering the target odorant repertoire could putatively render them unable to detect their mates. In that case, the gene drive would not be effective, but fortunately the effect will not be catastrophic (suppress the population) as by inhibiting mating we are essentially stripping the ability of the drive to be passed to the next generation.

Fig. 2: The Gypsy Moth (Lymantria dispar)

1) Identify a gene that is important for fitness in your target organism. Why must gene drives target genes important for fitness? A literature search may be required. You can use NCBI for information on genes if you wish, though Wormbase is generally superior and easier to use.

The gene we choose to target is the OrCo gene (sequence shown below from NCBI), which stands for olfactory receptor co-receptor. Knockout of Ors has revealed that OrCo proteins are required for insect olfaction. The gene must be important for fitness because if the gene makes the organism less competitive, then the organism will most likely die before propagating the gene to next generation.

Source: Identification and Knockdown of the Olfactory Receptor (OrCo) in Gypsy Moth, Lymantria dispar

Fig. 3: The target OrCo gene!

2) Find CRISPR target sites in the 3' end of the gene using an online tool. For C. elegans and many other organisms, a good user-friendly tool is GT-Scan. Recommended parameters: set the high-specificity mismatch limit to 3 and leave all other settings at default values. More exotic species require less user-friendly software. For bonus points, use sgRNAcas9, which allows you to analyze any downloaded genome sequence in fasta format. Warning: it's not particularly user-friendly. List the relevant criteria for inclusion of each target sequence (e.g. potential off-targets etc) as provided by the program you used.


We will answer this portion of the homework as a hypothetical. If the genome was hypothetically available, we would design the CRISPR target sites based on the sgRNAcas9 software (since this species is not available on GT-scan) -- we run the OrCo gene against the gypsy moth genome and design a GUIDE RNA that does not exist anywhere else in the moth genome. This ensures specificity. Also, it is important to include the target sequence at the 3' end of the OrCo gene.

Alternatively, we started a new search beginning with a common pest whose genome is sequenced. We found the Fall Armyworm (Spodoptera frugiperda), seen in this paper :: A draft genome assembly of the army worm, Spodoptera frugiperda. However, as the paper explains, the genome has not been annotated, at least not for olfactory receptors. What we did was run a BLAST search (http://blast.ncbi.nlm.nih.gov/Blast.cgi) of common known olfactory receptors against the sequenced armyworm. We found around 10 olfactory genes with around 95% match to the armyworm genome. From this data we can then run a GT-scan and design the proper guide-RNA and point of insertion near the 3' end.

3) Would it be possible for you to safely build this gene drive in your current laboratory? Which confinement strategies and safeguards would you use? See Akbari et al. and especially here for recommended confinement strategies.

It is definitely not safe for us to build the gene drive -as is- in our current laboratory as pests like Gypsy moth can be found all over the US and can be transferred accidentally on crop transportation. As a starter, we would use extrinsic confinement methods.

Moths can fly, so we would start with a barrier confinement made by strong air curtains on the doors of the laboratory and mandatory inspection of the lab clothing prior to exit for moths. The moths will be kept in multiple nested containers that are only ever opened when their occupants have been anesthetized.