***This is a first draft of an essay written in April 2019 for the Biochemical Society***
You may have heard of synthetic genomics. This headline grabbing, high-profile, big science topic is starting to emerge catalysed by the pioneering work of famous names in synthetic biology and biotechnology like George Church and Craig Venter. But what is synthetic genomics and what it is being used for? As a prominent researcher at a recent UK meeting said - “Is it just synthetic biology with bigger bits of DNA?” Well no, not quite...
Synthetic genomics and synthetic biology
Synthetic genomics certainly owes a lot to synthetic biology, making use of much of methods resources and jargon involved, yet it differs in crucial ways too. Synthetic biology - a major new interdisciplinary subject that has emerged this century - is focused on rewriting and reprogramming the DNA in cells using cycles of design and engineering to get improvements. It aspires to do this using engineering principles like modularity and standardisation that enable researchers to more quickly get to the ultimate aim of tailoring cells as technologies for specific tasks. Right now synthetic genomics lacks these formalities, as the goal of the work is not to optimise one cell behaviour over the rest but to produce a new understanding of DNA and biology, either directly or by enabling new experiments that can’t be done any other way.
New understanding of biology has already been one of the main outcomes of two decades of synthetic biology research. By building up synthetic gene systems from first principles scientists are better able to understand and mathematically model the key factors that define important networks and pathways where genes interact together in cells. Proponents of this aspect of synthetic biology often use a famous quote from physicist Richard Feynman “What I cannot create, I do not understand” which concisely postulates that the best way to learn about something works is by trying to build it. Indeed, wanting to determine the minimal requirements required for cells to genetically encode memory and rhythms led to the first significant achievements in synthetic biology; synthetic gene circuits that act as switches and oscillators.
Twenty years since these first steps in synthetic biology, academic labs and biotech companies around the world now use synthetic DNA to build a lot bigger than just systems of two or three genes (Figure 1). It is becoming routine to see dozens of genes used in synthetic DNA constructs for various tasks, and so naturally the cell’s own operating system - it’s genome - is increasingly within our sights. However, the true synthetic biology version of a synthetic genome, a genome designed and built from first principles from a kit of modular parts, is still a long way off, looming as a grand challenge that could even take another couple of decades to achieve. Right now we simply don’t know enough about all the genes and genetic regulation that is required to direct a cell to grow and perform a cell cycle, and so we cannot yet write a genome from scratch. The task also gets more complex by the day as researchers in cell and genome science continue to uncover new unexpected ways that DNA encodes regulation and function that will need to be taken into account.
So for now and the near future synthetic genomics is best placed to help us understand what we do and don’t know about cell biology and especially how the genome encodes an organism. Constructing and testing synthesised genomes and chromosomes that are increasingly different compared to natural genomes enables us to test our current understandings of genome biology, whilst also developing the methods and tools to one day build custom genomes to design. Most synthetic genomics projects right now are therefore aimed at delivering new knowledge of genome coding, content and organisation - aspects that are hard to be determined by other approaches . By tackling these interesting questions using a new synthetic approach to genome manipulation, these projects both push and pull the development of new technologies that one day will enable broader use of synthetic genomics within research or within applied synthetic biology.
A decade of synthetic genome progress
Impressively in just over 10 years, synthetic genomics efforts in bacteria have already advanced what is possible by several steps (Figure 2). In 2008 a full copy of a 580,000 bp Mycoplasma genome was constructed from chemically-synthesised DNA, and then in 2010 the same team showed that a synthetic copy of a 1 Mbp Mycoplasma genome could replace a natural genome and support the growth and division of a cell. This landmark work by the J. Craig Venter Institute gave us the first cell with a synthesised genome, albeit one with no major changes to its DNA sequence - it simply showed that synthesis and construction was possible.
In 2013, a team from Yale and Harvard then showed that a bacterial genome could be ‘recoded’ by using site-specific mutation (not genome synthesis) to remove all 321 occurrences of the rarest codon used in protein synthesis in E. coli. This Genomically Recoded Organism (GRO) now differed from almost all of the rest of natural biology in not using the same 64 codons in its genes to direct which amino acids are used to makes its proteins. It now only used 63, and so the spare codon in this cell could be reassigned to make E. coli add non-standard amino acids into proteins - a feature useful for both research and biotechnology applications.
While altering only 321 bases in a 4.6 million bp genome may seem like a minor change, this work showed that genomes could be made with recoding throughout their genes, changing the DNA that encodes the proteins without altering the protein itself. UK and US teams are now pushing to produce E. coli and Salmonella bacteria with substantially more DNA recoding in their genes and more codon reassignment, in all cases now doing it by constructing the recoded genomes from synthesised DNA, rather than by mutation.
The next step that bacterial synthetic genomics has taken beyond recoding is in genome minimisation. In 2016, the J. Craig Venter Institute constructed a synthetic, redesigned version of their 2010 Mycoplasma genome, leaving out the genes and DNA that they deemed not to be essential for growing this cell in the lab, which amounted to roughly half of the genome. No recoding of genes was done in this work, and where DNA remained it was the equivalent to its natural sequence. However, this achievement represents our ‘most synthetic’ genome to date as it has such huge differences in its gene content and layout compared to its natural equivalent.
Interestingly, in this minimised genome project the team tried pushing their work to an even further step towards the long term goal of a fully modular synthetic genome. As they synthesised and constructed their minimised genome, they also made a version where the order and layout of the remaining genes on the bacterial genome was totally changed, with the genes now arranged along the chromosome according to function. The team called this version ‘defragmented’, making an analogy to the process where computer files in a hard drive are relocated to common clusters to improve storage efficiency. For a 1/8th segment of the genome this defragmented design could replace its natural equivalent, but for the rest of the genome it could not. This tells us that the layout and order of the genes in the genome play a crucial role into whether they work correctly - revealing important new information on ‘genome design rules’ that will need to be considered in future efforts to construct custom genomes from modular DNA parts.
Synthetic genomes beyond bacteria
A synthetic genome for a eukaryote has yet to be realised, but the international synthetic yeast genome project (‘Sc2.0’) is rapidly approaching that goal by having a community of research groups around the world build synthetic chromosomes to a common new design. The Baker’s yeast S. cerevisiae has an 11 million bp genome naturally split into 16 different chromosomes, and synthetic versions of 7 of these have now been completed. The design of the synthetic genome includes gene recoding and some minimisation too, via the removal of unneeded non-coding elements such as transposons and introns. It also has an element of defragmenting as all tRNA genes are being removed from their normal locations in the main chromosomes to be now placed on a new synthetic tRNA chromosome. The Sc2.0 genome also has an inbuilt design feature that means further minimization and gene rearrangement can be done when desired. This is achieved by an inbuilt system called ‘SCRaMbLE’ where genes within the synthetic chromosomes can be randomly removed and rearranged inside the living yeast cells when they are given a specific chemical stimulus. Theoretically, continued SCRaMbLE of the complete Sc2.0 genome inside yeast growing in lab conditions would eventually led to a genome only containing the required genes for lab-based growth, and with these genes in a new layout that enabled this genome to function well.
While SCRaMbLE is not a direct way to remove or relocate large portions of the genome as desired, it still provides a powerful method to explore what genes are essential for a cell in various conditions and what gene order and genome arrangements are tolerated (and which one aren’t). Already work with SCRaMbLE on the completed synthetic yeast chromosomes has shown that the yeast genome can handle some serious rearrangement of its genes without many problems. Two teams have also shown that the 16 chromosomes of yeast can also be fused together so that the genome of yeast can be put on only 2 chromosomes with the cell functioning just fine. The whole genome can even be completely placed on just a single chromosome and still power a growing cell, albeit one that grows slower than usual. Clearly, there is significant plasticity in the chromosomal structure and gene layout in the yeast genome, which is a clue that eukaryotic genomes may in the end be more amenable than those of bacteria for the next steps for synthetic genomics, such as full genome reorganisation and ultimately modular design and construction.
So it seems after only ten years of synthetic genomics that genome recoding, genome minimisation and large-scale synthetic chromosomal reorganisation are all possible both in prokaryote and eukaryote microbes. These efforts are redefining how we think about genomes and the relative (lack of) importance of naturally evolved sequences, gene content and layout. We now know that cells can happily exist outside of nature’s standard genetic code where the same 64 codons encode the 20 amino acids of all proteins, and we’ve proven that genomes have no need to host transposable elements despite their ubiquity. Chromosome layout and content can be altered far beyond what we see in natural variation within species, but so long as key genes remain and are appropriately regulated, then cells are still viable and can even grow just fine.
These recent and ongoing advances all help towards the next major goals for synthetic genomics, which are to make viable, fully-refactored genomes and eventually realise completely modular genomes that are built-to-design from standard parts. At that point synthetic genomics would indeed return to being an engineering discipline like that of synthetic biology, where engineering tools (design and construction automation) and engineering principles (modularity, standardisation) can be used to accelerate and industrialise the work of making cells as technologies.
And while teams work towards achieving these goals in model microbes, the technologies for doing synthetic genomics can also benefit research elsewhere. For example, in more complex organisms like humans and mice, biomedical research is continually seeking to better understand how DNA sequence and the organisation of regulatory regions is important in determining gene function and how mutations lead to pathogenicity. The same tools and methods used to design and make megabase chromosomes for microbes can be used to recode and reorganise similar-sized regions within mammalian genomes, providing a new way to ask and answer questions on genome biology. Being able to synthesise, rearrange and relocate big DNA into mammalian genomes is now just beginning as a new approach to explore how the content and organisation of the large stretches of non-coding sequence (‘the dark matter of the genome’) are involved in the regulation and correct functioning of genes and cells. Via big DNA design and synthesis, researchers can make and test synthetic variants of important genomic loci, like regions containing key genes associated with cancer or development. They can then learn how the sequences, features and arrangements in these loci define how they work, helping to better understand how our own genomes function and how mutation in them can lead to diseases.
It’s early days, but if this ‘learn-by-building’ approach with big DNA pays off, then synthetic genomics in complex organisms may well become mainstream more quickly than we think. And so while making synthetic human genomes seems decades away right now when synthesising genomes a thousand times smaller is still an expensive and lengthy challenge, we need to be wary that technology in this area may well accelerate much faster than we think. Are we prepared for synthetic human genomes anytime soon? Custom-built synthetic microbial genomes is one thing, but the notion of synthetic human genomes raises many more pressing questions. They are not reality now, but very well may be real issues within in our lifetimes. It is therefore important that concerted efforts are made to engage widely, discuss and coordinate globally how synthetic genomics will advance over the next decades. These efforts have already begun, spearheaded by an international community of interested researchers, social scientists, engineers, lawyers and citizen science advocates who have formed the GP Write consortium. With community oversight, it is hoped that synthetic genomics can not only advance quickly to the benefit of science, but also advance safely to the benefit of society.
Further Reading
Hutchison, C.A. 3rd, Chuang, R.Y., Noskov, V.N., et al. (2016). Design and synthesis of a minimal bacterial genome. Science 351(6280):aad6253. doi: 10.1126/science.aad6253
Richardson, S.M., Mitchell, L.A., Stracquadanio, G., et al. (2017). Design of a synthetic yeast genome. Science 355(6329):1040-1044. doi: 10.1126/science.aaf4557
Wang, L., Jiang, S., Chen C., et al. (2018). Synthetic Genomics: From DNA Synthesis to Genome Design. Angew Chem Int Ed. 57(7):1748-1756. doi: 10.1002/anie.201708741
GP Write website containing information about the GP Write consortium, pilot and planned projects and roadmapping documents from the consortium working groups - https://engineeringbiologycenter.org/
Meeting report from the UK meeting “Synthesising a Human Genome: What could go right?” held September 2018. http://www.synbicite.com/media/uploads/files/Synthesising_a_Human_Genome_Meeting_Report.pd
Commentaires