Weeding the Gems

Guest post by Adam Talbot and Elspeth Ransom who work at the University of Warwick with GARNet Advisory Board member Dr Katherine Denby
Big data is an ever-increasing part of scientific research. With the expanding use of high throughput genomics, biologists of the future will have to not only possess laboratory skills, but will also need to become specialists in advanced computing. The sheer volume of data is increasing at an alarming and/or encouraging rate, depending on your point of view. This makes it more difficult for biologists to manage, process, manipulate and analyse their datasets.

The iPlant Collaborative is an US-based NSF-funded initiative that offers an easy-to-use, standardised and high performance platform for biologists across all areas to perform complex analyses from initial data evaluation to final visualization. It provides a computing cluster available for all academic biologists and a platform for safely storing and sharing data. The iPlant interface is easy to navigate, providing a large selection of software programs in a manner which makes them straightforward to use, both for those with little prior knowledge of big data analysis whilst also offering multiple running options for those with more experience.

In July 2015 we attended the iPlant Tools and Services Workshop at The Genome Analysis Centre, Norwich. It provided a great grounding in how to utilise the iPlant platform and tools that it offers, from simply storing and exchanging data to RNA-Sequencing and GWAS analysis. The course was well structured and presented, suitable for both iPlant novices and for those who wished to learn how to use iPlant to analyse data in different ways.

The course began with the basic functions of iPlant. We were shown how to manage and import data into the iPlant Discovery environment and how to then create public links and sharing folders to enable the sharing of data with colleagues, collaborators and other iPlant users. This included how to use iDrop, which allows users to directly import data from their hard drives and also iCommands, which allows interaction with the iPlant Data Store. The session also included how to properly annotate data within iPlant using their metadata and how to quickly and easily extract data using these metadata fields.

The workshop then moved to describe how to navigate the Discovery Environment – one of the main portals for exploring iPlant. The Discovery Environment allows you to visualise file storage and perform analysis through a web browser on the cluster. Here we were given a taste of the uses of iPlant by performing a large sequence alignment using MUSCLE and analysing RNA-Seq data using the Tuxedo pipeline (Tophat, Cufflinks and CummeRbund).

The final part of the course focussed on Atmosphere cloud computing. This interface is designed for a user to start and save a virtual computer running on the iPlant high-performance computers, allowing performance of many tasks. You can outsource some computing power to these machines, install software you can’t or don’t want to use on your own machine, and use the interface to set up an external server to view from your computer. We used Atmosphere to visualize our RNA-Seq results within a web browser using the software JBrowse.

This workshop gave a broad introduction to computational analysis using iPlant, as well as introducing us to the concept of traceable, reproducible computing. The iPlant platform was easy to use and fully supports the research analysis requirements of today’s life scientists. A must use resource for all biologists!

The BBSRC has just funded an iPlant UK node so look out on all the usual forums for the announcements of future iPlant tutorial workshops run by iPlant-UK in collaboration with GARNet.

The understanding of synthetic biology is slowing percolating through the plant science community. Whereas the microbial research community started to adopt this technology a decade ago, it is only over the past five years that plant scientists are becoming aware of the possibilities that might exist when using the principles of synthetic biology in their research. In 2012 New Phytologist held a synthetic biology workshop whilst in 2013 GARNet hosted a plant synthetic biology meeting that aimed to set the agenda for the coming decade. In addition and perhaps more importantly, the BBSRC chose to fund the joint Cambridge-Norwich OpenPlant initiative as one of its centres of synthetic biology.

The principles of synthetic biology put great importance on the ability to easily switch between the DNA parts used in different experiments. Widespread uptake of synthetic biology is dependent on a significant amount of front-end design and planning to prepare and define universal cloning strategies and then provide the DNA parts that might be used by a downstream researcher. To that end a number of cloning strategies are available that might fit this bill, each which their own benefits and drawbacks. These include ligation-based strategies such as Goldengate/MoClo/Goldenbraid or overlap based methods such as Gibson Assembly (reviewed here).

Therefore with this backdrop, an important Viewpoint article was recently published in New Phytologist where a collaborative of ~50 plant scientists define: ‘Standards for plant synthetic biology: a common syntax for the exchange of DNA parts’. This work is led by members of the OpenPlant collaboration, namely Dr Jim Haseloff at Cambridge University and Dr Nicola Patron who is Head of Synthetic Biology at The Sainsbury Laboratory in Norwich. In this article the authors introduce the Type IIS assembly method (otherwise known as MoClo assembly) as a common mode of production for DNA parts for plant synthetic biology.

Importantly the article sets out the aims and future principles of plant synthetic biology as having analogy with mechanical or electrical engineering, wherein (DNA) parts made by multiple manufacturers (researchers) are immediate interchangeable, which facilitates and enables technological innovation.

The power of MoClo/GoldenGate cloning comes in the use of the BsaI enzyme, whose 6bp non-palindromic recognition sequence is physically separated from a 4bp restriction site (whose sequence is unrelated to the recognition site). The design of these 4bp ‘fusion sites’ is critical for the modular assembly of DNA parts in MoClo, where multiple parts can be fused within a single reaction using a range of fusion sites across the entire transcriptional unit (TU) (see Figure 1 from Patron et al (2015)).

This Viewpoint article describes 12 ‘parts’ that span a TU from 5’ promoter regions through to 3’ termination sequences. However previous articles have introduced many more DNA parts that are already available to researchers interested in this technology, for example, the 96-well MoClo Tool kit or the items available on the GoldenBraid 2.0 website. In these collections each DNA part is held within an Universal Acceptor Plasmid (UAP) and can be mixed and matched to create many functional units as long as the ‘fusion units’ match up along the length of the clone.

Figure 2: The MoClo 96-well Toolkit

However across the grand scheme of things, the specific nature of these parts is mostly irrelevant. What is most important is that DNA parts that lie in a similar position within any TU contain the same fusion sites so that they can correctly fit into an amalgamated clone at the functionally relevant location. If this is correct then a TU can be made up of any number of available parts designed for a specific regulatory function.

Therefore one future aim of this technology might be to generate a plasmid library that includes the entirety of Arabidopsis promoter sequences, which would all contain the same directional fusion sites so that each of them could be used to drive expression of any particular gene of interest.

Maximising the potential of synthetic biology resources will come as the availability of parts increases. Of course having the DNA parts available is one thing but understanding how they work in different biological contexts is an entirely different question. As researchers looks to use a range of experimental ‘chassis’ such as Arabidopsis, Tobacco or Marchantia then it might be necessary to provide DNA parts that have been optimised for each ‘chassis’.

So what does a lab researcher with their ‘clone of interest’ need to do to take advantage of this type of synthetic biology resource?

Firstly their clone needs to be ‘domesticated’ which, in its most simple terms, involves removal of any internal BsaI sites. As with any cloning reaction, this may not be trivial since it involves two overlapping PCR reactions to incorporate a single mismatch designed to destroy the internal BsaI site. As an aid, the GB2.0 website contains an optimisation tool to help this process.

The researchers clone of interest is fully ‘domesticated’ into the UAP by addition of the appropriate 4bp fusion site that will allow cloning into the correct location of any designed TU. The power of this system truly emerges when considering the vast range of possible regulatory regions that can be attached to the clone, all of which are held within an appropriate binary vector for plant transformation*.

With this Viewpoint article the UK plant synthetic biology community proposes that, given its ease of use, any future DNA parts repository might be best based around MoClo-type technology. However this does not prevent the use of other useful assembly technologies and in time, it might be possible to link seemingly disparate cloning strategies to provide a truly universal mode of DNA-part preparation.

Ultimately the effectiveness of any centralized parts repository is dependent on a number of important factors:

1. Stimulating community involvement to assess which DNA parts would be most useful for the largest number of researchers
2. Providing incentives for academic researchers to develop parts for the community and ensure they get appropriate credit for doing so.
3. Investment in the hardware and software infrastructure to support the physical location of a repository and the upkeep of that facility once it is established.
4. Continued support for wider synthetic biology so that the DNA parts can be used to fully develop the technological capability of the burgeoning bioeconomy.

Finally it is worth noting that although all the synthetic biology tools can be provided for researchers to use, if the individual researcher is unable to amplify their (cursed?) gene of interest then it might all be for nothing!!! Therefore it is vital to keep supporting basic skills in research so that a fundamental knowledge of molecular biology is not lost to following generations.

*- A discussion of transformation efficiency into both model or non-model plants is beyond the scope of this blog, even though it remains a potential bottleneck in many organisms.

The High Value Chemicals from Plants (HVCfP) network is a BBSRC-funded consortium that aims to support the development of resources that will enable researchers to investigate different avenues to produce commercially relevant compounds from plants. These strategies might take advantage of naturally occurring novel biochemical pathways or use plant systems to produce heterologous products of commercial interest. This network is based at the University of York and led by Professor Ian Graham and by Professor Anna Osbourn (John Innes Centre). In their past research both these network-leaders have investigated the potential for increasing the production of high value chemicals, namely the anti-malarial compound Artesmisia or the defence-saponin Avenacin.

On July 13^th and 14^th the HVCfP network hosted a meeting in Dunston Hall, near Norwich that aimed to assess the potential of using synthetic biology (synbio) approaches in the production of high value chemicals. The use of synthetic biology is a key component of the growing bioeconomy, which aims to unlock the potential of biological systems for the production of useful compounds. Synbio was named as one of the governments ‘Eight great technologies’ and the subject of the RCUK ‘Synthetic Biology roadmap’. Principles of synbio encompass engineering molecular and metabolic systems for the economic and pastoral benefit of society. Although synthetic biology has largely been focused on single-celled experimental systems there is clear potential to transfer this technology to plants, as exemplified by the production of flu-vaccine in tobacco.

‘High value’ chemicals in plants are usually produced at low levels and/or in organisms that are currently not amenable to transgenic or synthetic biology procedures. Therefore at the start of this meeting it seemed that there was too much of a distance between principles of synbio and the production of high value chemicals for meaningful discussions to occur. However the meeting was structured in such a way to make it clear that there is not such the disconnect as initially thought, even though there is a long way to go until many of the goals are realized. The full meeting schedule can be obtained from Wendy Lawley.

The first session provided an overview to synthetic biology with Jim Haseloff (Cambridge) introducing general principles of the topic, including some work on the new ‘chassis’ that his lab are using for synbio, the liverwort Marchantia Polymorpha. Rowan McKibbin from the BBSRC then explained that the research councils continue to be committed to synbio as highlighted by a ‘refreshment’ of the Synthetic Biology Roadmap in coming months.

Latterly Joyce Tait (Edinburgh) provided an overview of the future ethical considerations that synthetic biologists might need to consider. Her overall message was that in terms of public acceptance of the technology, synthetic biology is currently in a strong position. Notably it differs from the debate with genetic modification in a number of critical ways 1. Most of the products from synbio are not food-stuffs and 2. The technology is not linked to multinational companies in the same way as, for example, GMO crops. However Kate Miller (Nottingham) made the important cautionary point that few members of the public have actually heard about synthetic biology at this point and so opinions might change….

Session Two focused on some of the technologies involved in synthetic biology. As Head of Synthetic Biology at The Sainsbury lab, Nicola Patron presented ideas about  standardization of molecular parts for use in synbio, highlighting a new paper in New Phytologist that attempts to develop a ‘common syntax for DNA parts in synthetic biology’ . The GARNet blog will have more on this paper in coming weeks but in short, it aims to bring together the community so that transfer of materials will be more routine in future.

Arguably the most striking presentation of the day was from Patrick Cai from the Edinburgh Genome Foundry who introduced approaches that have been developed in the multinational Synthetic Yeast 2.0 project. These include the construction of synthetic yeast chromosomes and attempts to SCRaMbLE the yeast genome, each of which have the grand goal of attempting to re-define what it takes to be a eukaryotic ‘genome’.

The session continued strongly with a classification-redefining presentation from Matias Zurbriggen (Freiburg) who demonstrated the feasibility of recapitulating auxin and light signaling in mammalian cell lines, with the aim of developing general molecular sensors. Marnix Medema (Wageningen) demonstrated the power of a bioinformatic approach to discover biosynthetic pathways in non-model organisms, which might be a critical technology in plants that produce important secondary metabolites.

The first plant image of the workshop appeared courtesy of Sarah O’Connor (JIC) who presented the Madagascar Periwinkle, which produces important monoterpene indole alkaloids.One of the strengths of this meeting was the breadth of ‘experimental chassis’ that were discussed. These ranged from conventional model organisms such a Saccharomyces cerevisae or Arabidopsis thaliana through the less widely used tobacco, microalgae or Dictyostelium and ultimately to plants such as Madagascar periwinkle or Dioscorea (Yam), whose potential is now being investigated due to the novel compounds they are able to produce.

The second day of talks had a more practical edge as Session Three looked at ‘Production Platforms’ currently in place including use of microalgae, tobacco or E.coli. Anil Day (University of Manchester) introduced the benefits of chloroplast expression, either of individual genes or as fusions to the large subunit of Rubisco. This technology appears to have great potential yet it is striking that no transplastomically-transformed plants have been licensed for growth. Dr Day thought that this is due to regulatory and not technological problems. However it was conceded that one potential problem with this approach is the current inability to transform monocots.

The final session of the workshop focused on ‘Metabolic Engineering’ where examples were given for the output and potential of different systems. Jon Marles-Wright (Edinburgh) discussed engineering bacterial nano- and micro-compartments for different applications. He highlighted three potential benefits: 1. They can protect enzymes from high-temperatures or non-physiological conditions, 2. They can safely hold toxins within cells and 3. They can carry heavy metals. However the challenge is now to go from basic research to develop an ‘Encapsulation Platform’. Watch this space…

Arguably the most ‘complete’ story was presented by Olga Sayanova who described a decade-long journey to move production of fish oils from algae into camelina. This has been a cause-celebre in recent years and an exciting exemplar of what is possible with GM technology. Olga did have some cautionary words about the potential of synthetic biology as ‘one-module’ might not fit all. The heterologous expression of algal genes worked fine in Camelina (just a ‘big Arabidopsis’) but did not work in Flax/Linseed. This highlights that each experimental ‘chassis’ might require their own set of molecular tools to enable expression of your product of interest.

Overall the meeting seemed to be a great success with researchers across a range of disciplines learning plenty outside of their comfort zones. It might be a many years until we are able to produce Galanthamine (or similarly complex molecules) in E.coli but the potential is there to move in that direction….

In addition to discussions of HVCfP, approximately 30 attendees were able to take advantage of a Science-Art-Writing workshop over dinner. This JIC-based initiative is led by Jenni Rant and Anne Osbourn with the aim of linking science and art to enthuse schoolchildren about scientific topics. After general discussions about the topic, the attendees were encouraged to write an original ‘ode or haiku’ about synthetic biology and then to blindly draw their table-mates (two representations of your author are below…), which led to great hilarity all around.

However the important message was that seemingly-disparate topics can be brought together to aid education. Hopefully the workshop attendees were given food for thought concerning their next pubic outreach engagement.

I will finish with my indifferently-received ‘Ode to Plant Synthetic Biology’:

O GREEN MACHINE.