katE - Visualising Stress, explores how genetic engineering and synthetic biology allow artists to tap into biological processes and literally visualise what is otherwise invisible. This work began by growing bacterial colonies unusually large and asking the question: Would it be possible to capture biological states such as stress caused by depleted food. Using techniques from synthetic biology I was able to extract a genetic switch involved in stress control from the genome of E.coli and produce a new genetic device that would report this stress using colour (i.e. Green Fluorescent Protein (GFP)).
The artwork was developed as part of an immersive laboratory practice undertaken by the artist who relied on evidence-based art practices where the methods used produce the work required scientific approaches in particular those related to genetic engineering and synthetic biology. Through a series of questions the work tried to unravel some of the genetic foundations for biological stress but took this further by actually engineering genetic circuits to visualise these processes. The work offers a way of seeing stress through colour.
In order to tap into processes such as stress in E.coli, I needed to understand the biochemical changes and how these mechanisms are registered on a genetic level. It turns out E.coli has a whole range of biological stress mechanisms whose genetic control is complex due to cascading effects making it difficult to pinpoint the specific genetic factors involved. However, one biochemical event taking place when E.coli becomes stressed, such as in the case of starvation, is the release of reactive oxygen (O2-)[i] toxic to the cell. To cope, the cells need to breakdown reactive oxygen and E.coli does this by producing enzymes such a Catalase. A known gene responsible for producing this enzyme, catalase HPII (III), is katE. It is also known that rpoS gene is a key factor in regulating the transcription of katE, the RpoS protein product has the effect of cascading transcription rather than producing a high yield of proteins needed for visual expression for stress. A question I asked was: Who tells the genome what to make? If there is such a thing as a switch on the genome and how is this activated? One component that can be understood as a switch is a promoter in that it either allows or blocks transcription (i.e. make RNA) in the presence of certain substances (keep in mind the modern dogma for biology: DNA -> RNA -> Protein). Many of these promoters work independently of the gene they are switching on or off allowing them to be used with other genes, much like electronic switches. In sum, getting the promoter for katE would thus serve as a candidate for studying microbial stress in that a colour generating gene could be switched on under these conditions.
Due to the artistic requirement for this work and learning process much time was spent on learning molecular cloning techniques.
The first phases used traditional genetic engineering methods and involved finding a genetic construct or plasmid with a Green Fluorescent Reporter containing an open slot for the promoter to be placed. Using a PubMed search, I was able to locate a paper “A comprehensive library of fluorescent transcriptional reporters for Escherichia coli” where two constructs (pUA66 and pUA139) had been used to study promoter activity with GFP as a fluorescent reporter.
Both plasmids were both good candidates and emails were sent (27th of April) to the principal investigator as referenced on the paper, Prof. Uri Alon and Dr. Anat Bren, requesting both plasmids (these arrived from Israel on the 19th of May 2010).
To locate the promoter for katE, a PubMed nucleotide database search was made. The result showed “Escherichia coli str. K-12 substr. MG1655): hydroperoxidase HPII(III) (catalase)” with the map of the region around the location of the gene.
Given that adjacent genes were pointing in the upstream direction and katE pointing downstreams this was a good indication that at least 2 promoters in the inter-genetic non-coding space between the upstream adjacent gene (cedA) and the katE. One reason to support this assumption is that each gene has its own promoter and is normally located within 30bp or so from the gene they control. Further, cedA and katE were pointing in different directions meaning that two promoter should be present in the region. Where pointing in the same direction and if the region was small it could be that the one promoter coded for both genes to be expressed. (A gene can run in both direction upstream (5’ to ‘3) or downstream (3’ to 5’) but the transcription only happens downstream.)
It was not necessary to know exactly where the promoter for katE was. The paper providing the plasmids suggested that a promoter should be retrieved by moving 50-100bp into the inter-genetic region, that is the adjacent genes (cedA and katE). PubMed’s sequence viewer provides a FASTA view or the sequence of the selected region.
Using New England Biolab’s NEBCutter program I checked if the selected region contained any unwanted restriction site making it incompatible with the plasmid’s insertion sites, fortunately no overlaps where present. With the sequence at hand, I would design primers, that is small sequences of DNA that can be synthesised by companies and allow me to physically extract or amplify the sequence from the genome. I used a program called Primer 3 Plus, and selected the wanted region to amplify. The program uses many parameters that optimises conditions and if successful return either one or several primers. Primers come in pairs in order to amplify each direction and are used with other reagents and a DNA template. The key to the processes is a repeated process of heating and cooling (PCR) resulting in more and more of specific section of DNA being produced such as my katE promoter.
To make the extracted promoter compatible with the plasmid in the sense of being able to slot it in front of the GFP, the sequence of the respective restriction sites where added to the primers, this meant that the amplified fragments could be cut to fit into the plasmid.
Finally, I also added a two random basepairs to each of the primers, this was done to help the digestion or cutting processes.
The plasmids arrived in a standard letter containing two eppendorf tubes with Agar where a stab of bacteria had been made and a statement letter for customs. The tubes labelled µC1655+U66 and µC1665+u139 and marked with a single blue dot indicating that the selective marker, Kanamycin. The bacteria were streaked out on plates containing Kanamycin. The following day, a single colony was inoculated in 10ml of LB-broth with Kanamycin and grown in a shaker overnight. Glycerol stocks were taken and the remaining broth spun down for mini-prep. The mini-prep’ed plasmids were run on 1% agarose gel and showed the presence of plasmids with low but sufficient yield for both plasmids.
A period of much work followed in order to get significant quantity plasmid given its low-copy number and in order to cut the plasmid. Whilst the PCR of the katE promoter went straightfoward and its sequencing proved correct, the process of ligation and selecting colony was time consuming. Finally, a new construct was successfully produced with katE promoter and GFP. Whilst both gels and sequencing returned consistent result, a problem arose in that the GFP expression was barley visible to the naked eye and from an artistic point of view made it problematic by requiring sensitive instrument to see differences.
My belief was that the low-copy number was causing the issue and therefore developed two routes to re-engineer the existing system into a high copy-number plasmid. One of the tricky part was extracting the large fragment using PCR, for this high-fidelity polymerase (transcription enzyme) was used. The route chosen, would extract almost the entire plasmid, except the origin of replication thought to be the reason for low-copy number. The origin of replication from a high-copy number plasmid was also extracted. Finally, these where put togther but again the outcome was unclear as there was insufficient GFP expression.
A new approach was taken, beginning with registering the laboratory as part of the partsregistery(MIT) and then requesting the synthetic biology toolkit. The toolkit arrived in December 2010 in a box containing three 96-well plates each well containing a plasmid. As materials changed, a series of new material was ordered in the beginning of 2011 to work with the library. Primers had to be re-designed to work with the biobrick standard. I preformed basic work using the library alone and ensured that techniques worked by successfully cloning or connecting a constitutive (always on switch) promoter with a GFP gene producing a green iridescent colour. With this working, I was able to clone the katE promoter with GFP resulting in a significantly higher expression. The parts where then sequences and gave a consistent result.
katE - Visualising Stress produced several interesting outcomes in terms of process but also showed that it is possible for artists to develop such techniques to tap into processes invisible to us. The process showed the need to be stringent in the arts as the resulting outcomes are shared with audiences and can place different demands on what is produced that the sciences where measurements can be quantified by sensitive machines. Whilst imagining the work, my interest was in differences produced over time. The process of visualising outcomes and engineering using genetics and synthetic biology are in some ways to different stages. katE is exhibited in a petri dish with softagar or in liquid culture. It is intended to be allowed to grow overtime, forming large colonies and showing the agitation of bacteria over time through the iridescent colour.