PROJECT DESCRIPTION
KEY. COLI - PROTECT YOUR GEMS WITH GERMS!
________________
Why do we need Key. coli?
Security, and keys of every type are doomed. Physical keys can be photographed from a distance and 3D printed ; high zoom cameras mean its possible for someone to take a photo of keys left on the side then design a 3D print from this and use that print to open your door. Digital keys can be hacked, and with the advent of quantum computers prime numbers involved in encryption will be void. The currently popular biometrics such as fingerprints and retina scanners can also be hacked with a piece of paper and a photo. This means that almost anywhere someone wanted to get in to, could be brute forced. (ie, they don’t need humans to make a mistake to get through, computational power can hack through doors).
Key.coli concept
In order to create a key it must be 1. Unique, 2. Measurable, 3. Unpredictable, eg: A door key is unique enough that it will only open your house, the door lock measures the key dimensions via pins, and the choice of unique key is random, therefore to some extent the door lock is unpredictable to an invader. A fourth factor is the speed of key production, if the key is to be sold.
The University of Nottingham iGEM team (iGEM being a competition of synthetic biology devices) have created a new kind of key based on fluorescence in bacteria. The team have taken three Jellyfish fluorescent proteins, each corresponding to a colour. Red, green and blue are the primary colours and are also the coloured proteins chosen by the team, therefore a combination of any of these colours (depending on intensity/expression of each) can create any visible colour, much like on a “Photoshop” colour slider. Unlike normal colour however, these fluorescent proteins cannot be seen to their full potential with the naked eye.
Jellyfish protein, and randomness
Jellyfish live under a giant blue filter. Water has a small tint of blue, and the deeper down you go, the more all colours are filtered, apart from blue. This creates an almost black and white abyss in the ocean where the jellyfish live. If I was to wear a red jacket in the deep ocean, it would look like a blue jacket. Jellyfish have many defences and want creatures around to know where they are. They have proteins that absorb the blue light all around, and emit a green light. The green light brings back colour to the otherwise monocolour world. This makes the jellyfish more obvious, saying “Hey I’m a jellyfish. I’m clearly here and green, get lost”. The iGEM team have taken this protein, with modifications by Tsien et al (receiver of a Chemistry Nobel Prize, 2008) and inserted it into bacteria in a modular, factory like way.
The colours of the jellyfish proteins can only be unlocked when the correct wavelengths of light are shone upon them. This means if a bacterial colony was photographed it couldn’t show the range of colours it truly has, as each protein has a spectrum of emission and absorption not visible in white light. In addition, the methods for decision of each colour chosen are decided by a ligation process which is driven by Brownian motion. Brownian motion involves quantum mechanics to orient particle movement, therefore the ligation decisions use a truly random process, that cannot be predicted.
Modular Design
Modular ligation process uses library of P/T parts, giving each protein a different expression randomly. The uniqueness of the key is therefore decided by the array of colours that extend beyond the visible spectrum, the combination of these creates over 11,000 combinations using just the three colours currently being used. In the future it is hoped that with the hundreds of fluorescent proteins that exist, over a google (10^100) combinations can be created. The unpredictability of the key is ensured by quantum biological processes. The mass production element of key design is done via modularity in the ligation and transformation process.
![](https://static.igem.org/mediawiki/2017/a/a1/T--UNOTT--brickstitching.jpeg)
Physical Device
Bacteria are delicate, therefore the team freeze dries the keys they create, similar to how some astronaut food is stored. This lets the keys last extremely long periods of time, even at room temperature. The frozen bacteria will be kept in a key sized device, then can be measured whenever necessary.
Measurement of the bacterial key is undertaken by an array of lasers that excite the jellyfish proteins to reveal their full spectra. An image recognition software designed by one of the Nottingham researchers then compares this spectrum with that of a correct spectrum, or of another freeze dried colony. Only correct spectrum ratios that the programme accepts are allowed access. The laser measurement device can be as small as a conventional door locks used today. Currently a “Raspberry pi” device is being used as a prototype to house all this programming.
Current design
There are two forms of key currently possible, one requires re-activation of the bacteria (which is a slow method) and another requires only a vial of freeze dried bacteria. If the re-activation method is chosen, bacteria typically take at least 20 minutes to revive. The re-activation method can be used at banks, thereby limiting how fast would-be bank robbers could steal money. The police would be there before any money could be taken. The fast method would be a slightly less secure version for personal use, and would be used similar to current keys, simply putting the device in the measurement groove to gain access.
Advantage of Key.coli
The advantage of the key iGEM Nottingham have created is its non-digital, unhackable design. Even with the current 11,000 combinations, it would take years of research for another lab to produce all 11,000 combinations and test them on a door; brute forcing would no longer be an issue. If a google of combinations are used, access would take the lifetime of the universe. It is hoped that when more proteins are used, the key could even be used as a form of ID, databanks holding freeze-dried bacteria would be used as verification on the other side. These databases would be unable to be hacked, especially on large scales such as the 143 million citizens hacked in the US last year, as access would be individually based.
Bacteria, or other biological keys could be the future of security, they might be necessary with quantum technology able to break our passwords coming to the commercial market very soon. However it will likely be years before such keys are available for private use.
How is this Achieved?
Our project aim is to provide a platform for multiple proteins to be expressed simultaneously at different levels. dCas9 will be targeted to repress a given promoter preceding a protein by using a short guide RNA (sgRNA) corresponding to this promoter. Our initial proof of concept system uses fluorescent proteins as the reporter signal and we have taken five promoter-sgRNA pairings from literature and constructed two plasmids which will give many different combinations of protein levels. One plasmid contains three promoters (although this will be expandable) joined up to three different fluorescent proteins. This plasmid will also express dCas9, which we will submit to the iGEM registry. The second plasmid will express 3 different sgRNAs which target each one of these promoters and, by having the option of a non-targeting sgRNA too, we have ON/OFF switches for these fluorescent proteins. In the future, we can expand this to different levels, rather than just two, by creating mismatches in the sgRNA seed region, which has been shown to reduce efficiency of repression by dCas9. Discernible combinations would only be limited by the reproducibility of signals and accuracy of the measurement.
For future expansion, although there is a vast repertoire of fluorescent proteins, any other protein could also be substituted in and measured using other methods. In this way, this system could be used for a vast range of applications, from optimising production of proteins or a metabolic pathway to being used as a biological password. As our assembly method is all interchangeable so any protein can be linked to any promoter, and placed in any position in the plasmid. We have chosen to focus our project on using this random assortment to create a natural random number generator, with application as a biological key as we felt this is the most exciting to the public and shows the design’s mechanism the best. We envisage a system where these plasmids can be assembled randomly (as this is how we designed our system) to produce an enormous number of combinations, which is a valuable characteristic in security. The potential impact from this application is that we will be providing a new, more secure, form of key for accessing content. This is a hot topic at present; numerous major hacking incidents fixate online security at the forefront of many business’s concerns, as accounts are being hacked and sensitive information stolen. Many large companies are deviating away from conventional online character passwords, which are proving to be unreliable in the hands of the public. For instance, many banks are now developing physical biometric authentication procedures to correctly identify the true owner of an account. This new direction opens a market for biological “passwords”. An ideal system would be as decoupled from online software programs as possible, while maintaining the complexity and uniqueness of a biometric system. Cells are effectively living computers so we can programme cells to act as a changeable biometric password. Our product would have no human influence over the outcome of combinations due to our design, making it less prone to issues commonly associated with human carelessness such as using simple, easy to guess passwords. This system combines aspects comparable to the traditional mechanical key, a digital key, and biometric keys - a fourth alternative to the paradigm. Also, this track of research will hopefully be of interest to many and stimulate further research into developing tools for advanced biological computation. The parallels between computation and genetic regulation are astounding, unlocking the potential for fully realised programmable genetic systems would be invaluable to society in numerous ways. For the physical product, we plan on freeze-drying our “programmed” bacteria for storage in a device. We will optimise the conditions for this to optimise reproducibility and allow accurate authentication. The two signals need to have as low a variance as possible, otherwise the accuracy and completeness of emission signals is invalidated and fewer discernible combinations are possible. Therefore, we are using the data collected to form models to predict the outcomes of fluorescence under all possible conditions, and allow each key to be categorised separately by their output.