IGEM:IMPERIAL/2006/project/Oscillator

Biological Oscillator Introduction

• jump to Biological Oscillator Specification Document
• jump to Biological Oscillator Parts Used
• jump to Biological Oscillator Design
• jump to Biological Oscillator Modelling
• jump to Theoretical Analysis of the Model

Motivation

Oscillators and oscillatory systems are ubiquitous in everyday life, from the alternating current electricity that we use to the circadian rhythms that control our sleep and wake cycle. Already, synthetic biology has begun to take on the challenge of creating the first biological computer, starting with ETH Zurich’s 2005 iGEM competition entry for creating a biological NOR logic gate and a two bit counter as well as Harvard’s BioWire design concept to transmit a signal down a length of bacteria. In computers, clocks are used to synchronise the components in order to prevent overflow of information within the system. On a broader scale, these clocks synchronise time around the world and are also used to determine the winner of eBay bids accurate to seconds, perhaps a more recognisable example to our modern life.

Why would we want to create a biological computer? Consider first the human brain, a complex network of cells intercommunicating to create our thoughts and conduct the human body symphoniously. If we can mimic this type of system in other non-neural type cells, we might be able to harness the massive scale parallel computing power inherent to biological systems.

Once we are able to create a biological oscillator, we can then move to synchronise several biological computers paving the way for an internet-like system controlled by bacteria. Further developments in biological to electrical interfacing could mean that communication between electrical devices and biological devices would be seamless. This can potentially integrate the existing infrastructure and novel biological approaches so the current technology would not be drastically displaced, but gradually replaced by biological machines. Moreover, the quest for self-reproducing machines has finally succeeded. Wouldn't it be great if our computers upgraded themselves? Made themselves faster every 30 mintutes? Genetically engineered bacteria indeed have this potential and are only limited by their lifespan and the biological reaction rates. Unfortunately, biological reaction rates are relatively slow when compared to electrical signals, but consider 100 years ago when we knew very little about electricity and how to harness the power of electricity. Biological engineering is at that stage now, and we cannot expect to surpass in a few years the engineering foundations that have been perfected throughout the ages.

Stable biological oscillations are seen to be produced with accuracy in predator-prey relationships, where we assume an exponential growth of prey and insatiable predators. The Lotka-Voltarra model for predator-prey interaction can be implemented given certain assumptions and given that we are able to find biological equivalents to predator and prey. Of course, a molecular predator-prey system would have different assumptions and thus different equations, but the fundamental predator-prey relationship can still hold. The assumptions and adaptations to the Lotka-Volterra system will be discussed further in the modelling document. Once we find molecules that can act similarly to predator-prey interactions, the next step is to successfully implement the system into bacteria!

Past Designs

As synthetic biological is relatively a new field, scientists have a propensity to approach this concept through an engineering approach. The simplest electrical equivalent to an oscillator would be the logic inverter, where a high voltage input is converted to a low voltage output and vice versa. Following this approach, Elowitz & Leibler’s repressilator design connected three biological inverters within a cell to create an oscillator. Although they were able to produce oscillations, the negative feedback inherently caused damping within the system causing the fluctuating green fluorescent protein (GFP) output to gradually die down. They chained three inverters together in order to create a delay mechanism; however, this only propagated the noise within the system and also led to a faster decline in the amplitude of the output sinusoid. If we are to create a more robust oscillator, then two options are possible. First, we improve on the design by ensuring that biological noise is minimised and that we have dependent translation/transcription mechanisms. Second, we can look to other biological oscillating systems for inspiration and mimic those. Indeed it is a trivial feat for an undergraduate electrical engineering student to fuse together 3 inverters to make an oscillator, but can this uniquely electrical engineering approach be used to engineer bacteria? If we attempt to force an engineering attitude upon a uniquely biological system, then we might not succeed in obtaining desired results due to current incompatibility between the two ideologies. As a result, our efforts could potentially lead to failure.

The iGEM competition is riddled with designs for oscillators, some getting to modelling stages, and others, destined to fail from the start, scrapped for better designs whilst in the brainstorming process. In 2004, MIT modelled a relaxation oscillator based upon designs springing from 2000. Harty, et al. in their paper published in 2001 proposed a two plasmid model using RcsA and CI with interacting networks. The potential provided by quorum sensing molecules for use in a population controlled oscillator was quickly exploited by MIT’s 2004 SMUG designs. AiiA, an enzyme that degrades N-acyl homoserine lactone (AHL) provides a way to quench the quorum sensing ability, giving the possibility of control over cell to cell communication. Not only did MIT exploit the quorum quenching enzyme, they successfully modelled a one-cell approach to an oscillator. Their population wide control mechanism would be based upon the AHL diffusing into other cells producing a controlled oscillation throughout the cell population. Although their design was excellent, it neglected noise within each cell and that cell mechanics are relatively slow. Each cell would supply its own oscillating system, but this design relied too much on diffusion of AHL throughout the entire population to synchronise the entire colony. Moreover, the most accurate oscillations would be present within individual cells, not within the medium that the cells were hosted. Thus, actual measurement of the internal oscillations would require minute biosensors currently unavailable on the market.

Aim

To create a biological system which can mimic self-perpetuating predator-prey oscillations, thus creating a constantly fluctuating sinusoidal chemical output without damping of oscillations induced by an initial stimulus.