Specialisation

In our project we are exploring the opportunities of devising a bacterial system with its tasks divided among different cells/populations. It is clear that such a system has several advantages, the main being the fact that more gets done through specialisation of individuals. iGEM projects are often organised in a similar way, due to time constraints and skill distribution in the team. To verify this view, we asked teams whether or not and how they subdivided their team and how they organised and assigned the tasks.

• All teams had used at least partial division of labour to organise their project work. The most common arrangement was to have wet-lab and dry-lab teams as separate entities, with further subdivision in modelling, Policy & Practices and biology teams.
• We asked teams how they came about to have the particular divisions. A common theme was that it just 'naturally happened that way' - A system will naturally evolve towards what seems to be the 'most efficient' arrangement.
• Some teams would first devise a list of tasks to do, and then assign these to sub-teams/people, while others would first divide into sub-teams and then choose particular tasks to do. There were indications that the latter was less efficient: People work on different areas. However, these do not cover all tasks, so there are things that are left undone.
• In all cases, skill was a significant factor in choosing/allocating the tasks: - Specialisation is highly important.
  Usually, the person who is best at a task will do it.
  I can do my own tasks more efficiently individually, because I have the most expertise individually.
  When I do a process repeatedly, it takes me less time as I gain more experience. I become more efficient.
• Fair allocation of tasks was also found to be crucial to the team's efficiency: The effect of 'parasites' on the system can be detrimental.
  There are people in our team who don't do any work at all, so we've ended up with too many tasks allocated to too few people. We don't have time to do everything.

Relevance for bacterial system design

The study of iGEM teams confirmed our aim of introducing more specialised strains into a population, each of which working together towards an output. The specialist metabolic breakdown of each strain, allowing population control can be considered as the first instance of specialisation. However, once this has been established we can then try to fine tune the specialist functions. A good example highlighting the advantage of these specialisation has been described by Pande et al. (2014). The researchers knocked out metabolic genes in two strains of bacteria and showed increased growth rates compared to the wild type strains. This confirms our initial aim of dividing up workload through metabolic wiring. Additionally we identified that parasitism might potentially be detrimental to a system and should hence be considered in our bacterial system design. Interestingly though, the Pande et al group also found that the synthetically cross-feeding strains even had a fitness advantage if in direct competition with non-cooperating strains. This could represent another advantage of the introduction of specialisation, as it means that the cooperating strains outperform potential parasites in a population.

So why is it that social systems like iGEM and bacterial systems alike prefer to utilise specialisation? Surely, in a bacterial system, it is not a conscious decision driven by rational thought... Lindi Wahl (2002) argues that this is simply a case of evolutionary dynamics - survival of the fittest. The beauty of it lies in the fact that this can actually be proven by the power of mathematical modelling! The argument goes as follows.

A population consists of several types of individuals: generalists, who can perform all the necessary tasks, and several types of specialists, who can each perform only a specific task, all of which are needed to achieve the ultimate aim. The frequency of each type in the population will therefore evolve as follows:

where pi is frequency of type i; Fi is its fitness; and Φ is the average population fitness. There are, of course, many factors influencing the fitness of each individual type; the costs and benefits of the task it is doing being one, group size being another. It can be shown that the equilibrium state will have a population that is highly specialised, like this:

What conclusion can we draw from this? It makes logical, mathematical sense for a system to evolve towards specialisation, as the probabilistic chance of success will be much higher. Besides, the larger the group, the more pronounced the specialisation is likely to be. So it most certainly makes sense to plan for specialisation from the very beginning and consider incorporating it in the design.

Pande et al (2014). Fitness and stability of obligate cross-feeding interactions that emerge upon gene loss in bacteria. The ISME Journal 8: 953-962.
Wahl, L.M. (2002). Evolving the division of labour: generalists, specialists and task allocation. Journal of Theoretical Biology 219: 371-388