Dynamic Behaviour of Fungi in Microfluidics - A comparative study

All multi-cellular living organisms are very complex microfluidics systems, which are assembled 'inside-out', as the result of a complex 'tug-of-war' process comprising both feed-forward modules, in particular the program embedded in the species' DNA, and feed-back processes, in particular the response to external environmental conditions. Living organisms have to solve a perennial problem: for an environment with limited resources (finite and spatial distribution of nutrients, geometrical limitations, competition from other species, etc.), finite growth rate and penalties applied to non-optimal behaviour, what is the best strategy that will maximise growth and survival of the individual and the species? Identifying how organisms resolve this problem is very difficult due to the high complexity involved. However, filamentous fungi and other simple organisms, provide a means to address these issues due to the availability of microscopic imaging and microfabricated geometries in which the behaviour can be tested. Semiconductor fabrication (photolithography, RIE etching, and soft lithography techniques) was used here to produce artificial micro-networks, which are physically and chemically structured 3D microenvironments. Continuous imaging of fungal behaviour in microfluidics networks required the use of PDMS, which is transparent and O-2-permeable. Using these structures, the dynamic growth behaviour of two fungal species Neurospora crassa and Armillaria mellea was observed microscopically in real-time. Growth parameters such as the tip extension velocities, branching angles and branching distances on plain surfaces differed vastly between the species. Features in the micronetworks of a size similar to the hyphal diameter induced the largest change in growth parameters such as, the largest decrease in branching distances in N. crassa and a near complete suppression for A. mellea. Despite these fundamental differences, both species negotiated the structures successfully. These results open the way to asking more fundamental questions, such as which species use algorithms that are most efficient for solving a particular type of mathematical problem coded in a network? Can these natural algorithms be used for either better control of fungal growth and metabolism for both medical and biotechnology applications; or even for new computation paradigms?