Two decades have passed since the landmark paper of Adelman. A major game changer has been the advance of synthetic biology, with novel concepts for bioengineering strongly based on systems theory. This led to trials for identifying, characterizing and standardizing biological parts useful for a general purpose computer. Major advances have been made in areas such as engineering of switches and logic gates, letting the dream of engineering a general Turing machine come close to reality. This dream is finally about our human superiority and rule over nature, making biocomputers one of the really exiting challenges in contemporary science, both in respect to engineering and ethics. We still face a couple of challenges before we will see biocomputers in our daily environment.
Novel concepts for Turing machines have been suggested, such as a deoxyribozyme based molecular walker, as this kind of machines have the ability to read and transform secondary cues. However, the general Turing machine requires the ability of erasing and writing of symbols. Recently, major advantages have been made in respect to genome wide codon replacement in vivo by applying multiplex automated genome engineering technology. This technology provides novel opportunities to implement a general Turing paradigm.
On this road we need to clarify whether the digital paradigm is in fact the best approach to molecular computing. The values of biological signals are typically analog, so we need to explore, if analog computing might be an alternative road to explore. In any case, we need to engineer signals, both as input and output with well-defined stable concentrations, thus do not fluctuate, and stable circuits. If we wish to use Boolean logic we need to be able to group signals in low expressed and high expressed. The engineering design of the logic gate based on the transcriptor mark the advances that have been made towards digitalization of signals and the engineering of clearcut thresholds.
Silicon computers have been a fruitful inspiration for the engineering of computing systems from biological materials. These engineered biological computers have some advantages over the silicon counterpart, as they can potentially self-organize and self-replicate. This has the potential to reduce engineering costs and efforts. However, the overall capabilities of today’s artificial engineered biological computers are still premature in many aspects in comparison to the silicon based one.
Today’s logic gates can only be concentrated for up to order 10 processing steps. The logic problems solved so far by biological computers are impressive, but also demonstrate the inferiority of such systems in comparison with their silicon counterparts, as they are still of relatively simple nature. These problems are both due to the novelty of the field, but also to system specific properties of the biological matter. Biochemical reactions have by nature often long reaction times. The input and output signals are of analog and not digital nature. Biochemical reactions are often in solution and not in all cases compartmentalized, which results in the lack of signal separation. Novel compartmentalization concepts, organizing signal transduction by binding mediators to a scaffold, might further contribute to signal separation. Although these kinds of inert material properties might define the natural limitations for the engineering of biological computers, one might consider a change in the computing paradigm applied, in order to engineer more in coherence with these material properties.
The analog computer paradigm, which uses continuous values, might be interesting in this respect. Daniel et al. have recently published a paper exploring analog computing in living cells and demonstrate that synthetic analog gene circuits can be engineered to execute sophistical computational functions in living cells. Moreover, further improvement might be possible to advancements in biological engineering. Much of the work necessary is in line with standard quality insurance in biological experiments such as system stability and consistency under different conditions, system quantification, and identification of system imperfections. Examples of such experimental problems are: systems might be unstable due to transient transfections.
Moreover, cell populations might be not homogenous due to heterogeneity of gene copies, rate constants and stochastic effects. Furthermore, system measurements are potentially difficult in respect to measuring intracellular input levels. Once experimental advances are made towards standardized and well defined parts, one of the major next engineering steps will be to combine the different units of the biological microprocessor to one complex system.
A challenge will be the spatial organization of such a complex system. Novel artificial scaffold systems might be necessary to develop for this purpose. Efficient manufacture methods might also be required. The emerging field of 3-D printing might provide novel ways for system engineering. Further advancements in engineering of biological control units might be necessary for powerful integrated systems. Altogether, this will push biological systems closer to the level of complexity and problem solving power of silicon computers. Such an integrated system will have much more computing power and advances the problem solving capability. Evidence for the potential of the potential computing power of a biological system is provided by the capabilities of nature’s most powerful biological computer, the human brain.
Novel areas for development are on the horizon. Hybrids of electronic semiconductor and biological machines might be interesting to explore; playing on the initial discussed feedback loop between biology inspired engineering and engineering inspired biology. Some interesting research is going on in this area both in academic labs and in industry.
Several promising biocomposites have been developed, such as cells treated with silicic acid; DNA as a mediator that arranges fullerenes, golden particles and DNA-templated nanowire formation; and DNA metamaterials and hydrogels with memory. Another interesting device under development is IBM’s DNA transistor. This system controls DNA translocation through the nanopore. It is composed of a metal/dielectric/metal/dielectric/metal multilayer nano-structure built into the membrane that contains the nanopore. The function of this system is based on the interaction of discrete charges along the backbone of a DNA molecule with the modulated electric field to trap DNA in the nanopore with single-base resolution. DNA might be moved through the nanopore at a rate of one nucleotide per cycle. This could lead among other to a nanopore-based reading device.
Via Dr. Stefan Gruenwald