2D printed multicellular devices performing digital and analogue computation

Much effort has been expended on building cellular computational devices for different applications. Despite the significant advances, there are still several addressable restraints to achieve the necessary technological transference. These improvements will ease the development of end-user applications working out of the lab. In this study, we propose a methodology for the construction of printable cellular devices, digital or analogue, for different purposes. These printable devices are designed to work in a 2D surface, in which the circuit information is encoded in the concentration of a biological signal, the so-called carrying signal. This signal diffuses through the 2D surface and thereby interacts with different device components. These components are distributed in a specific spatial arrangement and perform the computation by modulating the level of the carrying signal in response to external inputs, determining the final output. For experimental validation, 2D cellular circuits are printed on a paper surface by using a set of cellular inks. As a proof-of-principle, we have printed and analysed both digital and analogue circuits using the same set of cellular inks but with different spatial topologies. The proposed methodology can open the door to a feasible and reliable industrial production of cellular circuits for multiple applications.


Engineered cells library
The genetic architecture of the engineered cells used in this study is summarized in table S1 together with a schematic representation of these constructs, shown in figure S1. There are four groups of cells.
 CS supplier cells: These cells produce de AHL molecules, acting as CS, in different locations of the circuit surface.
o For digital circuits, cell S1 was built to produce AHL molecules in a constitutive manner by expressing LuxI downstream PTet promoter, acting as constitutive in E.coli Top10 strains.
o For analog circuits, cell S2 allows the production of AHL in response to an external input, i.e. arabinose, locating LuxI, synthase, responsible of AHL synthesis 1 , downstream the arabinose-inducible promoter Para. Finally, the PLac promoter [Elowitz & Leibier, 2000], which is repressed by LacI, regulates aiiA production.
 Amplifier cell (CA): In order to compensate AHL decay due to diffusion gradient, this cell produces AHL in presence of AHL. The implemented architecture is based in the Lux system from vibrio fischeri 4 . Receptor protein LuxR, which is constitutively expressed, binds to AHL molecules forming an active dimeric complex. This complex triggers the expression of LuxI under the PLux promoter.
 Reporter cell (CR): In order to validate the correct behaviour of the cellular circuits, this cell produces GFP in response to AHL levels. This cell architecture is based on the same Lux system than CA cell but expressing GFP under PLux promoter, instead of LuxI.  b. In S 2 , S 3 and S 4 cells, the expression of LuxI is regulated by an externally inducible promoter P k (arabinose-inducible promoter P BAD in S 2 cells, rhamnose-inducible promoter P rham in S 3 cells and mercury-inducible promoter P mer in cells S 4 ). K represents the receptor protein, AraC, RhaR and MerR respectively. c. In negative modulatory cells M -, Aiia expression is regulated by an external inducible promoter P k (P BAD , Ptet and Pram respectively). The corresponding receptor proteins K are Arac, aTc and RhaS respectively. d. In positive modulatory cells M + , the expression of LacI repressor is regulated by an external inducible promoter P k (similar to negative modulatory cells) which in turns, negatively regulate Aiia expression. For b, c, and d, the external input (green circles) is either aTc, arabinose or rhamnose. e. Auto-amplifier cells (CA) express LuxI in the presence of AHL, which binds to and dimerises the LuxR transcription factor, that subsequently induces more AHL. f. Reporter cells (CR) produce GFP as a final output in the presence of AHL. Symbol T represent a double-terminator sequence.  Temporal stability of printed circuits. As a case study, a simple transistor-like device (identical to figure 1) was analysed. Circuits were stored on the fridge at 4°C. Every 2 days, circuits were grown at 37°C and measured after 24h. Despite the fold change reduces when circuits are stored in the fridge, differences between the ON/OFF states are still significant after 10 days. Circuits frozen at -80ºC do not show effects on their performance and maintain their functionality, once thawed. Error bars are the standard deviation (SD) of three independent experiments. Data are presented as mean values +/-SD. Numerical quantification of the GFP acquired by scanning the surface of paper strips, with a schematic representation of the experimental setup. In order to determine the signal decay due to the presence of modulatory elements, different conditions were analysed four devices involving no modulators, 1 modulator, 2 modulators and 3 modulators. Additionally, auto-amplifier cells (CA) were added to re-establish the signal decay. a. Schematic representation of each cellular printed pattern with the corresponding distances. S 1 is located 10mm away from CR as it is the maximum distance used in all our experiments. b. Maximum GFP expression by reporter cells depending on the number of modulators. Error bars are the standard deviation (SD) of three independent experiments. Data are presented as mean values +/-SD. Several CR dots were printed to visualize the diffusion range of AHL molecules. GFP levels are normalized by RFP, which correlates with cell population. Below the chart, different pictures of the experiments can be shown.