Ordered self-assembly of proteins for computation in mammalian cells

Abstract

A cellular logic system capable of combinatorial and sequential logic operations based on bacterial protein-triggered cytotoxicity was constructed. Advanced devices such as a keypad lock, half-adder and several basic Boolean properties were demonstrated on the cells.Recent advances in our understanding of biology are critical to the development of future biocomputers, with several distinct advantages, such as flexible design, efficient energy usage, memory functions and error checking.1,2 These developed biocomputing systems can be generally grouped into two categories, the bottomup approach uses biomolecules for sensing and performing computation; on the other hand, the top-down method utilizes whole cells including prokaryotic or mammalian cells to implement sophisticated tasks. [3][4][5] For example, bacterial virulence proteins have been used as valuable synthetic biology tools to engineer mammalian cells for therapeutic and analytical applications. 6Compared to the silicon-based computers using sole electrical signals for information processing, natural cellular biocomputing systems harness various inputs including chemicals, biomolecules and even physical factors such as light and heat. After the extracellular inputs have been sensed by cells, they pass through the first barrier, the cell membrane, then activate downstream effectors in the cytoplasm, and eventually trigger different gene regulators in the nucleus leading to the expression of the corresponding proteins, which will execute specific tasks.7 All these events take part in the integral information flow from Input to Output, as shown in Scheme 1a. Whereas the information flow from DNA to RNA and to protein synthesis, according to the central dogma of molecular biology, is well known, 8 the crucial role of the cell membrane, the environment-cell interface, in sensing, filtering, amplification, and storage of external signals is still not fully understood. The environment-cell interface shows a phospholipid bilayer architecture to which various receptors and other recognition elements interacting with extracellular stimuli are anchored. 9 Elucidation of the process of information transduction between Input and the environmentcell interface will facilitate the development of biocomputers with useful human-machine interfaces. In this work, we utilize the selfassembly of a bacterial toxin complex on the cell membrane of immortalized mammalian cells as a 3-input framework to conduct different logic operations (Scheme 1b). This extracellular protein complex, the nonhemolytic enterotoxin (Nhe), is expressed by most strains of Bacillus cereus, but can also be produced by non-B. cereus strains. 10 The Nhe complex consists of the cytolytic protein NheA and the binding components NheB and NheC. Nhe has some unique properties: (i) the individual Nhe proteins are not toxic, (ii) toxicity can be evoked only by a specific ordered binding sequence, provided that (iii) the Nhe proteins are present in a specific molar ratio, and (iv) Nhe toxicity can be neutralize...

User Authorization at the Molecular Scale

“…The ability to create an entirely different family of molecular keypad locks, which utilize biological components, was first demonstrated by the group of Katz . Since then, a variety of molecular and biomolecular password systems that respond to biological inputs, such as to enzymes, proteins, and antibodies, have been developed, demonstrating novel mechanisms for differentiating among input sequences at the molecular level. An elegant bio‐security system, 3 , which combines an enzymatic cascade with a biofuel electrochemical cell, is described in Figure .…”

“…Hence, another way to obtain molecular password systems is by harnessing the ability of cells to discriminate among input sequences. The sequence‐dependent assembly of the nonhemolytic enterotoxin (Nhe) bacterial protein complex has inspired Märtlbauer to construct a cellular keypad lock (Figure ) in which mammalian cells are subjected to the sequential addition of the Nhe components: NheA (input A ), NheB (input B ), and NheC (input C ).…”

“…The latter are macroscopic analytical devices that, similar to the way the olfactory system operates, can interact non‐specifically with a wide range of analytes and discriminate among them by creating a wide range of unique identification patterns . Hence, the main difference between 8 and the molecular keypad locks discussed before is that instead of generating a single digital output (0 or 1) for each code entry, it associates each password with a unique emission “signature” that enables it to authorize multiple users.…”

An Enzyme‐Based Half‐Adder and Half‐Subtractor with a Modular Design

“…[48] Theu se of protein-based systems for the realization of arithmeticf unctions is limited to very few examples, including light-activated bacteriorhodopsin, [49,50] actin filaments [51] and proteins for computationi nm ammalian cells based on bacterial protein-triggered cytotoxicity. [52] The only enzyme-based system reported to date hasp erformedt he half-adder operation (the half-subtractor was claimed, but not experimentally realized). [53] The motivation behind enzymebased computing systems is their potentiali ntegration into multi-input binary-operating (YES/NOT format) biosensors, [54] which are particularly useful in biomedical sensing [55] as well as in diagnostic applications.…”

Sequential Logic Operations with a Molecular Keypad Lock with Four Inputs and Dual Fluorescence Outputs

“…Systems with three inputs [7] or dual outputs [8] have remained extremely rare. Apart from these molecular systems, a number of biomolecular keypad locks that are based on the use of enzymes, [9] DNA, [10] aptamers, [11] antibodies, [12] or bacterial toxins [13] have also been developed. Some of them work with three inputs, and one DNA-based system that can have five inputs has been described, [10f] but these systems only feature a single output channel.…”