LAUSR

It is tempting to speculate that sixty years ago, when Jacob and Monod presented their model of the lac operon (Jacob and Monod, 1961), they already had a glimpse of the future of the lacZ gene, not only as encoding a cleaver of disaccharides, nor as a component in a beautiful and groundbreaking model of gene regulation, but also as a universal reporter of gene activation. Indeed, reporter gene technology rapidly became a basic tool in studying the regulation of gene expression; several decades had to pass, however, before the same approach has led to the first report of a microorganism genetically engineered to perform an accurate, specific and sensitive analysis of an environmental pollutant (King et al., 1990). The term ‘whole cell biosensor’ soon entered into use, accompanied by some semantic controversy: purists view the term ‘biosensor’ as a hardware device, in which the biological entity (e.g. enzyme, antibody, oligonucleotide or a live cell) serves as its sensing component (IUPAC, 2017); according to this view, a microbial strain, notwithstanding the complexity of its re-engineering, may be called a ‘sensor strain’ or a ‘bioreporter’, but never a ‘biosensor’. Long before this linguistic polemic became an issue, however, a pioneering article from the Sayler group (King et al., 1990) described a bioluminescent Pseudomonas-based sensor of naphthalene. This publication was trailed by the first E. coli-based mercury sensor (Selifonova et al., 1993), soon to be followed by numerous others, all sharing the same basic structure: a gene promoter induced by the target compound (directly, or via the removal of a repressor), fused downstream of a reporter gene. The latter could code for a traceable protein (e.g. GFP) or – more often – for an enzyme, the activity of which could be monitored quantitatively in real time (van der Meer and Belkin, 2010). When necessary, regulatory elements had to be cloned as well, especially when the gene promoter acting as the sensing element was not native to the host organism. In view of the practically infinite number of gene promoters and regulatory proteins available as candidate sensor elements, the scope of possible sensing targets of such sensors is exceptionally broad. In parallel to the development of microbial sensors of specific compounds, bioreporter strains have also been described for the detection of global sample characteristics such as toxicity or genotoxicity/mutagenicity, parameters of importance for environmental health as well as for chemicals’ safety. The commercial SOS Chromotest (Quillardet et al., 1982), the forerunner of this group of assays, was followed by the umu-test (Oda et al., 1985). In both cases, the activation of gene promoters from the E. coli SOS repair regulon by DNA damaging agents was chromogenically monitored with lacZ as a reporter gene. Looking back over the last 15 years, possibly the most powerful innovator of microbial biosensor design was the coming of age of synthetic biology. While the term has been introduced to the scientific literature over a century ago (Leduc, 1910), its meaning has slowly changed over the years. Following the introduction of the Jacob and Monod model, microbial biotechnology horizons opened up with the advent of increasingly more sophisticated molecular tools, including numerous enzymes derived from diverse microorganisms and viruses, harnessed and retrained to perform cutting, pasting and editing tricks. The same horizons practically exploded when thermophilic variants of these enzymes were ingeniously employed in the invention of PCR technology, and turned essentially limitless when genome sequencing was made trivial and bioinformatic data (and tools for its analysis) became freely available to all. These advances have prepared the ground for the invasion of practitioners of additional disciplines into the realm of whole cell sensor design; when engineers, physicists and computer scientists started to practice biology in earnest, things have started to become truly interesting. In 2004, van der Meer et al. have claimed that one of the reasons current bioreporters’ performance cannot comply with environmental detection standards is the ‘lack of engineering principles’. More or less at the same time, the ‘Biobricks’ concept has been presented (Knight, 2003), aiming to provide ‘a set of standard and reliable engineering mechanisms to remove much of the tedium and surprise during assembly of genetic components into larger systems’. The trend embodied by these two examples paved the ground for engineering school graduates to advise ‘classical’ molecular biologists involved in