Novel Genetically Encoded Biosensors Using Firefly Luciferase Frank Fan † , Brock F. Binkowski † , Braeden L. Butler, Peter F. Stecha, Martin K. Lewis, and Keith V. Wood* Promega Corporation, 2800 Woods Hollow Road, Madison, Wisconsin 53711. † These authors contributed equally to this work. ABSTRACT Genetically encoded biosensors have proven valuable for real-time monitoring of intracellular phenomena, particularly FRET-based sensors incorporating variants of green fluores- cent protein. To increase detection sensitivity and response dynamics, we genetically engineered firefly luciferase to detect specific intermolecular interactions through modulation of its lumines- cence activity. This concept has been applied in covalent, noncovalent, and allosteric design con- figurations. The covalent design gives sensitive detection of protease activity through a cleavage- dependent increase in luminescence. The nonco- valent and allosteric designs allow reversible de- tection of the small molecules rapamycin and cAMP, respectively. These sensors allow detec- tion of molecular processes within living cells fol- lowing addition of the luciferin substrate to the growth medium. For example, the cAMP sensor al- lows monitoring of intracellular signal transduc- tion associated with G-protein coupled receptor function. These and other luminescent biosensors will be useful for the sensitive detection of cellu- lar physiology in research and drug discovery. T he observation of dynamic molecular processes within living cells is a long- standing desire of cell biology re- search. Accordingly, recent years have brought increased efforts toward engineer- ing reporter proteins to serve as intracellu- lar biosensors. For example, numerous FRET-based sensors have been configured to detect a range of physiological mediators such as Ca 2+ , cAMP, and phosphorylation (1, 2). These biosensors typically comprise an allosterically responsive domain linking a pair of fluorescent proteins, where the con- formational change influences the relative orientation and separation of the adjoining FRET energy donor and acceptor. Although this approach has proven use- ful for cellular imaging, its suitability for macroscopic measurements can be limited by insufficient detection sensitivity and dy- namic range. In contrast, luciferase reporter proteins are well-known for providing high sensitivity and a wide dynamic range, and they are commonly used for rapid quantita- tion within multiwell microplates. This has been especially valuable in automated screening applications where microplates may contain 1536 wells or more. Moreover, as intracellular probes, luciferases can be used effectively at much lower expression levels than fluorescent proteins, thereby minimizing their potential influence on cel- lular physiology. Luciferases are also com- monly used for molecular imaging in living organisms, such as mice. The beneficial properties of luciferases as intracellular probes have, nonetheless, been limited largely to their use as genetic reporters. To broaden their applicability for elucidating cellular phenomena, we have set out to develop new strategies for config- uring firefly luciferase as a biosensor of in- tracellular molecules or events. Firefly lucif- erase is a 61 kDa monomeric enzyme that catalyzes the oxidation of firefly luciferin in the presence of Mg·ATP and molecular oxy- gen to generate oxyluciferin, CO 2 , and AMP with the concomitant emission of yellow- green light. Crystal structures of luciferase reveal two domains, connected through a hingelike region, that rotate and close to- gether upon substrate binding (Figure 1, panel a) ( 3, 4). Our biosensor design strate- gies modeled this conformational change as the closing of a simple hinge, where each of the strategies was intended to restrict or modulate the motion of this hinge. Because the N- and C-termini are located on opposite sides of the hinge region, the first design strategy was to restrict hinge motion simply by connecting these termini together, thus impeding the luminescent re- action (Figure 1, panel b). This was done by creating a circularly permuted form of firefly luciferase, where the native N- and C-termini are relocated to a new position in the pro- tein structure ( 5). To identify sites tolerant of structural manipulations, we used transpo- son mutagenesis to create a library of mu- tant luciferases containing random inser- tions of five amino acids. One mutant from this library, retaining 75% of the parental lu- minescence activity, contained a polypeptide insertion following Pro233 (Figure 1, panel a). With this location for the new N- and C-termini, a gene encoding a circularly per- *Corresponding author, keith.wood@promega.com Received for review February 28, 2008 and accepted April 23, 2008. Published online June 20, 2008 10.1021/cb8000414 CCC: $40.75 © 2008 American Chemical Society L ETTER ACS CHEMICAL BIOLOGY • VOL.3 NO.6 www.acschemicalbiology.org 346