WO2015067302A1 - Sensor molecules and uses thereof - Google Patents

Abstract

The invention generally relates to methods, reagents and devices for determining a concentration of an analyte in a sample. Provided is a proteinaceous sensor molecule for detecting an analyte of interest (A), comprising a signal generating moiety tethered to a synthetic regulatory molecule capable of modulating signal generation by intramolecular binding, wherein (i) the synthetic regulatory molecule comprises a primary ligand (L1) capable of intramolecular binding to a primary partner (BP1) on the sensor molecule; wherein (ii) the sensor molecule comprises a secondary ligand (L2) capable of binding to a secondary binding partner (BP2) which is A or a binding partner thereof; and wherein (iii) binding of L1 and L2 to their respective binding partners is mutually exclusive, such that binding of BP2 to L2 influences binding of L1 to BP1, resulting in a conformational change within the sensor molecule such that the generated signal is modulated.

Description

Title: Sensor molecules and uses thereof.

The invention generally relates to methods, reagents and devices for determining a concentration of an analyte in a sample, such as an analyte in a sample of bodily fluid, as well as methods and devices which can be used to support the making of such determinations. More in particular, it relates to the field of in vitro detection methods using luminescence.

Luminescence is a phenomenon in which energy is specifically channeled to a molecule to produce an excited state. Return to a lower energy state is accompanied by release of a photon. Luminescence includes fluorescence, phosphorescence, chemiluminescence, and bioluminescence. Luminescence can be used, among others, in the analysis of free analytes or biological interactions.

In 2009, the inventors introduced an approach for the generation of semisynthetic protein-based biosensors for small molecule analytes. The

fluorescent biosensors were named SNAP-tag based Indicator protein with a Fluorescent Intramolecular Tether (SNIFIT). See Brun et al. J Am Chem Soc. 2009;131(16):5873-84 and Brun et al. J Am Chem Soc. 2011;133(40):16235-42. The sensor is constituted of a fusion of a binding protein, specific for the analyte of interest, and a synthetic portion constituted of a ligand for the binding protein and two fluorophores forming a FRET-pair tethered to the protein construct via self- labeling-tag fusion. The proteinaceous sensor molecule undergoes a conformational change resulting from the displacement of the tethered ligand by the analyte of interest. This causes a detectable change in FRET-efficiency between the two fluorophores. By choosing a suitable binding protein and its relative tetherable ligand, virtually any small metabolite can be sensed and several examples have been disclosed (Brun et al., , J Am Chem Soc 2012, 134, 7676-8; Masharina et al., J Am Chem Soc 2012, 134, 19026-34).

However, despite its flexibility, the SNIFIT sensor design suffers from severe limitations: the binding protein has to fulfil strict geometrical requirements, namely the relative spacing of the binding site and the protein terminus has to be small. When a suitable binding protein is available, an optimal ligand with suitable affinity must be found and successfully tethered. These requirements make the generation of a new sensor non-trivial and time-consuming. Moreover, the SNIFIT concept is limited to the change in FRET-efficiency as readout. To overcome these limitations, the inventors aimed at a substantial innovation in the SNIFIT sensor concept to allow a much broader application and to introduce a relative ease in novel sensor generation. Ideally, the new sensor concept should not be limited to resonance energy transfer as readout and facilitate a broad practical applicability, e.g. in a point-of-care device.

This was achieved by modifying an optimized sensor with a second ligand which is specific for the analyte of interest e.g. a protein or for a binding partner of a small molecule analyte. Provided is a proteinaceous sensor molecule for detecting an analyte of interest (A), comprising a signal generating moiety tethered to a synthetic regulatory molecule capable of modulating signal

generation by intramolecular binding, wherein (i) the synthetic regulatory molecule comprises a primary ligand (LI) capable of intramolecular binding to a primary partner (BPl) on the sensor molecule; wherein (ii) the sensor molecule comprises a secondary ligand (L2) capable of binding to a secondary binding partner (BP2), BP2 being A or a binding partner thereof; and wherein (iii) binding of LI and L2 to their respective binding partners is mutually exclusive, such that binding of BP2 to L2 influences binding of LI to BPl, resulting in a conformational change within the sensor molecule such that the generated signal is modulated.

In the absence of analyte of interest, the primary ligand (LI) is intramolecularly bound to a primary binding partner (BPl ) on the sensor. The sensor contains a moiety which can modulate a detectable signal depending on the conformational configuration of the sensor. The configuration is controlled by the interaction between LI and its primary binding site. Since the binding of primary and secondary ligand (L2) to their respective binding sites is mutually exclusive i.e. either LI or L2 is bound, the presence of analyte influences the extent of LI binding to BP1 and hence the signal of the signal generating moiety by inducing a conformational change.

For example, the binding of L2 to an analyte of interest displaces LI from its primary binding site on the sensor molecule, such that the sensor reaches an open conformation. As another example, in the absence of analyte L2 is bound to a binding partner of the analyte, thus preventing binding of LI to the primary binding site such that the sensor is in an open conformation. The presence of analyte displaces L2 from the binding partner, allowing LI to close the

conformation of the sensor via intramolecular binding to its binding site on the sensor. In both examples, the conformational change of the sensor modulates the signal generating moiety of the sensor. As will be described in more detail herein below, the conformational change e.g. from an open to closed conformation can result in a modulation of the detectable signal depending on the design and nature of the signal generating moiety. For instance, if detection is based on FRET or BRET between a donor and a fluorochrome acceptor comprised in the sensor molecule, an open conformation typically results in a decrease in resonance energy transfer efficiency. In contrast, if the binding site of LI is within the active site of an enzyme comprised in the sensor, opening of the sensor can release the inhibition and increase or induce enzymatic activity. This novel concept, referred to as Steric Displacement-SNIFITs (SD-

SNIFIT) opens the way to adapt the SNIFIT sensor concept to virtually any biochemical analyte and makes it adaptable to the most diverse readout systems. For example, detection can be based on any enzyme capable of generating a detectable signal, e.g. luciferases, enzymes catalyzing chromogenic or fluorogenic reactions, redox enzymes that can be coupled to electrochemical devices, etc.

Detector systems that rely on steric competition are known in the art. For example, US 3,935,074 relates to immunoassays for detecting a ligand of interest, and the use thereof in medical diagnosis. A reagent is provided having at least two epitopes, one of the epitopes being common with the ligand of interest, and the other epitopes being foreign to the ligand. The two epitopes are positioned in the reagent so that antibody bound to one of the epitopes sterically inhibits the binding of antibody to the second epitope. As another example, US 6,455,288 relates to methods for immunoassay of analytes employing mutant glucose- 6- phosphate dehydrogenase (G6PDH) enzymes as labels. Disclosed are methods for determining the presence or amount of an analyte in a sample suspected of containing the analyte, comprising the steps of: a) combining in an assay medium: 1) the sample, 2) a conjugate of an analyte analog and a mutant G6PDH wherein the G6PDH has at least one amino acid mutation per subunit as compared to precursor G6PDH wherein at least one of the mutations comprises the introduction of a cysteine residue proximate to an epitope recognized by an inhibitory anti- G6PDH antibody capable of simultaneously binding to two of the subunits within the same G6PDH molecule, 3) an antibody capable of binding the analyte and the analyte analog conjugate, and 4) substrates for the enzyme; and 5) measuring the activity of the enzyme. Also provided is an analyte-label conjugate, employing as the label mutant G6PDH enzymes having at least one mutation per subunit as compared to precursor G6PDH wherein the mutations are proximate to an epitopic site recognized by an anti-G6PDH antibody capable of inhibiting the activity of the precursor G6PDH.

However, the prior art reagents differ significantly from a sensor of the invention. For example, none of them comprises the use of tethered, intramolecular ligands that affect the conformation of the sensor.

The skilled person will appreciate that the novel approach taught in the present invention can be realized with sensors having widely diverse designs and geometries as long as LI and L2 have the functionalities as herein described above. Figures 1 and 2 merely represent a variety of exemplary sensor molecules of the invention and in no way exclude still further variants. X denotes the site of tethering of the synthetic regulatory molecule to the signal generating moiety.

In one embodiment, both LI and L2 are part of the synthetic regulatory molecule. See Figure 1A for a schematic drawing of an exemplary sensor molecule wherein LI is bound to a moiety of the sensor moiety referred to as BP1 comprising a primary binding site for LI. Since LI is bound and L2 is free, the sensor has a closed conformation in the absence of analyte (in Fig. 1A referred to as BP2), e.g. a protein of interest. However, the presence of BP2, which is capable of binding to L2, displaces LI from BPl, resulting in the opening of the sensor. Figure IB shows a schematic drawing of an exemplary sensor molecule wherein L2 is bound to BP2, wherein BP2 is also a binding partner of the analyte A. Since L2 is bound and LI is free, the sensor has an open conformation in the absence of analyte. However, the presence of analyte A displaces L2 from BP2 thus allowing LI to bind to BPl. This results in the closing of the sensor. As indicated by the dotted lines, BP2 can be a separate entity or it can be part of the sensor molecule. In one embodiment, BP2 is a separate entity. In another embodiment, BP2 is part of the regulatory molecule. In another embodiment, L2 is not part of the regulatory molecule but specifically tethered to or part of the proteinaceous region of the sensor molecule that contains the primary binding site. See for instance Figure IC or ID wherein L2 is conjugated or attached to BPl. In case of Figure IC, binding of analyte BP2 to L2 displaces LI from BPl, resulting in the opening of the sensor. In the sensor design of Figure ID, the presence of A displaces L2 from BP2 thus allowing LI to bind to BPl. Also here, BP2 can but does not have to be part of the sensor molecule.

The skilled person will appreciate that any combination of primary ligand and primary binding partner can be used. In one embodiment, BPl is the active site of dihydrofolate reductase (DHFR) or a circularly permuted variant thereof, preferably in combination with trimethoprim, methotrexate, or a variant thereof as LI. In another embodiment, BPl is the active site of human carbonic anhydrase (HCA), preferably in combination with an aromatic sulfonamide or variant thereof as primary ligand. As said, a sensor molecule of the invention is characterized by a secondary ligand capable of binding to a secondary binding partner, which can be the analyte of interest itself, or a binding partner (e.g. binding protein) of the analyte of interest. In one embodiment, the invention provides a sensor wherein the secondary ligand (L2) is capable of binding to BP2, BP2 being the analyte, such that in the absence of free analyte L2 is free and LI is bound to BPl and wherein the binding of analyte to L2 displaces LI from BPl resulting in an open

conformation state of the sensor molecule such that the signal generation is modulated. This concept is referred to as "normal steric displacement". See

Figures 1A and 1C for a schematic drawing of exemplary sensors based on normal steric displacement.

In an alternative embodiment, the secondary ligand (L2) is capable of binding to a binding partner of the analyte of interest (BP2), such that in the absence of analyte L2 is bound to BP2 and LI is free and wherein the binding of analyte to BP2 displaces BP2 from L2 allowing LI to bind to BPl resulting in a change in the conformational state of the sensor molecule such that the signal generation is modulated. This concept is referred to as "competition steric displacement'. BP2 can be a separate entity or it can be part of the sensor molecule. In both cases, analyte binding to BP2 displaces BP2 from L2, thus freeing LI which then can bind intramolecularly to BPl. See Figures IB and ID for a schematic drawing of an exemplary sensors based on normal steric displacement.

In one specific aspect, BP2 is not covalently bound to the sensor. See for example Figure 4. In another specific aspect, BP2 is part of the proteinaceous sensor molecule and binding of L2 to BP2 is intramolecular. The presence of analyte disrupts intramolecular binding of L2 to BP2 and induces intramolecular binding of LI to BPl. In some embodiment, e.g. for RET-based sensors (see further herein below), the relative positions of BP2 and BPl within the sensor molecule are designed in such a manner to ensure that the change in intramolecular binding alters the conformation of the sensor such that signal generation is modulated. For example, each of BP2 and BPl can be located at one of the termini of the sensor and the regulatory molecule comprising LI and L2 can be tethered to an

intervening region. See Figures IB or 5 for a representative example. Preferably, the synthetic regulatory molecule is tethered to the proteinaceous moiety in a site-specific fashion to ensure a single, homogenous product. The site of attachment can be chosen among any part of the proteinaceous moiety. The site of attachment of the synthetic regulatory molecule to the proteinaceous moiety is chosen such that it allows for a signal change when the sensor molecule switches between the different conformations upon analyte- induced steric displacement of either LI or L2 from its respective binding partner.

Site-specific attachment of the synthetic regulatory molecule can be achieved by methods known in the art. For example, an amino acid (natural or non- natural) showing a unique reactivity is suitably used. Suitable amino acids include cysteine and any (unnatural) amino acid that allows for a site-specific chemical conjugation reaction, such as click-chemistry, of an appropriate synthetic regulatory molecule. For example, the unnatural amino acid azidohomoalanine (AHA) can be used.In another embodiment, the synthetic regulatory molecule is site- specifically tethered to the proteinaceous moiety by means of a protein labelling tag. Preferably, the protein labelling tag is a self-labelling protein known in the art, such as SNAP-tag, CLIP-tag or Halo-Tag, and wherein the synthetic regulatory molecule is tethered via the appropriate reactive group. In one embodiment, the self-labeling protein tag is based on a human 0⁶-alkylguanine- DNA-alkyltransferase (hAGT) to which the synthetic regulatory molecule is tethered via a reactive group for hAGT. For example, the protein tag is a SNAP-tag or CLIP-tag. Preferably, the reactive group is a 0⁶-benzylguanine (BG), 04-benzyl- 2-chloro-6-aminopyrimidine (CP) or 0²-benzylcytosine (BC) derivative. In another embodiment, the self-labeling protein tag is based on a modified haloalkane dehalogenase to which the synthetic regulatory molecule is tethered via a chloroalkane (Halo-Tag). Alternatively, the protein labeling tag can be a tag that is labeled with the synthetic regulatory molecule through the action of an enzyme, such as sortase (and mutants thereof), lipoic acid ligase (and mutants thereof), biotin ligase (and mutants thereof), phosphopantetheine transferase (PPTase; and mutants thereof). Labeling can be achieved by directly transferring a molecule carrying the synthetic regulatory molecule to the protein tag or by a two-step procedure where in the first step a molecule comprising a bioorthogonal group is attached and in the second step the bioorthogonal group is reacted with the synthetic regulatory molecule comprising an appropriate functional group. For example, enzymatic transfer of a modified phosphopantetheine derivative carrying the synthetic regulatory molecule results in labeling of a specific serine within a certain peptide sequence derived from acyl carrier proteins and thus allows the synthetic regulatory molecule to be linked at exactly one residue present in the protein (see N. George et al. J Am Chem Soc. 2004 126, 8896). ACP-tag and MCP-tag are such sequences derived from acyl carrier protein. The presence of the phosphopantetheine transferase is required for the formation of a covalent link between the ACP-tag or MCP-tag and their substrates, which are derivatives of Coenzyme A (CoA). In the labeling reaction, the group conjugated to CoA is covalently attached to the ACP-tag or MCP-tag by the phosphopantetheine transferase. An example for the two-step strategy would be a labeling in which in the first step, a mutant of lipoic acid ligase (LplA) ligates a ircmscyclooctene derivate onto a LplA acceptor peptide which is part of the sensor molecule. In the second step, ligated ircms-cyclooctene is chemoselectively derivatized with a synthetic regulatory molecule conjugated to a tetrazine. Details of such a two step procedure are described by Liu et al. (J Am Chem Soc. 2012 Jan 18;134(2):792-5).

Alternatively, the synthetic regulatory molecule is site-specifically tethered to the proteinaceous moiety by means of intein-based labeling. For example, the use of so-called expressed protein ligation (T. Muir, Annu. Rev.

Biochem. 2003. 72:249-289) would entail expressing the proteinaceous moiety as fusion protein with a C-terminal intein and the subsequent isolation of the corresponding C-terminal thioester. This thioester is then reacted with a cysteine residue to which the synthetic regulatory molecule is attached, resulting in formation of functional sensor molecule. In split-intein-based protein labeling (Volkmann G, Liu X-Q (2009) PLoS ONE 4(12): e8381), the proteinaceous part of the sensor molecule can be expressed as a fusion protein with a C- or N-terminal split intein. Addition of an appropriate synthetic peptide that represents the other part of the split intein and that also carries the synthetic regulatory molecule results in formation of functional intein, the subsequent excision of the intein from the protein and formation of a functional sensor molecule (Volkmann G, Liu X-Q (2009) PLoS ONE 4(12): e8381).

In certain embodiments, like in sensors based on RET, , the site of specific attachment of the synthetic regulatory molecule in the sensor molecule may be connected via a proteinaceous linker moiety to the other part of the proteinaceous moiety. The linker moiety can be an artificial polypeptide sequence or a naturally occurring protein designed to ensure that analyte-induced ligand displacement causes a sufficient change in the conformational state of the sensor.

Poly-L-proline linkers can be used as precise molecular rulers due to their well-defined property of forming a stable and rigid helical structure (the polyproline II helix) with a pitch of 3.1 A per residue in aqueous solution.

Accordingly, the linker moiety is preferably a helical linker rich in prolines, which leads to structural rigidity and isolation of the synthetic regulatory molecule from the attached luciferase. Very good results were obtained with a poly-L-Proline linker consisting of at least 15 Pro residues, for instance Prois, or Pro3o or even longer. Brun et al. (2011) investigated polyproline linkers of varying length (0, 6, 9, 12, 15, 30, 60) that were inserted between SNAP- and CLIP-tag in the conventional SNIFIT-sensors. It was found that a length of 30 or 60 proline residues yielded an improved maximum ratio change of the sensor. Accordingly, in one embodiment the linker moiety consists of a poly-L-Pro linker comprising at least 15, preferably at least 20, more preferably at least 30, residues.

A sensor molecule of the invention may contain any type of signal generating moiety which is capable of generating a detectable signal. As described herein above, the activity of the signal generating moiety is modulated by a change in the binding ratio L1/L2 to their respective binding partners (LI to BP1 and L2 to BP2). In one embodiment, the signal generating moiety comprises an enzyme E. Preferably, E is selected from the group of enzymes catalyzing a chromogenic, luminescent, fluorogenic or redox reaction. Examples of E include peroxidases, luciferases, hydrolases, glucose oxidases, beta-galactosidases. Enzyme modulation can be direct, for example by designing BP1 to be part of E, preferably in or close to the active site of E. Therefore, in one embodiment of the invention a proteinaceous sensor molecule for detecting an analyte of interest comprises a signal generating moiety comprising an enzyme (E), to which is tethered a synthetic regulatory molecule capable of modulating signal generation by intramolecular binding, wherein (i) the synthetic regulatory molecule comprises a primary ligand (LI) capable of intramolecular binding to a BP1 on E, preferably wherein BP1 is in the active site or an allosteric site of E, and wherein (ii) binding of LI to the active site of E leads to inhibition of the corresponding enzymatic activity of E and wherein (iii) binding of LI and L2 to their respective binding site is mutually exclusive, such that binding of BP2 to L2 influences binding of LI to the active site of E, resulting in a change of the enzymatic activity of E within the sensor molecule such that signal generation is modulated. See Figure 2C for a schematic representation. Preferably, LI is an inhibitor of the enzyme such that binding of analyte to L2 ligand sterically displaces LI, resulting in activation of the reporter enzyme. We call this sensor design TURNON-reporter. For instance, Example 4 herein below which exemplifies a sensor using Renilla luciferase as reporter enzyme in a sensor tethered to a regulatory molecule comprising the luciferase inhibitor

coelenteramide as primary ligand and biotin as secondary ligand to sense the analyte protein streptavidin.

Again, as disclosed herein above, LI and L2 can both be part of the synthetic regulatory molecule or L2 can be specifically tethered to or part of the enzyme. Also, the sensor can be based on the normal or competitive displacement.

In one embodiment, the signal generating moiety is capable of generating a resonance energy transfer (RET) signal, and wherein the

conformational state of the sensor molecule influences the energy transfer efficiency. All RET methods are based on the use of compatible energy donor and acceptor pairs allowing RET to take place when donor and acceptor are in close proximity (<10nm). To be a compatible pair, the donor emission spectrum has to overlap with the acceptor excitation spectrum in order to gain energy transfer. The signal generating moiety may comprise the donor and acceptor pair, e.g. two fluorescent proteins. Alternatively, only the donor or acceptor is contained in the signal generating moiety, the other member of the pair being part of the regulatory molecule (see Figure 2A for an exemplary design). The donor and/or acceptor is suitably attached to the signal generating moiety via a self-labeling protein tag e.g. CLIP, SNAP, Halo-tag, or any of the other methods as described herein above. Alternatively, both the donor and acceptor have a proteinaceous character and are comprised in the amino acid sequence of the sensor (see Figure 2B). For example, a sensor may contain two fluorescent proteins or one fluorescent protein plus a luciferase enzyme.

In one specific embodiment, the signal generating moiety is capable of generating a fluorescence resonance energy transfer (FRET) signal. FRET is a distance-dependent interaction between the electronic excited states of two dye molecules in which excitation is transferred from a donor molecule to an acceptor molecule without emission of a photon. The efficiency of FRET is dependent on the inverse sixth power of the intermolecular separation, making it useful to monitor an analyte-induced conformational change in a sensor molecule of the invention. In most applications, FRET can be detected by the appearance of sensitized fluorescence of the acceptor or by quenching of donor fluorescence. Primary conditions for FRET are that donor and acceptor molecules must be in close proximity (typically 10-100 A) and that the absorption spectrum of the acceptor must overlap the fluorescence emission spectrum of the donor. In addition, donor and acceptor transition dipole orientations must be approximately parallel. For example, a sensor is provided wherein the signal generating moiety comprises a fluorescent acceptor and wherein the synthetic regulatory molecule comprises a fluorescent donor, or vice versa. Example 3 herein below demonstrates that the steric displacement concept of the present invention is advantageously used to improve the FRET-ratio change of a sensor molecule. Suitable FRET

donor/acceptor pairs for use in a sensor of the invention are known in the art.

A limitation of FRET is the requirement for external illumination to initiate the fluorescence transfer, which can lead to background noise in the results from direct excitation of the acceptor or to photobleaching. To avoid this drawback, a sensor of the invention is advantageously based on Bioluminescence Resonance Energy Transfer (BRET). In BRET, the donor fluorophore of the FRET pair is replaced by a bioluminescent donor protein (BDP) which, in the presence of a substrate, excites the acceptor fluorophore through the same resonance energy transfer mechanisms as FRET.

Accordingly, in one embodiment the signal generating moiety of a sensor provided herein comprises a BDP. Generally, the BDP is a luciferase, for example the luciferase from Renilla reniformis). Alternative BDPs that can be employed in this invention are enzymes which can act on suitable substrates to generate a luminescent signal. Specific examples of such enzymes are beta- galactosidase, alkaline phosphatase, beta-glucuronidase and beta-glucosidase. Synthetic luminescent substrates for these enzymes are well known in the art and are commercially available from companies, such as Tropix Inc. (Bedford, MA, USA).

In a preferred embodiment, the BDP has luciferase activity. Luciferases, and nucleic acid constructs encoding them, are available from a variety of sources or by a variety of means. Examples of bioluminescent proteins with luciferase activity may be found in U. S. Patent Nos. 5,229,285; 5,219,737; 5,843,746;

5,196,524; or 5,670,356. Preferred luciferases include NanoLuc luciferase, Renilla luciferase, firefly luciferase and Gaussia luciferase. Also encompassed are non- naturally occurring luciferases, e.g. a mutated luciferase. In a particular embodiment, a sensor of the invention comprises the previously described NanoLuc™ Luciferase (Nluc), a 19.1 kDa, monomeric, ATP independent enzyme that utilizes a novel substrate to produce high intensity, glow- type luminescence. See WO 2012/061530 and Hall et al. ACS Chem Biol.

2012;7(11): 1848-57. The enzyme was generated using directed evolution from a deep-sea shrimp luciferase, creating a luciferase that is much brighter than other forms of luciferase, including both firefly {Photinus pyralis) and Renilla reniformis. The high intensity luminescence of the NanoLuc enzyme combined with low autoluminescence of the furimazine substrate allows the sensitive detection of low levels of luciferase. The fluorescent acceptor molecule is chosen to function as BRET pair together with the BDP i.e. to accept the bioluminescence energy from the donor in the presence of an appropriate substrate. Furthermore, the fluorescent acceptor molecule is adapted to emit light after accepting the bioluminescence. The choice depends on luciferase emission spectrum and/or application of the sensor molecule. Suitable fluorescent acceptors to form a BRET pair include any fluorophore whose excitation spectra at least partially overlaps with the emission spectra of the respective luciferase. Tetherable fluorophores that can be used as luminescence acceptors in a sensor molecule of the invention comprising luciferase include Alexa Fluor dyes, in particular Alexa Fluor 488, Alexa Fluor 594; cyanine dyes such as Cy3, Cy3.5, Cy5, Cy7 and derivatives thereof, in particular sulfonated derivatives; SYTO dyes; SYBR dyes, Bodipy dyes; fluorescent proteins such as EGFP and mCherry; Atto Dyes such as Atto647N; rhodamine dyes such as carboxy- tetramethylrhodamine (TMR), Texas Red, silicon rhodamine; fluorescein derivatives such as carboxyfluorescein and FITC; Oregon Green; triarylmethane dyes as malachite green; naphthalimide dyes such as Lucifer Yellow; xanthene dyes such as SNARF-1; acridine dyes such as acridine orange; coumarins; IRDye stains such as IRDye 700DX. Very suitable acceptors include Cy3 and TMR.

The relative positioning of the BRET donor/acceptor pair within a sensor of the invention can vary. For example, the signal generating moiety comprises the BDP and the synthetic regulatory molecule comprises a fluorescent acceptor molecule. Other useful formats include the signal generating moiety comprising both the BDP and the fluorescent acceptor, either in the form of a fluorescent protein or a synthetic fluorophore tethered either using a self-labelling tag or by other means of attachment.

As will be appreciated by the skilled person, a sensor molecule of the invention is broadly applicable for the detection, quantification and/or imaging of any analyte of interest. The sensor finds its use in many different areas, ranging from clinical diagnostics, drug screening, and fundamental research. Provided is a method for in vitro detecting an analyte of interest in a sample, comprising the steps of: (a) contacting the sample with a sensor molecule according to the invention under conditions allowing for an analyte-induced conformational change within the sensor molecule to occur such that the signal generated by the signal generating moiety is modulated; and (b) analyzing a change in a signal generated by the signal generating moiety.

The analyte of interest is a substance for which the presence, absence, location and/or quantity is to be determined. The analyte can have a chemical, biological, synthetic or semi- synthetic nature. It can be a small molecule or a protein. For example, the analyte of interest is a drug, a metabolite, an inhibitor, a protein, a biomarker or a nucleic acid molecule. Typically, the analyte of interest has biological or pharmacological activity. Biologically active analytes are often of biological origin themselves, but can also be of synthetic or semi- synthetic origin.

In one embodiment, the sample is a biological sample or a fraction thereof, preferably a bodily fluid, more preferably selected from the group consisting of blood, serum, saliva, urine, spinal fluid, pus, sweat, tears, breast milk, (in text: wherein the sample absorbs light in the blue light region). In another embodiment, the sample comprises cells.

In one embodiment, a sensor or method of the invention is used for drug quantification. For many drugs, finding the balance between efficacy and toxicity requires quantification of their concentrations in the patient's blood. Many immunosuppressants, anti-epileptics, antibiotics and others have unpredictable pharmacokinetics and narrow therapeutic ranges, requiring monitoring of their concentrations in the body. This process, known as therapeutic drug monitoring (TDM), currently relies on immunoassays and chromatographic techniques but such methods require dedicated personnel and infrastructure. The development of fast and low-cost assays would improve safety and therapeutic outcome in regions with poor infrastructure and allow personalized dosage at bedside or at home. Quantifying drug levels at bedside or at home would have obvious advantages in terms of therapeutic outcome and convenience, but current techniques require the setting of a diagnostic laboratory. Moving TDM from the diagnostic lab to the patient requires tools that (i) are capable of handling minimal sample volumes down to a single drop, (ii) are quantitative, and (iii) permit readout with inexpensive devices. It was found that a bioluminescent sensor described herein permits precise measurements of drug concentrations in patient samples e.g. by spotting minimal volumes on paper and recording the signal using a simple digital camera. The sensors can be readily engineered to selectively recognize a wide range of drugs, including immunosuppressants, antiepileptics, anticancer agents, and antiarrhythmics. This low-cost point-of-care method could make therapies safer, increase the convenience of doctors and patients, and make therapeutic drug monitoring available in regions with poor infrastructure.

Another specific aspect of the invention relates to the use ratiometric RET sensors for quantification of analytes in complex samples that absorb light at the emission wavelengths of the sensor, e.g. serum or other bodily fluids. Analysis of such samples is prone to artefacts and often leads to unreliable assay outcomes. Whereas sensors based on luciferases as an internal light source (i.e. BRET) would in theory reduce the fluorescent background problem and potentially increase sensitivity, no ratiometric BRET-based sensors have yet been introduced that are suitably used for the quantification of analytes in light-absorbing samples.

The fact that no BRET-based, portable, mix-and-measure sensors for precise point-of-care quantification of analytes (e.g. for therapeutic drug

monitoring) are currently available despite major developments in (medical) applications of bioluminescence technology is illustrative of the technical difficulties encountered to generate such sensors. It was surprisingly found that by absorbing a BRET sensor of the invention to a solid carrier such as paper or by immobilizing the BRET sensors prior to measurement to a solid carrier such as a glass surface, interference from absorbance of the sample at the emission wavelength of the sensor is minimized. This then allows for analysis of complex samples, like serum. The invention therefore also provides an analytical device comprising a BRET sensor molecule as described herein above, wherein the sensor molecule is arranged in such a manner that, when the device is in use for detecting an analyte of interest in a sample, the photons that are emitted from the sensor molecule and that are collected by a detector pass through the sample for a distance shorter than 330 μηι. In one embodiment, the sensor molecule is immobilized or absorbed to a solid carrier, preferably a glass or transparent plastic. Alternatively, the sensor molecule is absorbed to a paper carrier or a gel, preferably to chromatography or filter paper. In yet another embodiment, the sensor molecule is comprised in a thin film, or confined in a tube, capillary or (microfluidic) chamber.

Still further applications of the invention relate to the in vivo targeting and/or imaging of cells. For example, cancer cells expressing a cancer- specific cell surface marker like a receptor can be targeted by a sensor of the invention carrying as secondary ligand that is specific for such cancer marker. Thus, the analyte of interest is not necessarily a free compound in solution. Sensor molecules based on a luciferase reporter enzyme which is inhibited by LI in the absence of analyte are particularly suitable for such application in view of their negligible background emission and the established use of luciferase for in vivo imaging. Hence, a sensor molecule of the invention, in particular one being based on the TURNON design, is suitably used as optical probe. The invention also provides a method for detecting a cell of interest, comprising the steps of contacting a sample known or suspected of comprising the cell with a sensor molecule according to the invention wherein L2 is capable of specifically binding to at least one surface marker expressed on said cell, under conditions allowing for an analyte-induced conformational change within the sensor molecule to occur such that the signal generated by the signal generating moiety is modulated; and analyzing a change in a signal generated by the signal generating moiety. In one embodiment, the cell is a cancer cell. The sample can be a clinical sample, like a tissue biopsy or bodily sample , or a research sample e.g. a cell culture. The invention also relates to a method for providing a sensor molecule of the invention. As is illustrated in Examples 3-7, the proteinaceous moiety and the synthetic regulatory molecule (or precursor thereof) are typically produced as separate entities, after which the synthetic molecule is tethered to the

proteinaceous molecule using the appropriate coupling reaction. Hence, the method comprises the steps of providing the proteinaceous moiety and the synthetic regulatory molecule or precursor thereof, and assembling both to yield the sensor molecule.

The proteinaceous moiety can be prepared using standard recombinant DNA techniques well known to those skilled in the art. For example, the nucleic acid sequence coding the proteinaceous region comprising BP1 can be genetically introduced into the multiple cloning site of a bacterial expression vector comprising a luciferase sequence such that the BP1 sequence is operatively linked to the Luc coding sequence. Other proteinaceous components, like a protein labeling tag and/or linker sequences, can also be incorporated using standard techniques. The DNA constructs for various configurations of the proteinaceous moiety of a sensor of the invention can be transfected/transformed in suitable cell lines (eukaryotic or prokaryotic) for its production. The various configurations of the fusion proteins produced in cells, are then purified or semipurified from the

transfected/transformed cells. A convenient procedure to purify a proteinaceous moiety is by affinity chromatography e.g. using a His- and/or Strep-tag engineered in the DNA construct. Standard biochemical techniques can be also used alone or in combination with affinity chromatography to purify to various levels the various fusion proteins. Finally, these purified fusion proteins can be also chemically or enzymatically modified before their tethering to the synthetic regulatory molecule.

In another embodiment, the proteinaceous moiety is produced by a combination of in vivo and in vitro methods. First a fusion protein is genetically engineered and expressed in cells using recombinant techniques. The fusion protein is then purified or semi-purified before being modified by chemically or enzymatically attaching a further proteinaceous element, e.g. an element which can serve as a binding protein such as an antibody. Attachment of the further element can be peptide-based or chemically-based.

The synthetic regulatory molecule or precursor thereof can be synthesized using methods known in the art. For example, the primary and the secondary ligand are linked in close proximity to each other and appended to the fluorophore-0⁶-benzylguanine derivative by chemical synthesis. The skilled person will understand that the methods used can be selected based on the chemical nature of the fluorophore and/or the ligand(s). The coupling of each element can essentially be performed according to what has been described in the art on conventional FRET-based SNIFITs. Also, the regulatory molecule or precursor thereof may contain an element which mediates tethering to the proteinaceous moiety. For example, if the synthetic regulatory molecule is to be site-specifically tethered to the proteinaceous moiety of the sensor molecule via a self-labelling protein such as SNAP-tag, CLIP-tag or Halo-Tag, the synthetic regulatory molecule must contain the appropriate reactive group such as a reactive group for hAGT, a 0⁶-benzylguanine (BG), 0⁴-benzyl-2-chloro-6-aminopyrimidine (CP), O²- benzylcytosine (BC) derivative, or a chloroalkane. Reactive groups mediating tethering may be advantageously coupled to the fluorophore acceptor molecule via spacer comprising several polyethylene glycol (PEG) units. For example, a spacer of 3-27 PEG units is suitably used. See for example Brun et al. J Am Chem Soc.

2009;131(16):5873-84, and the examples herein below.

A regulatory molecule to be used in combination with cysteine or enzyme-mediated coupling can be synthesized based on the examples below, wherein the BG is exchanged with a maleimide for cysteine coupling, or with a CoA derivative for coupling via phosphopantetheine transferases. LEGENDS TO THE FIGURES

Figure 1. Pictorial description of the sensing mechanism of exemplary sensor molecules. The ligands, which can be synthetic or proteinaceous, are denoted LI and L2 and their corresponding binding partners are labelled BPl and BP2. X represents the attachment position of the synthetic regulatory molecule. (A) "Normal steric displacement" with both LI and L2 situated on the synthetic regulatory molecule. (B) "Competition steric displacement" with both LI and L2 situated on the synthetic regulatory molecule. BP2 can either be a separate entity or part of the sensor molecule (represented by dashed line). (C) "Normal steric displacement" with LI being part of the synthetic regulatory molecule and L2 situated on the signal generating moiety. (D) "Competition steric displacement" with LI being part of the synthetic regulatory molecule and L2 situated on the signal generating moiety. BP2 can either be a separate entity or part of the sensor molecule (represented by dashed line). Figure 2. Examples of sensing mechanisms of SD-SNIFITs. For simplicity, only sensor molecules based on "normal steric displacement" are shown; the readout for "competition steric displacement" is analogous. (A) RET-based readout with either the donor or acceptor fluorophore being part of the synthetic regulatory molecule and the RET partner in the form of a second synthetic fluorophore, a fluorescent protein, or a luciferase being part of the signal generation moiety. Displacement of LI from BPl leads to a change in RET efficiency. (B) RET-based readout with both RET partners being part of the signal generation moiety. The RET partners can be synthetic fluorophores, fluorescent proteins, or a luciferase. Displacement of LI from BPl leads to a change in RET efficiency. (C) TURN-ON reporter. The binding partner of LI is an enzyme (E) that is capable of generating a detectable signal. LI can be an inhibitor that reduces enzymatic activity. Binding of L2 to BP2 prevents LI from inhibiting E and a signal is observed.

Figure 3. Exemplary sensor molecule for detecting streptavidin based on "normal steric displacement" and BRET as readout. (A) Pictorial description of the structure and sensing mechanism of the sensor molecule. The signal generating moiety comprises human carbonic anhydrase (HCA) as BPl, NanoLuc luciferase as BRET donor, a 30-proline linker for geometrical optimization, and SNAP-tag as attachment position for the synthetic regulatory molecule. The synthetic regulatory molecule comprises BG, which is recognized by SNAP-tag, the BRET acceptor Cy3, an aromatic sulfonamide (SA) as LI, and biotin (b) as L2. Binding of streptavidin (strept) to L2 leads to opening of the sensor and a decrease in BRET efficiency. (B) Chemical structure of the synthetic regulatory molecule. (C) Luminescence emission spectra of the sensor molecule in the presence of varying streptavidin concentrations. (D) Response curve of the sensor molecule titrated with

streptavidin. When the sensor lacks the secondary ligand L2 (BG-Cy3-SA) no response is observed. For details see Example 1.

Figure 4. Exemplary sensor molecule for detecting methotrexate based on

"competition steric displacement" and BRET as readout. (A) Pictorial description of the structure and sensing mechanism of the sensor molecule. Both the signal generation moiety and the synthetic regulatory molecule are identical to the sensor molecule described in Figure 3, except for the secondary ligand L2, which is a derivative of trimethoprim (T / TMP) in this case. Defined concentrations of DHFR are added as a binding partner of the analyte of interest, methotrexate (M). High concentrations of M can displace T from DHFR allowing LI to bind to HCA. (B) Structure of the synthetic regulatory molecule. (C) Response of the sensor molecule titrated with methotrexate in the presence of different defined DHFR

concentrations. The transition from blue to red is visible by naked eye but was enhanced in this image to make it visible in grayscale. For details see Example 2.

Figure 5. Exemplary sensor molecule for detecting acetylcholine esterase inhibitors based on "competition steric displacement" and FRET as readout. (A) Pictorial description of the structure and sensing mechanism of the sensor molecule. The signal generation moiety consists of HCA as BPl, CLIP-tag labelled with Cy3 as FRET donor, a 30-proline linker for geometrical optimization, SNAP-tag as attachment position for the synthetic regulatory molecule, and acetylcholine esterase (AChE) as BP2 - binding partner for the analyte of interest. The signal generation moiety is tethered to the surface of mammalian cells. The synthetic regulatory molecule consists of BG, which is recognized by SNAP-tag, the FRET acceptor Cy5, an aromatic sulfonamide (SA) as LI, and edrophonium (e / edro) as L2. The analyte tacrine (T) can displace the secondary ligand e from AChE and allow SA to bind to HCA, leading to an increase in FRET- efficiency. (B) Structure of the synthetic regulatory molecule. (C) Response of the sensor on the surface of mammalian cells perfused with the analyte tacrine. For details see Example 3.

Figure 6. Exemplary TURN-ON reporter for detecting streptavidin based on "normal steric displacement" and bioluminescence as readout. (A) Pictorial description of the structure and sensing mechanism of the sensor molecule. The signal generation moiety consists of Renilla luciferase as enzyme, E, and SNAP-tag as attachment position for the synthetic regulatory molecule. The synthetic regulatory molecule consists of BG, which is recognized by SNAP-tag,

coelenteramide (coent) as LI and inhibitor of the luciferase, and biotin (biot) as L2. Binding of the analyte streptavidin (strept) to L2 prevents coent from inhibiting the luciferase and a signal can be observed. (C) Picture of the sensor molecule in the absence and presence of 1 μΜ streptavidin. The change in bioluminescence intensity can be seen by eye. (D) Quantification of the bioluminescence intensity produced by the sensor molecule in the absence and presence of 1 μΜ streptavidin.

EXPERIMENTAL SECTION

The Examples below illustrate the design and construction of exemplary BRET sensors according to the invention and the use thereof in an analytical device or in an analytical method. Reagents and solvents were purchased from Sigma Aldrich (St. Louis, MO) or Acros Organics (Waltham, MA) and used without further purification. Peptide couplings were performed by activation of the respective carboxylic acid with 0-(N-Succinimidyl)-N,N,N',N'-tetramethyluronium tetrafluoroborate (TSTU) or N,N,N',N'-Tetramethyl-0-(lH-benzotriazol-l- yl)uronium hexafluorophosphate (HBTU) in the presence of diisopropylethylamine (DIEA) as base in anhydrous dimethylsulfoxide (DMSO) at room temperature. EXAMPLE 1: Sensor molecule for streptavidin or biotin sensing using biotin as secondary ligand

The basic principle underlying the Steric Displacement- SNIFITs is the mutually exclusive binding of two tethered ligands due to steric hindrance between the two protein targets. The two ligands, the primary and the secondary ligand, are linked in close proximity to each other and appended to the fluorophore-O⁶- benzylguanine derivative by chemical synthesis. The BG -fluorophore- double ligand molecule can be used to label an existing sensor protein of an optimized SNIFIT, retaining its original optimized FRET-ratio change, but making it sensitive to the analyte protein of interest. In absence of the analyte protein, the primary ligand is bound to the receptor protein and strong resonance energy transfer occurs. The analyte protein at a sufficient concentration can bind to the secondary ligand, but the steric hindrance between the analyte protein and the receptor protein leads to unbinding of the primary ligand from the receptor protein leading to a detectable decrease in RET-efficiency. This allows to adapt a previously optimized probe to virtually any different analyte, with minimal synthetic effort. SD-SNIFITs are not anymore limited to sense small molecules and we were able to extend the field of application to whole protein analytes.

A BRET sensor capable of sensing streptavidin concentrations was constructed by introducing biotin as a secondary ligand into a human carbonic anhydrase II (HCA) -based sensor molecule (see Figure 3A).

A molecule containing an 0⁶-benzylguanine (BG) group for SNAP-tag labeling, the fluorophore Cy3, a sulfonamide as primary ligand, and biotin as secondary ligand was synthesized according to scheme 1.

Scheme 1: Schematic representation of the synthesis of the molecule BG-Cy3-SA- biot.

4-carboxybenzene sulfonamide (A-l) was coupled to N-alpha-Boc-(L)- 2,3-diaminopropanoic acid (A-2) to afford A-3, that was deprotected by stirring at room temperature in trifluoroacetic acid (TFA) to obtain A-4. The product was then coupled to biotin to give the bifunctional ligand A-5. that was subsequently tethered to a short PEG2 tether by peptide coupling to l-N-Boc-3,6-dioxa-l,8- diaminooctane (A-6) to afford A-7. Deprotection was performed by stirring in TFA at room temperature to give A-8. BG-EG11-NH2 (A-10) and Cy3 (A-9) were prepared as previously described (Mujumdar et al. Bioconjugate Chemistry 1993, 4, 105-111 Masharina et al. J Am Chem Soc. 2012;134(46):19026-34) and the three building blocks were coupled together with A-8 to give the polyfunctional molecule BG-Cy3-SA-Biot (A-ll).

A fusion protein of SNAP-tag, a 30-proline linker, NanoLuc luciferase (Promega, Fitchburg, WI) and HCA was constructed by replacing the coding sequence of CLIP-tag in the previously described sensor SNAP-PP30-CLIP-HCA (Brun et al. J Am Chem Soc. 2011;133(40):16235-42) by the coding sequence of NanoLuc luciferase using standard cloning techniques. The fusion protein was expressed in the E. coli strain Rosetta-gami and purified using a C-terminal His- tag as well as an N-terminal Strep-tag.

The sensor molecule was assembled by labeling SNAP-tag with the synthetic molecule BG-Cy3-SA-biot (Figure 3B). The purified protein was diluted to a concentration of 1 μΜ in HEPES buffer (50mM HEPES, 50mM NaCl, pH 7.2) and incubated with a 4-fold molar excess of the synthetic compound BG-Cy3-SA-biot for 1 hour at room temperature. Excess of the synthetic compound was removed using centrifugal filter devices (Amicon Ultra-0.5, Merck Millipore, Billerica, MA).

To test the response to different streptavidin concentrations, the assembled sensor molecule was diluted to a concentration of 10 nM in 100

HEPES buffer containing defined concentrations of streptavidin in white non- binding 96-well plates (Greiner Bio-One, Kremsmiinster, Austria). The solutions were incubated at room temperature for at least 10 minutes to ensure that the sensor had reached equilibrium. Bioluminescence was measured on an EnVision Multilabel Reader (Perkin Elmer): 5 seconds before the measurement, 100 1 g/mL coelenterazine-h (NanoLight, Pinetop, AZ) in HEPES buffer was added into the wells using the instrument's injector and the signal was collected using an emission filter for Umbelliferone (wavelength: 460 nm, bandwidth: 25 nm) to record NanoLuc emission and a filter for Cy3 (wavelength: 595 nm, bandwidth: 60 nm) to record Cy3 emission. Figures 3C and 3D show the response of the sensor to different streptavidin concentrations. At low concentrations, the sensor is in its closed conformation, permitting efficient resonance energy transfer from NanoLuc to Cy3 and leading to a low NanoLuc / Cy3 emission ratio. At high streptavidin

concentrations, biotin is bound to streptavidin, preventing sensor closing. In this open conformational state, resonance energy transfer from NanoLuc to Cy3 is inefficient, leading to high NanoLuc / Cy3 emission ratios.

This shows that a sensor of the invention can be used for an accurate

determination of streptavidin. HCA-binding molecules such as sulfonamide drugs can in principle interfere with the streptavidin quantification and this should be taken into account for some application, but the presence of both the compounds is rather uncommon. Interestingly, it is also possible to adapt the sensor to quantify biotin or biotinylated proteins, because the presence of an unknown concentration of biotin derivative in the sample will require a higher amount of streptavidin to open the sensor because of the competition with the free biotin derivative. The biotin concentration can thus be calculated from the shifting observed in the titration curve compared to a blank titration performed in the absence of biotin derivative. This approach has interesting clinical applications. Biotin is a coenzyme for five carboxylases in the human body (propionyl-CoA carboxylase,

methylcrotonyl-CoA carboxylase, pyruvate carboxylase, and 2 forms of acetyl-CoA carboxylase. Therefore, biotin is essential for amino acid catabolism,

gluconeogenesis, and fatty acid metabolism. Biotin is also necessary for gene stability because it is covalently attached to histones. Biotinylated histones play a role in repression of transposable elements and some genes. Normally, the amount of biotin in the body is regulated by dietary intake, biotin transporters

(monocarboxylate transporter 1 and sodium- dependent multivitamin transporter), peptidyl hydrolase biotinidase (BTD), and the protein ligase holocarboxylase synthetase. When any of these regulatory factors are inhibited, biotin deficiency could occur. Since biotin is in many foods at low concentrations, deficiency is rare except in locations where malnourishment is very common. Pregnancy, however, alters biotin catabolism and despite a regular biotin intake, half of the pregnant women in the U.S. are marginally biotin deficient. If biotin metabolism is defective, all four carboxylases will be deficient. Biotin is covalently linked to a key lysine residue in each carboxylase by action of

holocarboxylase synthetase. When the carboxylase proteins are degraded, biotinoyl-lysine is subsequently cleaved by biotinidase releasing free biotin that can be reutilized. The two defects in biotin metabolism associated with Multiple Carboxylase Deficiency are caused by deficient activity of holocarboxylase synthetase and biotinidase. The disorders tend to present clinically at different ages, with holocarboxylase synthetase deficiency being known as early-onset (neonatal) multiple carboxylase deficiency and biotinidase deficiency referred to as late-onset multiple carboxylase deficiency. Both respond to biotin supplementation. A sensor molecule of the present invention is advantageously used in support of diagnosis, treatment, and other clinical decision- making.

EXAMPLE 2: Methotrexate Sensor

A BRET sensor capable of sensing methotrexate concentrations was constructed by introducing the dihydrofolate reductase (DHFR) inhibitor trimethoprim as a secondary ligand into a human carbonic anhydrase II (HCA)- based sensor molecule. DHFR binds to this secondary ligand, keeping the sensor in its open state. Methotrexate competes for binding to DHFR and can thus be detected by competition steric displacement (see Figure 4A). A molecule containing an 0⁶-benzylguanine (BG) group for SNAP-tag labeling, the fluorophore tetramethylrhodamine (TMR), benzenesulfonamide as primary ligand, and trimethoprim as secondary ligand was synthesized according to scheme 2. 4-Demethyltrimethoprim (B-l) was alkylated with methyl 5- bromopentanoate (B-2) in the presence of anhydrous potassium carbonate in dimethylformamide (DMF). The reaction mixture was then poured in 1 M aqueous sodium hydroxyde to give B-3. The derivative was then coupled to compound A-4 resulting in the bifunctional molecule with benzenesulfonamide and trimethoprim B-4. A short PEG2 linker was added by coupling with A-6, followed by acidic deprotection to give the derivative B-6. Scheme 2: Schematic representation of the synthesis of the regulatory molecule BG-TMR-SA-tmp

B-S

BG-EG11-TMR-COOH (B-7) was prepared as previously described (Brun et al. J Am Chem Soc. 2009;131(16):5873-84; Kvach et al. Bioconjug Chem. 2009, 20(8), 1673-82) was subsequently coupled to B-6 to give the polyfunctional molecule BG-TMR-SA-tmp (B-8).

A fusion protein of SNAP-tag, a 30-proline linker, NanoLuc luciferase (Promega, Fitchburg, WI) and HCA was constructed by replacing the coding sequence of CLIP-tag in the previously described sensor SNAP-PP30-CLIP-HCA (Brun et al. J Am Chem Soc. 2011;133(40):16235-42) by the coding sequence of NanoLuc luciferase using standard cloning techniques. The fusion protein was expressed in the E. coli strain Rosetta-gami and purified using a C-terminal His- tag as well as an N-terminal Strep-tag. The sensor molecule was assembled by labeling SNAP-tag with the synthetic molecule BG-TMR-SA-tmp (Figure 4B). The purified protein was diluted to a concentration of 1 μΜ in HEPES buffer (50mM HEPES, 50mM NaCl, pH 7.2) and incubated with a 4-fold molar excess of the synthetic compound BG-TMR-SA- tmp for 1 hour at room temperature. Excess of the synthetic compound was removed using centrifugal filter devices (Amicon Ultra-0.5, Merck Millipore, Billerica, MA).

To sense methotrexate, the assembled sensor molecule was diluted to a concentration of 10 nM in 100

HEPES buffer supplemented with 100 μΜ

NADPH, a defined concentration of DHFR, and different concentrations of methotrexate in white non-binding 96-well plates (Greiner Bio-One,

Kremsmiinster, Austria). The solutions were incubated at room temperature for at least 15 minutes to ensure that the sensor had reached equilibrium and a picture was taken with a Canon EOS 600D digital camera. A sharp transition was clearly visible between the spots containing methotrexate concentrations below and above that of DHFR (Figure 4C). We performed the titration with different DHFR concentrations and demonstrated that the amount of methotrexate required for the transition to occur varies accordingly. Due to the strong interaction of trimethoprim with DHFR, the sensor molecules are quantitatively bound to the free DHFR via the secondary ligand, thus preventing the binding of the primary ligand to HCA and resulting in a low BRET-state (blue color observed). Methotrexate interacts with DHFR even stronger than

trimethoprim, thus when the methotrexate concentration exceeds the DHFR concentration threshold, there is no more free DHFR available to bind to the secondary ligand and the primary ligand can bind to HCA, bringing the sensor molecule in its high-BRET state (red color observed). A sharp transition is only observed if the sensor concentration is much lower than the concentration of free DHFR. This condition is met since the high brightness of the sensor molecule allows to easily distinguish its color by naked eye even at a concentration as low as 10 nM, thus well below the bottom end of the therapeutic concentration range of methotrexate. EXAMPLE 3: Acetylcholine Sensor

A FRET sensor capable of sensing the concentration of acetylcholine as well as acetylcholine esterase inhibitors was constructed by introducing

decani ethonium as a secondary ligand into a human carbonic anhydrase II (HCA)- based sensor molecule to which acetylcholine esterase had been fused (see Figure 5A). Acetylcholine esterase binds intramolecularly to the secondary ligand, keeping the sensor in its open conformation until it is displaced by acetylcholine or related molecules.

A synthetic regulatory molecule containing an 0⁶-benzylguanine (BG) group for SNAP-tag labeling, the fluorophore Cy5, a sulfonamide as primary ligand, and the acetylcholine esterase inhibitor edrophonium as secondary ligand was synthesized according to scheme 3.

The synthetic procedure involved three convergent steps. The edrophonium derivative was prepared by reacting Phloroglucinol C-l with dimethylamine to obtain C-2 that was alkylated in DMF with 10-azido-l- bromodecane in the presence of potassium carbonate to afford C-3. Treatment with ethyl iodide in acetonitrile gave the cationic C-4 that was coupled to

propargylamine via click-chemistry using copper(II) sulfate, tris[(l-benzyl-lH- l,2,3-triazol-4-yl)methyl]amine, sodium ascorbate in DMSO to introduce the free amine in C-5. The benzenesulfonamide derivative was prepared by coupling 2-N- Boc-diaminopropionic acid (C-6) with 4-carboxy-benzenesulfonamide (C-7) to obtain C-8. C-5 and C-8 were coupled using TSTU to afford the bifunctional ligand C-9, that was deprotected from the Boc group by treatment with TFA to obtain C- 10. BG-EG11-NH2 (A-10) and Cy5 (C-ll) were prepared as previously described (Mujumdar et al. Bioconjugate Chemistry 1993, 4, 105-111 Masharina et al. J Am Chem Soc. 2012;134(46):19026-34) and were coupled together also with 6- aminohexanoic acid C-12 to give C-13. C-13 and C-10 were coupled using TSTU to obtain the polyfunctional molecule BG-Cy5-SA-edro (C-14). Scheme 3: Schematic representation of the synthesis of the regulatory molecule BG-Cy5-SA-edro.

A fusion protein of acetylcholine esterase, SNAP-tag, a 30-proline linker, CLIP-tag and HCA was constructed by inserting the coding sequence of acetylcholine esterase into the previously described sensor SNAP-PP30-CLIP-HCA in pDisplay (Brun et al. J Am Chem Soc. 2011;133(40):16235-42) using standard cloning techniques. The plasmid encoding for the senor was transfected into HEK293 cells using Lipofectamine 2000 (Life Technologies, Carlsbad, CA) following the supplier's recommendations. The sensor molecule was assembled on the cell surface by labeling SNAP-tag with the synthetic molecule BG-Cy5-SA-deca (Figure 5B) as well as CLIP-tag with the fluorophore CLIP- Surface 547 (New England Biolabs, Ipswich, MA). The response of the sensor to inhibitors of acetylcholine esterase was then investigated by live-cell microscopy. The methods for labelling and imaging have been previously described (Brun et al. J Am Chem Soc. 2011;133(40):16235-42).

Figure 5C shows the response of the sensor to perfusion of the cells with tacrine. In the absence of tacrine, the secondary ligand is bound to acetylcholine esterase. In this state, resonance energy transfer from CLIP- Surface 547 to Cy5 is inefficient, leading to a high FRET ratio. In the presence of tacrine, the secondary ligand is displaced making it possible for the primary ligand to bind HCA, bringing the two fluorophores closely together. In this state, resonance energy transfer is efficient, leading to a high FRET ratio.

EXAMPLE 4: TURN-ON Sensor Molecule

Thus far, the SNIFIT concept was limited to the change in FRET- or

BRET-efficiency as readout. In contrast, the steric displacement approach of the present invention can allow the quantification of metabolites or proteins by any detection means. In this example, a reporter enzyme is used as signal generating moiety and tethered as primary ligand an inhibitor of such reporter enzyme. As secondary ligand we chose a binding partner of the analyte protein to be detected. When no analyte protein is present, the reporter enzyme is in its off-state, because the primary ligand inhibits it. The binding of the analyte protein to the secondary ligand, sterically displaces the primary ligand, resulting in activation of the reporter enzyme. We call this sensor design TURN-ON reporter. A turn-on luciferase sensor capable of sensing streptavidin

concentrations was constructed by labeling a fusion of SNAP-tag and Renilla luciferase with a synthetic regulatory molecule that contains the luciferase inhibitor coelenteramide as a primary ligand and biotin as a secondary ligand (see Figure 6A,B). A molecule containing an 0⁶-benzylguanine (BG) group for SNAP-tag labeling, coelenteramide, an inhibitor of Renilla luciferase, as primary ligand and biotin as secondary ligand was synthesized according to scheme 4. Coelenteramide was purchased from NanoLight (Pinetop, AZ).

2-Amino-3-azido-propionic acid (D-l) was coupled to biotin to afford D-2. BG-EGn- NH2 (A-10) was coupled to D-2 to obtain the azido-modified 0⁶-benzylguanine- biotinylated derivative D-3. Separately, coelenteramide (D-4) was O-alkylated with propargyl bromide with sodium carbonate in dimethylformamide to give the O- propargyl-coelenteramide D-5. Click-chemistry with copper(II) sulfate, tris[(l- benzyl-lH-l,2,3-triazol-4-yl)methyl]amine, sodium ascorbate in DMSO was used to couple D-5 to D-3 and form the polyfunctional labeling compound BG-coel-biot (D- 6).

Scheme 4: Schematic representation of the synthesis of the regulatory molecule BG-coel-biot.

A fusion protein of SNAP-tag and humanized Renilla luciferase was constructed by replacing the coding sequence of the 30-proline linker, CLIP-tag, and HCA in the previously described sensor SNAP-PP30-CLIP-HCA (Brun et at. J Am Chem Soc. 2011;133(40):16235-42) by the coding sequence of humanized Renilla luciferase using standard cloning techniques. The fusion protein was expressed in the E. coli strain Rosetta-gami and purified using a C-terminal His- tag as well as an N-terminal Strep-tag. The sensor molecule was assembled by labeling SNAP-tag with the synthetic molecule BG-coel-biot (Figure 6B). The purified protein was diluted to a concentration of 1 μΜ in HEPES buffer (50mM HEPES, 50mM NaCl, pH 7.2) and incubated with a 4-fold molar excess of the synthetic compound BG-coel-biot for 1 hour at room temperature. Excess of the synthetic compound was removed using centrifugal filter devices (Amicon Ultra-0.5, Merck Millipore, Billerica, MA).

To sense streptavidin, the assembled sensor molecule was diluted to a concentration of 50 nM in 100 HEPES buffer containing different

concentrations of streptavidin in white non-binding 96-well plates (Greiner Bio- One, Kremsmiinster, Austria). The solutions were incubated at room temperature for at least 10 minutes to ensure that the sensor had reached equilibrium.

Bioluminescence was measured on an EnVision Multilabel Reader (Perkin Elmer): 5 seconds before the measurement, 100 1 g/mL coelenterazine-h (NanoLight, Pinetop, AZ) in HEPES buffer was added to the wells using the instrument's injector and the luminescence intensity was measured using an integration time of 1 second.

Figure 6D shows the response of the sensor to streptavidin. In the absence of streptavidin, the primary ligand is bound to the luciferase and inhibits the enzyme. When streptavidin is added, it binds to biotin and displaces the primary ligand from the luciferase, bringing it into an active state and leading to a 17-fold increase in luminescence intensity. This change is also clearly visible by naked eye; a picture taken with a digital camera is shown in Figure 6C.

The TURN-ON luciferase sensor shows negligible background emission and the application of luciferases for in vivo imaging has been widely demonstrated (Roda et. al. Trac-Trend Anal Chem 2009;28:307). This allows its application for targeting and visualizing specific cells in living animals, e.g. cancer cells expressing a specific receptor and tethering a secondary ligand specific for such a receptor. Moreover, the TURN-ON reporter concept is not limiting to

bioluminescence but has a very wide application. An enzyme used for colorimetric assay can be used in place of the luciferase with one of its inhibitors as primary ligand, to provide a colorimetric assay to determine the concentration or the activity of any analyte protein of interest, by tethering one of its binders as secondary ligand. A redox sensor is envisioned analogously.

Claims

Claims

1. A proteinaceous sensor molecule for detecting an analyte of interest (A), comprising a signal generating moiety tethered to a synthetic regulatory molecule capable of modulating signal generation by intramolecular binding, wherein (i) the synthetic regulatory molecule comprises a primary ligand (LI) capable of intramolecular binding to a primary partner (BPl) on the sensor molecule;

(ii) the sensor molecule comprises a secondary ligand (L2) capable of binding to a secondary binding partner (BP2), BP2 being A or a binding partner thereof ; and

(iii) binding of LI and L2 to their respective binding partners is mutually exclusive, such that binding of BP2 to L2 influences binding of LI to BPl, resulting in a conformational change within the sensor molecule such that the signal is modulated.

2. Sensor molecule according to claim 1, wherein both LI and L2 are part of the synthetic regulatory molecule.

3. Sensor molecule according to claim 1, wherein L2 is specifically tethered to a region of the sensor molecule that contains BPl.

4. Sensor molecule according to claim 3, wherein L2 is a proteinaceous ligand that is part of the protein sequence of the sensor comprising BPl.

5. Sensor molecule according to any one of claims 1-4, wherein the secondary ligand (L2) is capable of binding to BP2, BP2 being the analyte of interest (A), such that in the absence of analyte A, L2 is free and LI is bound to BPl and wherein the binding of analyte A to L2 displaces LI from BPl resulting in an open conformation state of the sensor molecule such that the signal generation is modulated.

6. Sensor molecule according to any one of claims 1-4, wherein the secondary ligand (L2) is capable of binding to BP2, BP2 being a binding partner of the analyte of interest (A), such that in the absence of analyte A L2 is bound to BP2 and LI is free and wherein the binding of analyte A to BP2 displaces BP2 from L2 allowing LI to bind to BP1 resulting in a closed conformation state of the sensor molecule such that the signal generation is modulated.

7. Sensor molecule according to claim 6, wherein BP2 is part of the proteinaceous sensor molecule and binding of L2 to BP2 is intramolecular.

8. Sensor molecule according to any of the preceding claims, wherein the synthetic regulatory molecule is site-specifically tethered to the signal generating moiety, preferably via a natural or unnatural amino acid or via a protein labelling tag.

9. Sensor molecule according to claim 8, wherein the protein labelling tag is a self-labelling protein, such as SNAP-tag, CLIP-tag or Halo-Tag, and wherein the synthetic regulatory molecule is tethered via the appropriate reactive group.

10. Sensor molecule according to claim 8, wherein the protein labelling tag is a tag that is labelled through the action of an enzyme, such as sortase, lipoic acid ligase, or phosphopantetheine transferase.

11. Sensor molecule according to claim 8, wherein the amino acid is cysteine or an unnatural amino acid that allows for a site- specific chemical conjugation reaction, such as click-chemistry.

12. Sensor molecule according to any one of the preceding claims, wherein the signal generating moiety comprises an enzyme (E) capable of generating a detectable signal.

13. Sensor molecule according to claim 12, wherein E is selected from the group of enzymes catalyzing a chromogenic or fluorogenic reaction and redox enzymes, preferably wherein E is a peroxidase, luciferase, hydrolase, glucose oxidase or beta-galactosidase.

14. Sensor molecule according to claim 12 or 13, wherein

(i) the synthetic regulatory molecule comprises a primary ligand (LI) capable of intramolecular binding to BP1 and wherein BP1 is comprised within E;

(ii) binding of LI to the BP1 leads to inhibition of the corresponding enzymatic activity of E; and

(iii) binding of LI and L2 to their respective binding partners is mutually exclusive, such that binding of BP2 to L2 influences binding of Llto E, resulting in a change of the enzymatic activity of E within the sensor molecule such that signal generation is modulated.

15. Sensor molecule according to any one of claims 1-14 , wherein the signal generating moiety is capable of generating a resonance energy transfer (RET) signal, optionally together with a fluorescent group contained in the synthetic regulatory molecule, and wherein the conformational state of the sensor molecule influences the energy transfer efficiency.

16. Sensor molecule according to claim 15, wherein a fluorescent donor or acceptor is attached to the signal generating moiety via a self-labeling protein tag, preferably CLIP, SNAP or Halo-Tag.

17. Sensor molecule according to claim 15 or 16, wherein RET is FRET.

18. Sensor molecule according to claim 17, wherein the signal generating moiety comprises a fluorescent acceptor and wherein the synthetic regulatory molecule comprises a fluorescent donor, or vice versa.

19. Sensor molecule according to claim 15 or 16, wherein RET is BRET.

20. Sensor molecule according to claim 19, wherein the signal generating moiety comprises a bioluminescent donor protein, preferably wherein said donor protein has luciferase activity, and wherein the synthetic regulatory molecule comprises a fluorescent acceptor.

21. Sensor molecule according to claim 20, wherein said bioluminescent donor protein is NanoLuc luciferase (NanoLuc).

22. An analytical device comprising a BRET sensor molecule according to any one of claims 19 to 21, wherein the sensor molecule is arranged in such a manner that, when the device is in use for detecting an analyte of interest in a sample, the photons that are emitted from the sensor molecule and that are collected by a detector pass through the sample for a distance shorter than 330 μιη.

23. Device according to claim 22, wherein the sensor molecule is immobilized or absorbed to a solid carrier, preferably a glass or transparent plastic.

24. Device according to claim 23, wherein the sensor molecule is absorbed to a paper carrier or a gel, preferably to chromatography or filter paper.

25. Device according to claim 23, wherein the sensor molecule is comprised in a thin film, or confined in a tube, capillary or (microfluidic) chamber.

26. A method for in vitro detecting an analyte of interest in a sample, comprising the steps of:

(a) contacting the sample with a sensor molecule according to any one of claims 1-21 under conditions allowing for an analyte-induced conformational change within the sensor molecule to occur such that the signal generated by the signal generating moiety is modulated; and

(b) analyzing a change in a signal generated by the signal generating moiety.

27. A method according to claim 26 for in vitro detecting an analyte of interest in a sample using bioluminescence resonance energy transfer (BRET), comprising the steps of: (a) contacting the sample with a BRET sensor molecule according to any one of claims 19 to 21 under conditions allowing for an analyte-induced BRET change to occur; and

(b) analyzing energy resonance transfer.

28. Method according to claim 27, wherein step (b) is performed in solution.

29. Method according to claim 27, comprising the use of an analytical device according to any one of claims 22-25.

30. Method according to any one of claims 27-29, wherein said analyte of interest is a drug, a metabolite, inhibitor, a protein, a biomarker or a nucleic acid molecule.

31. Method according to any one of claims 25-30, wherein the sample is a biological sample or a fraction thereof, preferably a bodily fluid, more preferably selected from the group consisting of blood, serum, saliva, urine, spinal fluid, pus, sweat, tears, breast milk.