WO2005040412A1

WO2005040412A1 - Detection method

Info

Publication number: WO2005040412A1
Application number: PCT/SE2004/001563
Authority: WO
Inventors: Per-Åke Nygren; Olof Nord
Original assignee: Affibody Ab
Priority date: 2003-10-29
Filing date: 2004-10-28
Publication date: 2005-05-06
Also published as: SE0302840D0

Abstract

A method for determining whether , β-lactamase activity is present in a prokaryotic cell is provided, which comprises the steps of: providing a fluorescent substrate for , β-lactamase; providing a prokaryotic cell which comprises a gene encoding a heterologous esterase enzyme, which enzyme has the ability to convert the substrate from a diffusing form to an active form; applying said substrate, in its diffusing form, to said prokaryotic cell, said application enabling said heterologous esterase enzyme to convert the substrate to its active form; culturing said cell under conditions such that any β-lactamase activity present in the cell cleaves said substrate, yielding a fluorescent signal; and detecting fluorescence from said cell.

Description

DETECTION METHOD

Field of the invention The present invention relates to a method for determining whether /3-lactamase activity is present in a prokaryotic cell. In particular, it makes the use of fluorescent substrates in such a method possible, through the provision, in the prokaryotic cell, of a ^'heterologous esterase enzyme.

Background The present invention may be put to use for example in the field of protein-protein interaction studies, including studies of the inhibition of protein-protein interactions. Depending on the application and the context in which the interacting reagents are available, such studies can be performed by numerous techniques, including biosensor-based techniques, enzyme-linked immunosorbent assays (ELISA) , flow cytometry, western blotting, atomic force microscopy, immunoprecipitation (including pull-out experiments) , nuclear magnetic resonance and chromatographic or centrifugal techniques. A particular field of protein-protein interaction studies is related to the work with large collections (libraries) of variants of oligopeptides or polypeptides, such as antibody fragments, proteins (including enzymes) or protein domains. In the construction of such polypeptide libraries, different sources can be recruited to provide genetic material encoding the collection of polypeptide sequences. For example, to construct antibody fragment libraries, genes encoding variable domains of immunoglobulins can be obtained from human or animal donors, or by construction of immunoglobulin encoding gene pools using in vi tro synthesized oligonucleotides, or by combined strategies . Typically, polypeptide libraries are screened for their content of particular polypeptide species capable of interacting (binding or capable of enzymatic conversion) with a desired molecule used as target (or bait) during screening or selection procedures that allow the majority or all variants to be present simultaneously. Using various methods, such isolated polypeptide species can be identified via their corresponding encoding gene fragments which have remained physically linked to their respective gene product polypeptides during the procedure. Several systems allowing such genotype-phenotype linkage have been described, including phage display technology, mRNA display, cell display, viral display, yeast two-hybrid system, ribosomal display, STABLE, plasmid display, in vitro compartmentalisation (IVC) , covalent display and protein fragment complementation assay (PCA) . These systems differ in their implementation in several aspects, including (i) the mode of physical linkage between the encoding nucleic acid (DNA or RNA) and the translated polypeptide, (ii) a dependence on cells or cell-free extracts for the transcription and translation of the encoded polypeptide, (iii) mode of presentation of the polypeptide library members to the surroundings, (iv) requirement of the characteristics of the polypeptide library members for functional expression, (v) mode of mechanism for selection or screening including competitive preparative enrichment by interaction or interaction-dependent host cell survival or reporting, (vi) requirement for target molecule status including a sample containing the target substance or having access to a gene fragment encoding a polypeptide target . As mentioned above, one type of selection system for protein-protein interactions is the protein fragment complementation assay (PCA) , which relies on an interaction-dependent reconstitution of the activity of a reporter enzyme. This enzyme is expressed in a cell divided into two separate fragments, encoded by two different expression cassettes. Production of the enzyme fragments per se in the same cell does not generally lead to re-association of the enzyme fragments. However, through genetic fusion of one of two interacting polypeptides to each of the enzyme fragments, the association of the two polypeptide-enzyme fragment fusion proteins is promoted. This, ideally, leads to the formation of non-covalently reconstituted enzyme molecules with an enzymatic activity that can be monitored. The system was originally described using a mouse dihydrofolate reductase (mDHFR) enzyme, which was genetically split in two fragments that could be reassociated into a functional enzyme via a protein- protein interaction exerted by genetically fused polypeptide moieties (see e g WO98/34120; Pelletier et al (1999), Nat Biotechnol 17:683-690). PCA-based systems also allow for the screening of substances capable of inhibition such protein-protein interaction, in which case a decreased enzymatic activity is monitored. The mDHFR system has been shown to be applicable in both prokaryotic and mammalian cells. A mDHFR activity can be monitored through a different mechanism, including the binding of a fluorescently labeled methotrexate molecule or an ability to grow under selective conditions. Typically, when mDHFR-based PCA technology is used in bacteria, reconstituted mDHFR activity can be monitored through the capability of host cells to grow in the presence of trimethoprim (see WO00/07038) . This substance effectively inhibits the bacterial DHFR activity, rendering the cells dependent on the reconstituted mouse homologue for survival. Other examples of the application of mDHFR PCA are presented e g in WOOl/88168 and O01/00866. The mDHFR PCA system has been followed by other examples of enzymes used in PCA. It has been shown that the bacterial enzyme 3-lactamase is possible to genetically split into fragments that, after genetic fusion to interacting protein pairs, are reconstituted into an active enzyme (see O01/94617; Galarneau et al (2002), Nat Biotechnol 20:619-622; ehrman et al (2002), Proc Natl Acad Sci USA 99:3469-3474) . This system has been described as being associated with a significantly lower background (lower number of false positives) than the mDHFR system. For this system, alternative methods exist for the monitoring of a reconstituted /3-lactamase activity. When implemented in prokaryotic host cells, the addition of /3-lactam-based antibiotics to the growth medium has been described, resulting in a growth advantage for cells that have a reconstituted /3-lactamase activity. Thus, it has been shown that the activity of this enzyme can be monitored, in a prokaryotic host cell, by a capability of the cell to grow under conditions involving exposure to certain /3-lactam antibiotics such as ampicillin and carbenicillin. A fluorogenic substrate denoted CCF2/AM has been successfully used to monitor /3-lactamase activity in mammalian cells (Zlokarnik et al (1998), Science 279:84- 88) . However, this substrate must be enzymatically modified by the host cell for activation and in order to be kept inside the cell after cell membrane translocation. This is a prerequisite to allow discrimination of 3-lactamase positive cells by optical means, including for example microscopy and flow cytometry. In the mammalian cells tested, the necessary enzymatic modification is carried out by naturally present esterase enzymes, and results in the removal, by hydrolysis, of substrate end groups resulting in a negatively charged and activated substrate. Importantly, the host cell esterase-mediated substrate activation as seen in mammalian cells has hitherto not been described for prokaryotic cells. This is despite the obvious advantages of such cells over eukaryotic cells in terms of economy, time, ease of handling and others known to the person of skill in the field. Known assays for the determination of /3-lactamase activity in prokaryotic cells have been based on survival of cells in media containing penicillin-derivatives, for example ampicillin or carbenicillin, where a quantitive measure of /3-lactamase activity levels can be obtained from use of increasing concentrations of these compounds. In addition, spectroscopic or pH-stat-based assays have been described using chromogenic substrate derivatives (e g Nitrocefin) or through monitoring of substrate conversion-specific pH effects, respectively. No methods employing fluorescent substrates for /3-lactamase in prokaryotic cells have been described.

Disclosure of the invention Thus, it is an object of the present invention to enable the use, in assays in prokaryotic cells, of β- lactamase with fluorescent substrates thereof. Another object of the present invention is to make possible the rapid classification, through e g fluorescence activated cell sorting, of positive/negative prokaryotic cells in a /3-lactamase PCA or other assay depending on presence or absence of /3-lactamase activity. Yet an object of the present invention is to provide an alternative to fluorescent, /3-lactamase-based assays in eukaryotes, in which the convenience inherent in working with prokaryotic cells is utilized. These objects, and others that will be evident to the skilled person from the description herein, are attained by the invention as claimed. Thus, the invention provides a method for determining whether /3-lactamase activity is present in a prokaryotic cell, which method comprises the steps of: - providing a substrate for /3-lactamase, which substrate i) is capable of diffusion through cellular membranes in a diffusing form thereof, ii) is incapable of diffusion through cellular membranes in an active form thereof, iii) comprises at least one first and one second photo-active group, chosen and arranged such that fluorescence resonance energy transfer occurs between them, yielding a fluorescent signal at a first wavelength, and iv) comprises a /3-lactam ring positioned between said photo-active groups; - providing a prokaryotic cell which comprises a gene encoding a heterologous esterase enzyme, which enzyme has the ability to convert the substrate from its diffusing form to its active form; - applying said substrate, in its diffusing form, to said prokaryotic cell, said application enabling said heterologous esterase enzyme to convert the substrate to its active form; - culturing said cell under conditions such that any /3-lactamase activity present in the cell cleaves said substrate, thus separating the photo-active groups and disrupting fluorescence resonance energy transfer, yielding a fluorescent signal at a second wavelength; and - detecting fluorescence from said cell, the first wavelength indicating absence and the second wavelength indicating presence of /3-lactamase activity in the cell. Thus, the invention is largely based on the surprising discovery by the present inventors that the engineering of prokaryotic cells to express a heterologous esterase enzyme enables the trapping in the cells of a fluorogenic /3-lactamase substrate. Substrates of this kind may now, for the first time, be used in prokaryotic cells, provided that these cells are engineered according to the present invention to express the proper heterologous enzyme. This opens the possibility for a wide range of applications in which the detection of /3-lactamase activity in prokaryotes is of interest, since such detection may now be performed quickly and conveniently using known principles for fluorescence detection and fluorescence activated cell sorting (FACS) . Among such applications, particular mention may be made of protein fragment complementation assays (PCA) and yeast two-hybrid assays, in which interactions between polypeptides may be studied. Coupling fluorescent detection with the use of prokaryotes will greatly increase the capacity and usefulness of these methods. A step in the method according to the invention is the provision of a substrate. As substrate in the present invention, any molecule may be used which fulfils the criteria listed. Thus, the substrate is able to adopt at least two different forms. In one form, denoted the diffusing form, it is capable of diffusing across cellular membranes. In the method of the invention, the substrate is applied to a prokaryotic cell suspected of harboring /3-lactamase activity in this diffusing form, which allows the substrate to enter the cell. In another form, denoted the active form, the substrate is no longer capable of passing cellular membranes. Use of such a substrate in a cell is dependent on the conversion from the diffusing form to the active form, since the substrate, were it present in its diffusing form, would leave the cell before having performed its intended function of serving as a /3-lactamase substrate. One type of substrate contemplated for use in the method according to the invention comprises at least one ester group in its diffusing form. The non-charged nature of the ester group makes it possible for the substrate to pass through cellular membranes. Conversion of the substrate to the active form will then involve cleavage of the at least one ester group to reveal a negatively charged group, whose presence will prohibit passage through cellular membranes . Furthermore, the substrate comprises at least one first and one second photo-active group, which may be two identical or different fluorescent groups, or one fluorescent group and one non-fluorescent quenching group. The first and second photo-active group are chosen such, and arranged in the substrate molecule in such a way, that fluorescence resonance energy transfer (FRET) occurs between them. As is previously known in the field of fluorescent molecules, this phenomenon may occur between several known pairs of photo-active groups. Thus, examples of pairs of photo-active groups for use in the substrate in the method of the invention include, but are not limited to, fluorescein - tetramethylrhodamine; IAEDANS - fluorescein; EDANS - dabcyl ; fluorescein - fluorescein; BODIPY FL - BODIPY FL; fluorescein - QSY 7; fluorescein - QSY 9; Cy3 - Cy5 and fluorescein - coumarin. The photo-active groups in the substrate are preferably chosen such that FRET only occurs in the active form of the substrate, and not in the diffusing form thereof. In the case outlined above of a substrate comprising at least one ester group, the photo-active groups may also be esterified in the diffusing form of the substrate. Conversion into the active form through cleavage of ester groups will then serve the dual purpose of i) preventing passage through the cellular membrane, and ii) FRET activation. It is not excluded, however, that FRET occurs in both the diffusing and the active form. The photo-active groups, when present in an active substrate, will thus emit a fluorescent signal at a certain wavelength λi, which is dependent on the FRET occurring between the at least two groups. Being a substrate for /3-lactamase, the substrate also comprises a /3-lactam ring. This structure constitutes the cleavage site for the /3-lactamase enzyme. Thus, any /3-lactamase activity present in a prokaryotic cell, which has been "loaded" with a substrate in accordance with the present invention, will act on this structure and cleave the substrate. The /3-lactam ring in the substrate is positioned between the photo-active groups, so that cleavage of the substrate by /3-lactamase activity will separate them. As substrates of interest in the method according to the invention, mention is made of substrates that, in their diffusing form, belong to the group consisting of acetate esters and acetoxymethyl esters. In particular, the substrate chosen may, in its diffusing form, be an acetoxymethyl ester. Substrates chosen from the group consisting of CCF2/AM and its derivatives and analogues are considered as particularly useful . The particular properties of the substrate for use in the method of the invention will now be explained using the above-mentioned substrate CCF2/AM as a non- limiting example. Reference is made to Figures 1 and 2. As illustrated in Figure 1, the non-fluorescent and non-polar CCF2/AM substrate (diffusing form) comprises two photo-active groups, coumarin (donor) and fluorescein (acceptor) , connected via a cephalosporin /3-lactam ring (indicated) . Four ester groups (Ac, acetyl ; Bt, butyryl ; AM, acetoxymethyl) prevent the excitation of the coumarin donor in the CCF2/AM molecule when the substrate is excited at a wavelength of 409 nm. The proper enzyme has the ability to catalyze the hydrolysis of the four ester groups, which activates the CCF2/AM substrate to CCF2. When exciting the substrate at 409 nm, fluorescence energy transfer (FRET) will occur from the coumarin donor to the fluorescein acceptor, yielding green fluorescein fluorescence at 520 nm. If /3-lactamase is present, it will catalyze the cleavage of the /3-lactam ring, interrupting the FRET between the coumarin donor and the fluorescein acceptor. Thus, when both the hydrolyzing enzyme and /3-lactamase are present, the substrate will emit blue light when excited at 409 nm, due to blue fluorescence at 447 nm from the coumarin photo-active group attached to the cephalosporin, whereas the released mercaptofluorescein hardly absorbs the 409 nm light and is further quenched by its free thiol . In Figure 2, the CCF2/AM substrate is schematically depicted as Y-shaped. Its two photo-active groups, the coumarin donor and the fluorescein acceptor, have been denoted FI and F2, respectively. As described above, the four protecting ester groups makes the substrate non- fluorescent and non-polar. Due to these non-polar properties, the CCF2/AM substrate can diffuse across a cell membrane (left panel of Figure 2) . If the host cell possesses the necessary enzyme, the substrate is activated within the host cell through conversion by ester hydrolysis of CCF2/AM (diffusing form) to CCF2 (active form) . Once activated in the cell, the CCF2 substrate is intracellularly trapped due to its polyanionic charge and cannot cross the cell membrane (right panel of Figure 2) . In this form, FRET occurs between FI and F2 , and the substrate is susceptible to cleavage by /3-lactamase . Another step in the method of the invention is the provision of a prokaryotic cell, which comprises a gene encoding a heterologous esterase enzyme, which enzyme has the ability to convert the substrate from its diffusing form to its active form. As used herein, an "heterologous" enzyme is intended to mean an enzyme that is not present in the native prokaryotic cell, but whose expression has been intentionally provided for through the use of genetic engineering or other techniques. The cell may have been provided with a plasmid carrying a gene encoding the heterologous enzyme. Alternatively, the gene may have been inserted in the genome of the prokaryotic cell. Methods and means for performing such a manipulation of prokaryotic cells to express heterologous enzymes are well known in the art. For reasons of availability, economics and ease of handling, a preferred prokaryotic cell for use in the present invention is an Escherichia coli cell. As the heterologous esterase enzyme, any enzyme capable of converting the substrate from its diffusing form to its active form may be employed. In the case of a substrate comprising ester group (s) in its diffusing form, a heterologous esterase enzyme which has the ability to cleave the esterase grou (s) of the substrate may be used. As is known by those of skill in the art, esterase enzymes may be classified in those that require surfacial activation and those that do not. Esterases with a "lid region" that covers the active site need surfacial activation, e g through presentation at their surface of a hydrophobic micelle or a lipid drop (an example of such an esterase is the Thermomyces languinosa esterase (Cajal et al (2000), Biochemistry 39(2) :413- 423) ) . Without wishing to be bound by any particular theory, it is contemplated by the present inventors that esterases with a lid region which requires surfacial activation are unsuitable for use as the heterologous esterase enzyme in the present invention. Rather, the heterologous esterase enzyme expressed by the prokaryotic cell should be a "lid-less" esterase. This characteristic is regarded as beneficial for activity towards substrates present as monomers in a bacterial periplasm. As such "lid-less" esterases, mention is made of enzymes that belong to the cutinase family of lipase/esterase enzymes. Preferred are fungal esterases, in particular cutinases originating from genus Fusarium. Particular mention is made of the lipase cutinase from the species Fusarium solani pisi (Soliday et al (1989), J Bacteriol 171:1942- 1951) . Other fungal esterases that may be used in the present invention are lipases from genus Candida . In this case, particular mention is made of the lipase B from Candida antarctica (see eg Uppenberg et al (1994) , Structure 2:293-308; Ottosson et al (2002), Biochim Biophys Acta 1594:325-334). The genes for "lid-less" esterases, such as the cutinases and lipases mentioned above, may readily be obtained from various organisms and cloned into expression systems within the prokaryotic host cell using known techniques. It is to be noted that the heterologous esterase enzyme need not be expressed in the prokaryotic cell in the exact same form as it is present in its native organism; conventional modifications such as fusions with other polypeptide domains, additional amino acids for detection or purification, point mutations or other minor changes, provision of tags detectable by antibodies or radioactivity measurement and others may be made by the skilled person without departing from the spirit and scope of the invention, as long as the heterologous esterase enzyme retains the ability to convert the substrate from its diffusing form to its active form. The method according to the invention further comprises applying the substrate, in its diffusing form, to the prokaryotic cell. The result of such application, provided that the substrate is applied in a proper concentration readily determined by the skilled person, is that the substrate will enter the cell via diffusion through the membrane. Once present inside the cell, the action of the heterologous esterase enzyme converts the substrate from its diffusing form to its active form, preferably through cleavage of ester groups to yield charged groups that preclude membrane passage. As a result, the active substrate, being incapable of passing the membrane, will be present inside the cell, ready to act as a substrate for any 3-lactamase activity in the cell . An example of performing this step of the method of the invention is illustrated in Figure 3. Again, the substrate CCF2/AM is used as a non-limiting example. The figure illustrates application of this substrate to two different prokaryotic cells, exemplified by E. coli cells. In the comparative panel on the left, the substrate is applied to a cell that has not been provided with a heterologous esterase enzyme. This cell thus lacks the necessary means for converting CCF2/AM to CCF2 , and the substrate will pass freely in and out through the cellular membrane. In the panel on the right, illustrating the invention, the cell has been provided with the proper heterologous esterase enzyme. This enzyme catalyses the formation of CCF2 from CCF2/AM. CCF2 will then remain in the cell, susceptible to cleavage by any /3-lactamase present. The method further comprises culturing of the prokaryotic cell (containing activated substrate) under conditions that are such that any /3-lactamase activity in the cell cleaves the substrate. Cleavage of the substrate occurs at the /3-lactam site between the photo-active groups, with the result that the photo-active groups are separated. Thus, since they are no longer kept in each other's vicinity, the FRET between the photo-active groups is disrupted. One of the photo-active groups then yields a fluorescent signal with a second wavelength λ₂, which is different from the "FRET wavelength" λi. The final step of the method according to the invention comprises the detection of fluorescence from the prokaryotic cell. In this regard, detection of the first wavelength λi will indicate the absence of β- lactamase activity. Radiation of this wavelength implies intact FRET between the photo-active groups, which in turn is a sign that the substrate is still present in an uncleaved form in the cell. However, if the second wavelength λ₂ is detected, this is a sign that FRET has been lost and the photo-active groups separated, i e that the substrate has been cleaved by /3-lactamase activity. The method according to the present invention may be useful in any situation in which assaying for /3-lactamase activity in the prokaryotic cell is of interest. Thus, it may find application in a situation where a prokaryotic cell expresses a /3-lactamase enzyme as part of the normal expression pattern, i e comprises a /3-lactamase gene in its genome or on a natural or engineered plasmid. In this case, it may be of interest to assay for deletional mutations or other mutations that have disrupted the function of the original _/β-lactamase gene, to assay for loss of a plasmid containing a /3-lactamase gene, to analyze the effect of RNAi or antisense RNA towards the mRNA produced by the /3-lactamase gene etc. The method may also find application in prokaryotic cells normally lacking /3-lactamase activity. For example, such a cell may be one that is used for a protein fragment complementation assay (PCA) , such as that described for mammalian cells in WO01/94617 and Galarneau et al (2002), Nat Biotech 20:619-622. In analogy to the situation in mammalian cells described in these documents and in the Background section, the prokaryotic cell normally lacking /3-lactamase is, in such an assay, provided with two gene fragments that encode separate and complementary fragments of a functional /3-lactamase enzyme. In a positive PCA, these /3-lactamase fragments are each fused to genes encoding different fusion moieties, the interaction of which is to be analyzed. In a preferred situation, the fusion moiety fused to one of the /3-lactamase fragments is a polypeptide target, for which suitable interacting partners is desired. The fusion moiety fused to the other /3-lactamase fragment may then be designed as a large set, or library, of variants or mutants of a common "scaffold" polypeptide, in which library one wishes to find a suitable binding partner to the target . When interaction occurs between these fusion moieties, the /3-lactamase fragments are brought into contact with each other, and associate to form a reconstituted, active enzyme. Thus, the /3-lactamase activity detected by the method according to the invention may appear as a result of reconstitution of active /3-lactamase enzyme through association of fragments thereof in a positive protein fragment complementation assay. The interaction of the fusion moieties with each other may thus be detected as a switch in fluorescence from the wavelength λi (indicating FRET, i e intact substrate, i e no /3-lactamase activity) to the wavelength λ₂ (indicating no FRET, i e cleaved substrate, i e /3-lactamase activity) . In an alternative PCA setup, termed negative PCA, the study instead focuses on the disruption of interaction between the fusion moities. As before, β- lactamase activity is reconstituted through association of enzyme fragments with the aid of fusion partners that interact with each other. However, in this case, the fusion partners are chosen among those having a known affinity for each other, so that the association of functional /3-lactamase is expected for all cells provided with the different fusions of /3-lactamase fragments and interacting moities. The expected fluorescent signal from these cells is then the wavelength λ₂. This situation may subsequently be exploited to study the ability of a compound or set of compounds, e g each of a library of small molecule compounds, to disrupt the known interaction between the fusion moieties. When a compound having this disruptive ability is applied to the cell, this will lead to the dissociation of functional β- lactamase, as a consequence of the disruption of the interaction between the fusion moities. As a result, the discovery of a compound that disrupts the interaction between the fusion moieties may be detected as a switch in fluorescence wavelength from λ₂ back to λi. In another setup of a negative PCA assay, the small molecule compound or set of compounds mentioned above may be replaced by a gene expressing a protein, whose effect on the known interaction pair is to be studied. If the protein encoded by the added gene disrupts the interaction between the fusion moieties, the same switch in fluorescence wavelength as described for a "disruptive" small molecule compound will be observed. The opposite effect of small molecules on interactions between the fusion moieties may also be studied, in another form of a positive PCA set-up. Here, the fusion partners are chosen among those that do not interact with each other in their initial states. Thus, no functional /3-lactamase is expected in any cells provided with the different fusions between /3-lactamase fragments and non-interacting moities. Application of small-molecule compounds to cells expressing such constructs enables studying whether a certain, or a group of, small-molecule compound (s) has the ability to promote the interaction of the initially non-interacting fusion moieties. The presence of a small-molecule compound with the dual ability to bind to one of the fusion moieties and to influence that fusion moiety (e g through promotion of a conformational change) so that it will interact with the other fusion moiety may then be detected as a signal, since such a presence will lead to re-association of functioning /3-lactamase. In another approach to studying interactions between macromolecules such as polypeptides, the potential β- lactamase activity is instead provided as a reporter gene, which is switched on or off depending on the existence in the cell of a functional transcription factor. As in the case of PCA, this type of analysis is preferably performed in a prokaryotic cell that is otherwise free from endogenous /3-lactamase activity. The same reasoning as in the case of PCA applies, with the exception that the fusion moieties are each coupled to a fragment of said transcription factor, so that the interaction of fusion moieties with each other results in a functional transcription factor, whereas disruption of this interaction leads to dissociation of the transcription factor. As is known by those of skill in the analogous field of yeast two-hybrid systems, the transcription factor may either be a repressor or an enhancer of reporter gene expression. Depending on the properties of the transcription factor, the assay may be designed so that successful interaction yields a signal from λi to λ₂ or vice versa, whereas the disruption of interaction yields the opposite switch in wavelength, as discussed above for PCA. Regardless of the underlying purpose of detection of /3-lactamase activity, the switch between fluorescence wavelengths λi and λ₂ may readily be exploited through the use of a fluorescence activated cell sorter (FACS) . In the case of interaction studies where one of the fusion moieties is a target and the other a number of variants of a scaffold, it is particularly important to be able to easily isolate those cells that express the correct interacting pair. The use of FACS for this purpose offers advantages over the heretofore known methods, in that no cell-growth assays need be performed to assay for β- lactamase activity. Furthermore, the invention makes it possible to use flow cytometry to analyze and isolate cells with spectral properties that are characteristic of a /3-lactamase dependent conversion of the substrate. This also provides means for a ranking of cells that show different fluorescence intensities, by ratiometric fluorescence detection. The different intensities in turn reflect the amount of /3-lactamase activity in the cells. This activity can be related to the number of active β- lactamase enzymes. In an investigation of protein-protein interactions with association of /3-lactamase fragments, the number of reconstituted /3-lactamase enzymes in turn reflects the efficiency in binding between a particular polypeptide library member and a target protein. The invention will be described in more detail through the following recital of experiments performed in accordance therewith. This description should not be construed as limiting.

Brief description of the drawings Figure 1 shows the properties of CCF2/AM in its diffusing form, in its active form and following cleavage with /3-lactamase. Figure 2 is a schematic representation of how the membrane-permeable fluorogenic CCF2/AM substrate is taken up by a cell and subsequently activated by intracellular esterases present in mammalian cells. Figure 3 is a schematic description of the invention using the lipase (esterase) cutinase to activate the fluorogenic CCF2/AM substrate to CCF2 in Escherichia coli cells . Figure 4 is a schematic representation of the plasmids (A) pE318, (B) pE318-cutinase, (C) pE318- cutinase-CAT, and (D) pEZZ-cutinase . These plasmids were used for in vivo and in vi tro protein production. Figure 5 shows the results from fluorometric analyses of (A) the CCF2/AM substrate alone, (B) cutinase alone and (C) cutinase incubated with CCF2/AM. Emission spectra (435-550 nm range) were monitored for all samples after excitation at 409 nm using a fluorometer. Figure 6 shows the results from fluorometric analyses of E. coli periplasmic fractions incubated as indicated with the fluorogenic CCF2/AM substrate.

Emission spectra (435-550 nm range) were monitored for all samples after excitation at 409 nm using a fluorometer. Figure 7 shows the results from fluorescence microscopy analyses (excitation at 405 ± 10 nm, emission detected using a 435 nm long pass filter) of four different E. coli cell populations incubated with the fluorogenic CCF2/AM substrate. (A) A non-plasmid bearing E. coli RR1ΔM15 cell incubated with CCF2/AM. (B) An E. coli RR1ΔM15 cell transformed with the plasmid pE318 and incubated with CCF2/AM. (C) An E. coli RR1ΔM15 cell transformed with the plasmid pE318-cutinase-CAT and incubated with CCF2/AM. (D) An E. coli RR1ΔM15 cell transformed with the plasmid pE318-cutinase and incubated with CCF2/AM. Figure 8 is a schematic representation of the use of the invention for selection of protein ligands from a polypeptide library using the fluorogenic CCF2/AM substrate for detection of interacting pairs of library members (LM) and target protein (T) .

Examples

Example 1. In vi tro analysis of CCF2/AM substrate activation by cutinase lipase The CCF2/AM fluorogenic substrate was analyzed for its spectral characteristics under different conditions, to investigate if a recombinant esterase derived from Fusarium solani pisi , denoted cutinase (EC 3.1.1.74), could be used to remove end groups linked to the substrate by ester linkages and thus activate the fluorescent properties of the substrate. The cutinase esterase was produced in E. coli from the plasmid pEZZ- cutinase (Figure 4D) as fused to a divalent IgG-binding affinity handle ZZ and purified by IgG-affinity chromatography as described by Bandmann N and coworkers (Bandmann N et al (2000), J Biotechnol 79:161-172).

Emission spectra (435-550 nm range) were monitored for samples in HEPES (Sigma, Saint Louis, MO, USA) buffered saline (pH 7.3), containing the CCF2/AM substrate (Aurora Biosciences, San Diego, CA, USA; 0.6 μM final concentration prepared according to manufacturer's recommendations) , after excitation at 409 nm using a Perkin-Elmer LS 50B fluorometer (Perkin-Elmer Instruments, CT, USA) . Three different samples were analyzed in this fashion. The samples corresponded to (A) CCF2/AM (0.6 μM) alone, (B) ZZ-cutinase fusion protein alone (40 μM) , and (C) CCF2/AM (0.6 μM) incubated with ZZ-cutinase fusion protein (40 μM) for one hour at room temperature. The results shown in Figure 5 show that a small emission peak at approximately 450 nm can be seen for sample A. This is a wavelength characteristic for emission from coumarin of activated CCF2 substrate. This emission from untreated CCF2/AM substrate could be due to a described chemical instability of the substrate, resulting in spontaneous loss of protective end groups and cleavage of the β- lactam ring (Cavrois et al (2002) , Nat Biotechnol 20:1151-1154) . It is to be noted that no emission at wavelengths corresponding to emission from fluorescein can be seen (520 nm) . This shows that no FRET occurs in the CCF2/AM molecule under these conditions. In the control sample containing only purified ZZ-cutinase fusion protein, no emission signal could be observed, showing that the ZZ-cutinase fusion protein had no intrinsic fluorescence activity under the conditions used (Figure 5B) . In contrast, a distinct emission peak from the fluorescein can be seen in sample C, corresponding to

CCF2/AM substrate incubated with the ZZ-cutinase fusion protein (Figure 5C) . This shows that the added cutinase activity is adequate for activation of the CCF2/AM substrate into its activated form CCF2.

Example 2. Cleavage of cutinase-activated CCF2/AM substrate by β-lactamase The /3-lactamase substrate CCF2/AM was further analyzed for its spectral characteristics under different conditions to investigate if the cutinase described in Example 1 together with TEM-1 /3-lactamase from Escherichia coli (Sutcliffe (1978) , Proc Natl Acad Sci USA 75:3737-3741), would cause activation of the CCF2/AM to its fluorescent form CCF2 and subsequent cleavage of the /3-lactam ring between the coumarin and fluorescein photo-active groups in the CCF2 molecule. Cutinase and β- lactamase were produced in the periplasm of E. coli by transforming the cells with the plasmid pE318-cutinase (Figure 4B) . Two negative controls were included; E. coli wild-type cells without plasmid (RR1ΔM15; Rϋther (1982), Nucleic Acids Res 10:5765-5772) and E. coli cells harboring the pE318 plasmid (Figure 4A) , which encodes the production of /3-lactamase protein. After overnight cultivation of the cells at 37°C in 100 ml tryptic soy broth (TSB; Merck, Darmstadt, Germany) , supplemented with yeast extract (Merck) and ampicillin (100 μg/ml; Sigma, St Louis, MO, USA) , all three samples were pelleted by centrifugation (2600 x g; Sorvall RC26 Plus; Kendro, Asheville, NC, USA) . After removal of TSB media, the pellets were treated with osmotic shock solution (Nossal and Heppel (1966), J Biol Chem 241:3055-3062) to release the periplasmic proteins. The osmotic shock fractions were subsequently incubated with or without CCF2/AM substrate (Aurora Biosciences, San Diego, CA, USA; prepared according to manufacturer's recommendations; final concentration in samples A-E: 0.6 μM; in sample F: 0.4 μM) in HEPES (Sigma, St Louis, MO, USA) buffered saline (pH 7.3), and excited at 409 nm using a Perkin- Elmer LS 50B fluorometer (Perkin-Elmer Instruments, CT, USA) , followed by an emission scan from 435 nm to 550 nm. The results of this fluorometric analysis of the six samples is illustrated in Figure 6, wherein the panels correspond to (A) the periplasmic fraction from non- plasmid bearing E. coli RR1ΔM15 cells, (B) the periplasmic fraction from E. coli RR1ΔM15 cells transformed with plasmid pE318 encoding /3-lactamase, (C) the periplasmic fraction from E. coli RR1ΔM15 cells with plasmid pE318-cutinase encoding β-lactamase and cutinase, (D) the periplasmic fraction from non-plasmid bearing E. coli RR1ΔM15 cells incubated with CCF2/AM, (E) the periplasmic fraction from E. coli RR1ΔM15 cells transformed with plasmid pE318 encoding /3-lactamase incubated with CCF2/AM and (F) the periplasmic fraction from E. coli cells transformed with plasmid pE318- cutinase encoding /3-lactamase and cutinase incubated with CCF2/AM. In Figures 6A-6C, it can be seen that no fluorescence could be detected for the samples incubated without the CCF2/AM substrate. For sample D (Figure 6D) , a minor blue fluorescence peak can be detected at approximately 450 nm and a fluorescein emission shoulder around 520 nm, indicating some activation of CCF2/AM to CCF2. Spontaneous hydrolysis of the /3-lactam ring due to chemical instability (Cavrois et al (2002) , supra) of the CCF2/AM substrate is probably responsible for the emission of blue fluorescence at 450 nm. The results obtained in these experiments indicate that the E. coli RR1ΔM15 cells do not have the necessary esterases in the periplasm for activation of CCF2/AM to CCF2. For sample E (Figure 6E) , the fluorescence around 450 nm is higher compared to that for sample D, indicating that β- lactamase present in the periplasmic space cleaves the β- lactam ring, yielding the blue coumarin fluorescence. Interestingly for cells not expressing cutinase, the coumarin fluorescence is relatively low, indicating that there is very little activation of the CCF2/AM substrate, which correlates to the results obtained for D. In contrast, for sample F, a major peak can be seen for the blue coumarin fluorescence around 450 nm and a tendency of a fluorescein emission shoulder around 520 nm. These results indicate a sufficient production of cutinase in the periplasm of the E. coli cell, mediating activation of CCF2/AM substrate to CCF2. The activated CCF2 molecules are then further cleaved by /3-lactamase, resulting in a very high blue fluorescence peak from the coumarin photo-active group observed in the fluorometric analysis (see Figure 6F) . Noteworthy, the CCF2/AM concentration had to be lowered 1.5-fold to be able to measure the coumarin fluorescence within the fluorometer' s dynamic range, suggesting an effective activation and cleavage of the CCF2/AM substrate by cutinase and β-lactamase, respectively.

Example 3. Fluorescence microscopic analysis of Escherichia coli cells incubated with CCF2/AM This example demonstrates that CCF2/AM can be used in prokaryotic cells to detect /3-lactamase activity in vivo, subsequent to activation of CCF2/AM to CCF2 by cutinase produced in the periplasm. E. coli RR1ΔM15 cells were transformed with the pE318-cutinase plasmid, which encoded the production of both cutinase and β-lactamase. After an overnight cultivation of E. coli cells harboring the pE318-cutinase plasmid at 37 °C in 100 ml tryptic soy broth TSB (Merck, Darmstadt, Germany) supplemented with yeast extract (Merck) and ampicillin (100 μg/ml, Sigma, Saint Louis, MO, USA) , approximately 2 x 10^δ cells were taken out and pelleted by centrifugation (1500 x g; Biofuge Pico, Kendro, Asheville, NC, USA) . The cells were subsequently incubated with CCF2/AM (Aurora Biosciences, San Diego, CA, USA; 1 μM final concentration, prepared according to manufacturer's recommendations) for 1 h in room temperature with continuous agitation, followed by centrifugation at 6000 x g (Biofuge Pico, Kendro, Asheville, NC, USA) . The cells were washed with 100 μl HEPES (Sigma, Saint Louis, MO, USA) buffered saline (pH 7.3) and re-centrifuged at 6000 x g and re-suspended in 100 μl HEPES buffered saline (pH 7.3) at a concentration of 4 x 10⁴ cells/μl. Fluorescence microscopy analysis was performed with an Olympus BX 51 microscope (Olympus, Tokyo, Japan) , equipped with a CCF2 filter set (Chroma Technology, Brattleboro, VT, USA; excitation 405 ± 10 nm, 425 nm dichroic mirror, 435 nm long pass emission) and approximately 1.6 x 10⁵ cells were analyzed. Additionally, to examine if cutinase alone (without the presence of /3-lactamase) , would activate CCF2/AM, yielding green fluorescein fluorescence upon exciting of the cells at 409 nm, the plasmid pE318-cutinase-CAT was constructed by PCR amplification of the chloramphenicol actetyl transferase gene (CAT) from the plasmid pACYC184 (Rose (1988), Nucleic Acids Res 16:355). PCR amplification was performed on a GeneAmp^® PCR System 9700 (PE Biosystems, Foster City, CA, USA) for 30 cycles (95 °C, 30 s; 60 °C, 30 s; 72 °C, 1 min) using the oligonucleotides N00L-31 (5' -GGGGACGTCGTTGGCAGCATCACCCGA- C-3 ' ) and NOOL-32 (5 ' -CCGGGACTCCCCGTCCACGGTCACACTGCTTCCG- 3') as primers (the oligonuleotides were synthesized by MWG-Biotech Ebersberg, Germany) . Five pmol of each primer was added to a PCR mixture containing 5 μl 0.2 mM dNTPs, 5 μl lOx cloned Pfu DNA polymerase reaction buffer (Stratagene, La Jolla, CA, USA), 2.5 U PfuTurbo® DNA polymerase (Stratagene, La Jolla, CA, USA) and water to a final volume of 50 μl . Standard agarose gel electrophoresis analyses of nucleic acids were performed using ethidium bromide for staining. The CAT-fragment was restricted with Aatll and Ahdl and subsequently ligated into Aa il /Ahdl-restricted pE318-cutinase plasmid. Standard cloning work, including standard plasmid preparations, restriction enzyme cleavage, ethanol precipitation of nucleic acids and ligations etc, were performed as described in Sambrook and Russell, Molecular Cloning: a laboratory manual, 3^rd edition, Cold Spring Harbor Laboratory press, Cold Spring Harbor, New York, (2001) and according to suppliers' recommendations. Restriction enzymes were purchased from New England Biolabs (Beverly, MA, USA) and ligation was performed with Rapid DNA Ligation Kit (Roche Diagnostics GmbH, Roche Applied Science, Penzberg, Germany) . In the obtained pE318-cutinase-CAT plasmid (depicted schematically in Figure 4C) , the /3-lactamase gene was replaced with the gene encoding chloramphenicol actetyl transferase (CAT) , resulting in resistance against the antibiotic marker chloramphenicol instead of ampicillin. Thus, E. coli RR1ΔM15 cells transformed with the pE318- cutinase-CAT plasmid produce cutinase in the periplasm, but not /3-lactamase. Two control populations with E. coli RR1ΔM15 cells were also included in the analysis; non- plasmid bearing __.. coli RR1ΔM15 cells incubated with CCF2/AM and E. coli cells transformed with plasmid pE318 (only producing /3-lactamase protein) incubated with CCF2/AM. All three cell populations were grown overnight at 37 °C in 100 ml tryptic soy broth (TSB; Merck, Darmstadt, Germany) supplemented with yeast extract (Merck) and ampicillin/chloramphenicol (100 μg/ml and 20 μg/ml, respectively; Sigma, St Louis, MO, USA) . Approximately 2 x 10⁶ cells were taken out and pelleted by centrifugation (1500 x g; Biofuge Pico, Kendro, Asheville, NC, USA) . The cells were subsequently incubated with CCF2/AM (Aurora Biosciences, San Diego, CA, USA; 1 μM final concentration, prepared according to manufacturer's recommendations) for 1 h in room temperature with continuous agitation, followed by washing with 100 μl HEPES buffered saline (pH 7.3) . The cells were subsequently re-centrifuged at 6000 x g and re-suspended in 100 μl HEPES buffered saline (pH 7.3) at a concentration of 4 x 10⁴ cells/μl. Fluorescence microscopy analysis of approximately 1.6 x 10⁵ cells for each sample was performed as described above . The four different cell populations analyzed with fluorescence microscopy (Figure 7) thus corresponded to (A) E. coli RR1ΔM15 cells incubated with CCF2/AM, (B) E. coli RR1ΔM15 cells transformed with plasmid pE318 encoding /3-lactamase, and incubated with CCF2/AM, (C) E. coli RR1ΔM15 cells transformed with plasmid pE318- cutinase-CAT encoding the production of cutinase and CAT, and incubated with CCF2/AM, and (D) E. coli RR1ΔM15 cells transformed with plasmid pE318-cutinase encoding the production of both /3-lactamase and cutinase, and incubated with CCF2/AM, respectively. Fluorescence microscopy showed that E. coli RR1ΔM15 cells in samples A and B (Figure 7) show very little or no fluorescence at all. In sample C, where cells expressing cutinase were incubated with CCF2/AM, the cell population was highly green fluorescent. This indicates that the CCF2/AM substrate was taken up by the E. coli cell, converted to CCF2 and retained in the cell due the negative charge of CCF2, making it non-permeable with respect to the E. coli cell membrane (cf also Figure 3, right panel) . Also, it correlates with the in vi tro results described in Example 1 (see Figure 5C) , suggesting a sufficient production of cutinase in the periplasm for activation of CCF2/AM to CCF2. For sample D, the E. coli cells were highly blue fluorescent (447 nm) , indicating that CCF2/AM had been activated by cutinase, and thereby retained in the E. coli cell, in turn making it susceptible to /3-lactamase cleavage of the /3-lactam ring in the CCF2 molecule. As earlier observed for the corresponding in vi tro samples, a correlation can be seen between sample D and the results from the use of the periplasmic fraction of E. coli cells transformed with the pE318-cutinase plasmid (see Figure 6F) . In conclusion, the obtained results in these in vivo experiments show that /3-lactamase activity can be efficiently monitored using a substrate with the characteristics of CCF2/AM in prokaryotic host cells that express a heterologous esterase enzyme.

Example 4. Detecting β-lactamase activity in connection with protein fragment complementation assay (PCA) As described previously, prokaryotic cells expressing a heterologous esterase enzyme can be used in conjunction with /3-lactamase PCA for selection and analyses of pairs of interacting proteins, wherein the detection of /3-lactamase activity is performed in accordance with the present invention. Figure 8 schematically illustrates such a PCA method. As an illustrative example, the library members are expressed together with the target protein in the periplasm of an E. coli cell. The library of putative binding partners for the target may for example be constructed as a library of different, mutated DNA sequences that encode variants of a scaffold protein, the mutations being created at predefined positions known to be involved in the interaction with other proteins. Different examples of possible such scaffold proteins have been described, and include, but are not limited to, domains of bacterial receptins, fibronectins, protease inhibitors, retinol binding proteins, bilin binding proteins, amylase inhibitors, CTLA-4, cytochromes and cellulose binding proteins. As examples of preferred scaffolds, domains from bacterial receptins are mentioned. Of such domains, domains derived from the group consisting of staphylococcal protein A, streptococcal protein G and Peptostreptococcus magnus protein L are especially preferred. Thus, in an embodiment, an affinity ligand for use in the method of the present invention is constructed using any immunoglobulin binding domain of staphylococcal protein A as scaffold. Preferably, this domain is the B domain of staphylococcal protein A, or the "protein Z" derived therefrom. In another embodiment, any immunoglobulin binding domain of Peptostreptococcus magnus protein L is used as scaffold. In yet another embodiment, any immunoglobulin binding domain of streptococcal protein G is used as scaffold. In yet another embodiment, an albumin binding domain of streptococcal protein G is used as scaffold. The DNA sequences encoding variants of the scaffold (library members) and the DNA sequence encoding the target protein are each fused to DNA encoding one of the two halves (1, 2) of the _/β-lactamase enzyme, respectively. Each individual cell in the population then expresses two fusion proteins, one being a fusion between one unique library member and a /3-lactamase half, and the other being a fusion between the target protein and the other _/β-lactamase half. The cell population as a whole thus comprises cells expressing a variety of unique library members, but a single cell expresses only one library member. Inventively, cutinase is also expressed in the periplasm of the E. coli cells, for activation of the CCF2/AM substrate to CCF2. Thus, when a certain library member (LM) binds to the target protein (T) , the enzymatic activity of _/β-lactamase is restored. This leads to cleavage of the /3-lactam ring of the activated CCF2 and to disruption of fluorescence resonance energy transfer (FRET) between the two photo-active groups, yielding blue coumarin (FI) fluorescence upon excitation of the cell at 409 nm. In contrast, if there is no interaction between the library member and the target protein, there will be not be any complementation of the _/β-lactamase restoring the enzyme activity. In this case, the CCF2 substrate will remain intact, emitting green fluorescein fluorescence at 520 nm upon excitation at 409 nm, due to the intact FRET between the coumarin and the fluorescein.

Claims

1. Method for determining whether /3-lactamase activity is present in a prokaryotic cell, comprising the steps of: - providing a substrate for _/β-lactamase, which substrate i) is capable of diffusion through cellular membranes in a diffusing form thereof, ii) is incapable of diffusion through cellular membranes in an active form thereof, iii) comprises at least one first and one second photo-active group, chosen and arranged such that fluorescence resonance energy transfer occurs between them, yielding a fluorescent signal at a first wavelength, and iv) comprises a /3-lactam ring positioned between said photo-active groups; - providing a prokaryotic cell which comprises a gene encoding a heterologous esterase enzyme, which enzyme has the ability to convert the substrate from its diffusing form to its active form; - applying said substrate, in its diffusing form, to said prokaryotic cell, said application enabling said heterologous esterase enzyme to convert the substrate to its active form; - culturing said cell under conditions such that any jβ-lactamase activity present in the cell cleaves said substrate, thus separating the photo-active groups and disrupting fluorescence resonance energy transfer, yielding a fluorescent signal at a second wavelength; and - detecting fluorescence from said cell, the first wavelength indicating absence and the second wavelength indicating presence of /3-lactamase activity in the cell.

2. Method according to claim 1, in which said prokaryotic cell is an Escherichia coli cell.

3. Method according to any one of the preceding claims, in which said heterologous esterase enzyme lacks a lid region, thus being independent on surfacial activation for its activity.

4. Method according to any one of the preceding claims, in which said heterologous esterase enzyme is a fungal esterase.

5. Method according to claim 4, in which said fungal esterase is from the genus Fusarium.

6. Method according to claim 5, in which said fungal esterase is the lipase cutinase from Fusarium solani pi si .

7. Method according to claim 4, in which said fungal esterase is from the genus Candida .

8. Method according to claim 7, in which said fungal esterase is the lipase B from Candida antarctica .

9. Method according to any one of the preceding claims, in which the substrate, in its diffusing form, is an acetate ester or acetoxymethyl ester.

10. Method according to claim 9, in which tine substrate, in its diffusing form, is an acetoxymethyl ester.

11. Method according to claim 9 or 10, in which the substrate, in its diffusing form, is selected from the group consisting CCF2/AM and its derivatives and analogues .

12. Method according to claim 11, in which the substrate, in its diffusing form, is CCF2/AM.

13. Method according to any one of the preceding claims, in which appearance of /3-lactamase activity in the cell is a result of reconstitution of active β- lactamase enzyme through association of fragments thereof in a positive protein fragment complementation assay.

14. Method according to any one of claims 1-12, in which loss of _/β-lactamase activity in the cell is a result of disruption of active β-lactamase enzyme through dissociation of fragments thereof in a negative protein fragment complementation assay.

15. Method according to any one of claims 1-12, in which appearance of _/β-lactamase activity in the cell is a result of expression of a /3-lactamase reporter gene.

16. Method according to claim 15, in which said expression is dependent on the formation of a functioning transcription factor which drives _/β-lactamase expression.

17. Method according to any one of claims 1-12, in which loss of _/β-lactamase activity in the cell is a result of interruption of expression of a /3-lactamase reporter gene .

18. Method according to claim 17, in which said interruption is dependent on the disruption of a functioning transcription factor which drives _/β-lactamase expression.