WO2025068399A1

WO2025068399A1 - Novel screening assays for the detection of protein homodimerization

Info

Publication number: WO2025068399A1
Application number: PCT/EP2024/077118
Authority: WO
Inventors: Liza ULČAKAR; Marko NOVINEC; Matej BUTALA
Original assignee: Univerza Ljubljana v Fakulteta za Farmazijo
Current assignee: Univerza Ljubljana v Fakulteta za Farmazijo
Priority date: 2023-09-29
Filing date: 2024-09-26
Publication date: 2025-04-03
Anticipated expiration: 2026-03-29

Abstract

The present invention provides plasmids, recombinant bacteria and uses thereof which enables rapid and simple detection and quantification of homodimerization of soluble proteins. The present invention relates to e.g. a plasmid for the detection of protein homodimerization of the protein in soluble state comprising: a nucleotide sequence encoding the expression of a chimeric protein comprising a protein of interest (POI) fused to a DNA-binding domain (DBD) via a flexible linker, said DBD being from the bacterial transcription factor LexA or ToxR; a nucleotide sequence encoding a reporter protein which is under transcriptional control being activated or repressed by said chimeric protein comprising a protein of interest (POI) fused to a DNA-binding domain (DBD) via a flexible linker once said protein of interest (POI) associates such that said chimeric protein homodimerizes; an origin of replication, and a selection system.

Description

Novel screening assays for the detection of protein homodimerization

Technical field of the invention

The present invention relates to the the field of protein engineering, biotechnology and recombinant gene technology. In particular the present invetion relates to protein homodimerization, screening assays to detect protein homodimerization and its use in medical and pharmaceutical applications.

Background

The already diverse world of protein folds is made even more complex by the ability of some proteins to associate with one another to form supramolecular structures. Oligomers can be categorized into heterooligomers and homooligomers, the latter being more common in nature due to being symmetrical and therefore more energetically favorable ¹. Oligomerization is known to be beneficial because it can improve some existing protein properties such as catalytic efficiency and resistance to denaturation and even lead to previously nonexistent ones, for example, cooperativity ²- ³. In particular, stability and activity are among the critical enzyme properties that need to be improved before use in industrial biotransformations ⁴. Conversely, homooligomerization may be the reason for product contamination in industrial biocatalysis. Most often, enzymes are immobilized on a carrier during industrial processes, which increases their stability towards organic solvents and high temperatures, reduces downstream processing costs, and facilitates biocatalyst recycling and separation from the final product ⁵. If the enzyme is oligomeric, not all subunits are necessarily involved in the immobilization, so product contamination may occur during dissociation. Furthermore, even though oligomerization leads to an increase in stability, it can sometimes come at the expense of a decrease in enzyme activity due to negative cooperativity (e.g., 'half-site' reactivity) ³- ⁶.

The study of homooligomerization of proteins is not only interesting for industrial applications, but also from a medical and pharmaceutical point of view. Although homooligomerization plays a central role in many cellular processes, it can be the main cause of a number of diseases. For example, overexpression and subsequent homodimerization of HER2 receptor tyrosine kinase has been found to play a central role in various cancers, especially breast cancer ⁷ Moreover, in amyloidogenic proteins, dimerization is often considered to be the first step of protein aggregation and is therefore an important aspect of various neurodegenerative diseases ⁸- ⁹. In addition, homodimeric bacterial and viral proteins have been proposed as potential targets for curing infectious diseases ¹⁰ - ¹¹.

For all the aforementioned reasons, homodimerization has become a focal point of many research groups. As it is associated with various diseases, it has become an important target in drug discovery, and due to the needs of industry, the number of attempts to artificially homooligomerize and monomerize enzymes using protein engineering techniques has increased. Consequently, there is a need for the development of appropriate methods to study and detect these interactions. Especially in the case of protein engineering researchers often employ semirational and combinatorial methods that require a screening assay for analysis. Namely, these methods mimic evolution by randomly mutating specific protein surfaces or the entire protein ¹². The resulting libraries of mutant protein variants are then screened for positive results, in this case homodimers or monomers. The most commonly used screening assays for binary protein-protein interactions are genetic technologies, namely protein fragment complementation assays and two-hybrid systems ¹³. These systems are designed to detect interactions directly within cells, which means that the screening process is fast and simple, and the proteins are expressed under physiological conditions. The disadvantage of two-hybrid systems, when it comes to engineering homooligomeric proteins, is that their mechanisms of action rely on two protein libraries - 'bait' and 'prey'. Therefore, the results of two-hybrid screening are both homo- and heterotypic interactions. There are as yet no commercially available screening assays for the detection of homodimerization. Nevertheless, scientists have occasionally developed their own homodimerization assays for their research. Most of these in-house developed assays are based on modular bacterial transcription factors that contain a DNA-binding domain (DBD) and a dimerization domain and are active only in their homodimeric form. The dimerization domain is then replaced by the protein of interest (POI). If the chimeric transcription factor homodimerizes, it binds to DNA and modulates transcription of a reporter gene. For example, scientists have used the transcription factors LexA from E. coli and ToxR from V. cholerae as the basis for their assays. LexA is a cytosolic repressor, whereas ToxR is a transmembrane activator. Both have an N-terminal wHTH motif that binds DNA and a C-terminal domain that allows homodimerization ¹⁴“¹⁶. TOXCAT and GALLEX are assays derived from ToxR and LexA, respectively ¹⁷- ¹⁸. Both are designed to detect homotypic interactions of transmembrane regions, although GALLEX has also been modified to detect heterotypic interactions. LexA-based systems for the detection of homo- and heterodimerization of soluble proteins have also been described ¹⁹. Importantly, none of the abovementioned assays are suitable for screening purposes.

There is a need for flexible and fast assays for the detections and quantification of protein homodimerization, in particular homodimerization of soluble proteins.

Summary of the invention

The present invention is new homodimerization assays constructed on the basis of bacterial transcription factors LexA and ToxR and the plasmids used for such assays. The assays of the present invention are user-friendly as well as very cost- and time-efficient. Furthermore, the assays have a wide range of applications from engineering of oligomeric proteins to drug discovery and basic study of homodimerizing interactions.

In a first aspect the present invention provides a plasmid for the detection of protein homodimerization of the protein in soluble state comprising:

- a nucleotide sequence encoding the expression of a chimeric protein comprising a protein of interest (POI) fused to a DNA-binding domain (DBD) via a flexible linker, said DBD being from the bacterial transcription factor LexA or ToxR.

- a nucleotide sequence encoding a reporter protein which is under transcriptional control being activated or repressed by said chimeric protein comprising a protein of interest (POI) fused to a DNA-binding domain (DBD) via a flexible linker once said protein of interest (POI) associates such that said chimeric protein homodimerizes,

- an origin of replication, and

- a selection system.

In a second aspect the present invention provides a library of plasmids as defined in the first aspect, wherein the plasmids in the library are identical except for the part encoding the POI of the chimeric protein.

In a third aspect the present invention provides a bacterium transformed with the plasmid as defined in the first aspect of the invention or the library of plasmids as defined in the second aspect of the invention. In a fourth aspect the present invention provides the use of the plasmid as defined in the first aspect of the invention for the qualitative or quantitative detection of homodimerization of the protein of interest (POI), said use comprising transforming said plasmid into a bacterium, cultivation of said bacterium and detection or quantification of the increase or decrease of the expression of the reporter gene.

In this work, we describe two new LexA- and three new ToxR-based homodimerization assays. All five systems are adapted for use in E. coli, meaning that protein-protein interactions occur and are studied in vivo, suggesting greater physiological relevance than studying interactions in vitro. In addition, the use of E. coli makes these systems user-friendly and cost- and time-effective. The LexA-based systems allow detection of homodimerization of proteins in the cytosol, whereas the ToxR-based systems allow detection of homodimerization in the periplasm. Due to the oxidative environment and the presence of the (Dsb) system in the periplasm, the latter systems are particularly advantageous when the stability of the POI depends on the formation of disulfide bonds or when the formation of homodimers is expected to occur via the formation of disulfide bonds ²⁰.

The LEXGFP and TOXGFP systems allow detection of homodimerizing interactions by whole-cell fluorescence measurements. The results of this study demonstrate that the fluorescence intensity of both systems can be reliably measured directly in cultures grown in nutrient-rich medium in multiwell plates. This is particularly advantageous for drug discovery, as the systems can be used to screen libraries of potential low molecular weight drugs for destabilization or stabilization of homodimerizing interactions. In this case, the TOXGFP system offers some advantage because the drug under investigation does not need to be transported across the cytoplasmic membrane ²¹.

In addition, TOXGFP and LEXGFP can easily be used for screening directly on agar plates, for example when screening mutant protein libraries. However, an appropriate imaging instrument is required. To make these systems more accessible for screening mutant protein libraries, we have developed new variants of both systems, namely LEXBLUE and TOXBLUE. These variants use p-galactosidase as the reporter, and the results of this study clearly demonstrate that the systems allow blue-white screening on solid agar plates, p-galactosidase activity can also be measured in cultures grown in multiwell plates using the Miller assay; however, the GFP variants of the systems allow for much less labor-intensive and more cost-effective detection of reporter expression. In addition, detection of fluorescence is much more sensitive, which is particularly advantageous for ToxR-based assays where induction of reporter expression is weak. The TOXKAN system was developed to simplify the screening process by requiring only one agar plate to screen the entire library. In addition, TOXKAN can be used for screening in multiwell plates, for example, when screening libraries of potential stabilizers or destabilizers of homodimerization. Furthermore, TOXKAN can be used to discriminate between weak and strong homodimers if an appropriate kanamycin concentration is used.

There are numerous potential applications for the assays described above, ranging from drug discovery and engineering of homooligomeric proteins to studying the evolution of protein homodimerization. While LexA-based assays provide a seemingly binary response to the oligomerization state of proteins, the response of ToxR-based assays, particularly TOXGFP and TOXKAN, depends on the affinity for association of monomers. Therefore, if quantitative analysis of homodimerization is desired, the use of TOXGFP or TOXKAN is more appropriate. On the other hand, LexA-based systems are more robust because the induction of reporter expression is stronger. Therefore, these systems are not interchangeable, and the choice of the most appropriate system should depend on the objective of each individual study.

The present invention is further defined by the following items:

1. Plasmid for the detection of protein homodimerization of the protein in soluble state comprising:

- an origin of replication, and

- a selection system.

2. The plasmid according to item 1, wherein the DBD is from the bacterial transcription factor LexA.

3. The plasmid according to item 1 or 2, wherein the DBD is the LexA mutant LexA408 (SEQ ID NO:1). 4. The plasmid according to any one of items 1-3, wherein said chimeric protein has the structure LexA408 DBD + linker + POL

5. The plasmid according to item 4, wherein the part of the chimeric protein having the structure LexA408 DBD + linker is SEQ ID NO:2.

6. The plasmid according to item 1, wherein the DBD is from the bacterial transcription factor ToxR, such as SEQ ID NO:3.

7. The plasmid according to any one of items 1-4, wherein the expression of the chimeric protein is under control of the native promoter for the DBD.

8. The plasmid according to any one of items 6-7, wherein the expression of the chimeric protein is under control of the ToxR promoter (SEQ ID NO:4).

9. The plasmid according to any one of items 4-6, wherein said chimeric protein has the structure ToxR DBD + ToxR transmembrane region + linker + POL

10. The plasmid according to item 7, wherein the ToxR transmembrane region is SEQ ID NO:5.

11. The plasmid according to item 7, wherein the part of the chimeric protein being ToxR DBD + ToxR transmembrane region + linker is SEQ ID NO:6.

12. The plasmid according to any one of items 1-5, wherein the flexible linker is a G/S rich linker, such as having more than about 50% amino acid residues being G or S, or having more than 75% amino acid residues being G or S.

13. The plasmid according to any one of items 1-6, wherein the flexible linker has a length of about 6 to 25 amino acid residues, of about 6 to 18 amino acid residues or of about 10 to about 15 amino acid residues.

14. The plasmid according to any one of items 1-7, wherein the flexible linker is EASSGGGSGGGSSR (SEQ ID NO: 7) or ASSGGGSGGGSSR (SEQ ID NO:8).

15. The plasmid according to any one of items 1-8, wherein the reporter protein is green fluorescent protein (GFP), p-galactosidase (P-gal) or kanamycin resistance (KanR).

16. The plasmid according to item 15, wherein the reporter protein is green fluorescent protein (GFP).

17. The plasmid according to item 15, wherein the reporter protein is p-galactosidase (P-gal).

18. The plasmid according to item 15, wherein the reporter protein is kanamycin resistance (KanR).

19. The plasmid according to any one of items 1-9, wherein the reporter protein is expressed under control of the cholera toxin promoter (P_rtx) (SEQ ID NO: 9), the A-41G mutant cholera toxin promoter (SEQ ID NQ:10) or the mutant SulA promoter (PsuiA4os) (SEQ ID NO: 11). 20. The plasmid according to any one of items 1-5 and 6-15 which is LEXGFP or LEXBLUE.

21. The plasmid according to any one of items 1-5 and 6-15 which is LEXGFP.

22. The plasmid according to any one of items 1-5 and 6-15 which is LEXBLUE.

23. The plasmid according to any one of items 1 and 6-15, which is TOXGFP, TOXBLUE or TOXKAN.

24. The plasmid according to item 23, which is TOXGFP.

25. The plasmid according to item 23, which is TOXBLUE.

26. The plasmid according to item 23, which is TOXKAN.

27. Library of plasmids as defined in any one of items 1-26, wherein the plasmids in the library are identical except for the part encoding the POI of the chimeric protein.

28. Bacterium transformed with the plasmid as defined in any one of items 1-26 or the library of plasmids as defined in item 27.

29. The bacterium according to item 28, which is Escherichia coli such as E. coli DH5a.

30. Use of the plasmid as defined in any one of items 1-26 for the qualitative or quantitative detection of homodimerization of the protein of interest (POI), said use comprising transforming said plasmid into a bacterium, cultivation of said bacterium and detection or quantification of the increase or decrease of the expression of the reporter gene.

31. Use of the library of plasmids as defined in item 27 for the detection or quantification of homodimerization of a multitude of proteins of interest (POI), wherein said library of plasmids comprises plasmids having the genes encoding the multitude of POI, said use comprising transforming said library of plasmids into bacteria, cultivation of said bacteria, and detection or quantification of the increase or decrease of the expression of the reporter gene.

32. The use according to item 30 or 31, wherein the detection or quantification is done by flow cytometry.

Description of the figures

Figure 1. A) Schematic representation of the LexA-based homodimerization assay. The chimeric transcription factor consists of a mutant LexA DNA-binding domain (grey box) at the N terminus, followed by a flexible linker and the POI (grey circle) at the C terminus. The latter is expressed in the cytosol. Upon homodimerization, the chimera binds to its respective operator region in the mutant sulA promoter (PsuiA4os) and represses transcription of the reporter protein. B) Schematic representation of the ToxR-based homodimerization assay. The chimeric transcription factor consists of an N-terminal ToxR DNA-binding domain (grey box), followed by the ToxR transmembrane region (grey bar), a flexible linker, and the POI (grey circle) at the C terminus. The latter is expressed in the periplasm. Upon dimerization, the chimera binds to the cholera toxin promoter (P_ctx) and initiates transcription of the reporter protein.

Figure 2. Measurement of GFP fluorescence in whole cells transformed with the control constructs of LEXGFP. A) Spectral scan of whole cells washed in PBS. A wavelength of 485 nm was used for excitation and emission was measured between 500 and 600 nm. Upper curve is LEXGFP-ctrl and middle curve is LEXGFP-MBP. B) Endpoint fluorescence measurements of whole cells grown in LB broth. A wavelength of 485 nm was used for excitation and emission was measured at 510 nm. The error bars in the graph represent the standard error of the mean and the asterisks indicate the statistically significant deviation of the mean values compared to the mean fluorescence of the untransformed cells (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Figure 3. Fluorescence imaging of agar plates containing E. coli DH5a transformed with LEXGFP controls. Imaging was performed using Bio-Rad Chemidoc Imaging System and images were processed in Image Lab 6.1. Fluorescence intensity is directly proportional to the colour intensity of the colonies in the negatives.

Figure 4. Quantitative and qualitative characterization of the LEXBLUE system. A) Results of the Miller assay performed on cells transformed with the LEXBLUE controls. The error bars in the graph represent the standard error of the mean and the asterisks indicate the statistically significant deviation of the mean values compared to the mean p-galactosidase activity of the untransformed cells (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). B) E. coli DH5a, transformed with the LEXBLUE control constructs, plated on solid agar plates containing IPTG and X-gal.

Figure 5. Measurement of GFP fluorescence in whole cells transformed with the control constructs of TOXGFP. A) Spectral scan of whole cells washed in PBS. A wavelength of 485 nm was used for excitation and emission was measured between 500 and 600 nm. The curves from upper to lower curve are TOXGFP-LexAC, TOXGFP-LexAC G124D, TOXGFP-AncHb, TOXGFP-MBP, untransformed DH5a and TOXGFP-ctrl, respectively. B) Endpoint fluorescence measurements of whole cells grown in LB broth. A wavelength of 485 nm was used for excitation and emission was measured at 510 nm. The error bars in the graph represent the standard error of the mean and the asterisks indicate the statistically significant deviation of the mean values compared to the mean fluorescence of the untransformed cells (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Figure 6. Fluorescence imaging of agar plates containing E. coli DH5a transformed with TOXGFP controls. Imaging was performed using Bio-Rad Chemidoc Imaging System and images were processed in Image Lab 6.1. Fluorescence intensity is directly proportional to the colour intensity of the colonies in the negatives.

Figure 7. Quantitative and qualitative characterization of the TOXBLUE system. A) Results of the Miller assay performed on cells transformed with the TOXBLUE controls. The error bars in the graph represent the standard error of the mean and the asterisks indicate the statistically significant deviation of the mean values compared to the mean p-galactosidase activity of the untransformed cells (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). B) E. coli DH5a, transformed with the TOXBLUE control constructs, plated on solid agar plates containing IPTG and X-gal.

Figure 8. Characterization of TOXKAN. A) Growth curves of E. coli DH5a transformed with TOXKAN controls in the presence of increasing kanamycin concentration. The error bars in the graphs represent the standard error of the mean. B) Growth of E. coli DH5a transformed with TOXKAN controls on LB agar plates containing 100 pg/mL kanamycin.

Figure 9. Maps of plasmids used in the LexA-based assays (A - LEXGFP, B - LEXBLUE). The chimeric transcription factor in composed of the mutant LexA DBD, followed by a flexible linker, and the POL Reporter - eGFP in the case of LEXGFP and LacZa in the case of LEXBLUE - is under transcriptional control of P_SUIA408. Both plasmids also contain the resistence to ampicillin and the origin of replication. Restriction sites Xbal and BamHI were chosen for cloning of cDNA for the POI into the LEXGFP and LEXBLUE plasmids. Both plasmid maps were prepared using Snapgene.

Figure 10. Maps of plasmids used in the ToxR-based assays (A -TOXGFP, B -TOXBLUE, C - TOXKAN). The chimeric transcription factor in composed of the ToxR DBD, followed by the ToxR transmembrane region, a flexible linker, and the POL Reporter - sfGFP in the case of TOXGFP, LacZa in the case of TOXBLUE and KanR in the case of TOXKAN - is under transcriptional control of P_ctx- All plasmids also contain the resistence to ampicillin and the origin of replication. Restriction sites Xbal and BamHI were chosen for cloning of cDNA for the POI into the plasmids.

Figure 11. Endpoint fluorescence measurements during culturing of E. coli DH5a transformed with LEXGFP controls. The ratios between fluorescence intensities of different controls stabilize in the stationary phase. The error bars in the graphs represent the standard error of the mean and the asterisks indicate the statistically significant deviation of the mean values compared to the mean fluorescence of the untransformed cells (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). B) Growth curve of E. coli DH5a transformed with LEXGFP controls. The error bars in the graph represent the standard error of the mean.

Figure 12. Endpoint fluorescence measurements during culturing of E. coli DH5a transformed with TOXGFP controls. The ratios between fluorescence intensities of different controls stabilize in the stationary phase. The error bars in the graphs represent the standard error of the mean and the asterisks indicate the statistically significant deviation of the mean values compared to the mean fluorescence of the untransformed cells (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). B) Growth curve of E. coli DH5a transformed with TOXGFP controls. The error bars in the graph represent the standard error of the mean.

Description

Unless specifically defined herein, all technical and scientific terms used have the same meaning as commonly understood by a skilled artisan in the fields of biochemistry, genetics, and molecular biology.

All methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, with suitable methods and materials being described herein. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will prevail. Further, the materials, methods, and examples are illustrative only and are not intended to be limiting, unless otherwise specified.

The practice of the present invention will employ, unless otherwise indicated, conventional techniques of cell biology, cell culture, molecular biology, transgenic biology, microbiology, fermentation technology, recombinant DNA, and assays, which are within the skill of the art. Such techniques are explained fully in the literature.

In the broadest aspect the present invention provides a plasmid for the detection of protein homodimerization of the protein in soluble state comprising:

- a nucleotide sequence encoding the expression of a chimeric protein comprising a protein of interest (POI) fused to a DNA-binding domain (DBD) via a flexible linker, - a nucleotide sequence encoding a reporter protein which is under transcriptional control being activated or repressed by said chimeric protein comprising a protein of interest (POI) fused to a DNA-binding domain (DBD) via a flexible linker once said protein of interest (POI) associates such that said chimeric protein homodimerizes,

- an origin of replication, and

- a selection system.

The function of the plasmids and their use for detecting and/or quantifying protein homodimerization is illustrated here by two novel homodimerization screening assays based on ToxR and LexA (Fig. 1). In contrast to GALLEX and TOXCAT, we designed these assays to detect homodimerization of soluble proteins. As described above, the mechanism of action of both systems is based on the activity of the chimera produced by exchange of the dimerization domain with the POI. In the case of LexA, the POI is expressed in the cytosol, and in the case of ToxR, it is expressed in the periplasm. In both cases, the POI is connected to the DBD via a flexible linker, which allows considerable freedom of rotation of the POI. Upon association of the POI, the chimera homodimerizes and activates transcription of the reporter protein in the case of ToxR and represses it in the case of LexA. We prepared two variants of the LexA-based system and three variants of the ToxR based system, wherein the variants differ in the type of the reporter protein. TOXGFP and LEXGFP assays use green fluorescent protein (GFP) as the reporter, LEXBLUE and TOXBLUE use - galactosidase and the reporter in the TOXKAN system is aminoglycoside phosphotransferase, which confers resistance to kanamycin.

To make these systems as user-friendly as possible, we chose E. coli DH5a as the chassis for screening. Since ToxR is only present in the Vibrio genus, we did not expect any interference with the chassis. However, E. coli endogenously expresses LexA, so interference was expected. To ensure orthogonality with the chassis, we used a DBD of a mutant LexA, LexA408, and the corresponding mutant operator region not bound by endogenous repressor ²². Unlike existing LexA-based assays, the use of specific LexA knock-out strains is unnecessary, making screening simpler and less expensive. In addition, to prepare the TOXKAN system, we used an attenuated form of the promoter carrying a point mutation (A-41G) ²³. By weakening the promoter, we were able to reduce its leakage and thus enable selection at lower concentrations of kanamycin. Especially due to their screening potential, the novel LexA- and ToxR-based assays significantly expand the range of applications offered by two-hybrid systems and the already existing homodimerization assays described above. The potential applications of the novel systems are as follows:

1. In vivo study of protein oligomerization state

2. Drug discovery of homodimerization inhibitors and activators

3. Protein engineering of oligomeric or monomeric proteins

4. Study of the evolution of protein oligomerization

The plasmid systems

In an embodiment the plasmid according to the invention comprises the DBD from the bacterial transcription factor LexA. In another embodiment the DBD is the LexA mutant LexA408 (SEQ ID NO:1).

In another embodiment the chimeric protein has the structure LexA408 DBD + linker + POL In another embodiment the chimeric protein has the structure LexA408 DBD + linker is SEQ ID NO:2.

In another embodiment the DBD is from the bacterial transcription factor ToxR, such as SEQ ID NO:3.

In another embodiment the expression of the chimeric protein is under control of the native promoter for the DBD.

In another embodiment the expression of the chimeric protein is under control of the ToxR promoter (SEQ ID NO:4).

In another embodiment said chimeric protein has the structure ToxR DBD + ToxR transmembrane region + linker + POL In another embodiment the ToxR transmembrane region is SEQ ID NO:5. In yet another embodiment the part of the chimeric protein being ToxR DBD + ToxR transmembrane region + linker is SEQ ID NO:6. Many different linkers will be applicable, the main point being the linker is flexible. In an embodiment the flexible linker is a G/S rich linker, such as having more than about 50% amino acid residues being G or S, or having more than 75% amino acid residues being G or S. In another embodiment the flexible linker has a length of about 6 to 25 amino acid residues, of about 6 to 18 amino acid residues or of about 10 to about 15 amino acid residues. In another embodiment the flexible linker is EASSGGGSGGGSSR (SEQ ID NO: 7) or ASSGGGSGGGSSR (SEQ ID NO:8).

In another embodiment the reporter protein is green fluorescent protein (GFP), p-galactosidase ( -gal) or kanamycin resistance (KanR). In another embodiment the reporter protein is green fluorescent protein (GFP). In another embodiment the reporter protein is p-galactosidase (P-gal). In another embodiment the reporter protein is kanamycin resistance (KanR).

In another embodiment the reporter protein is expressed under control of the cholera toxin promoter (P_rtx) (SEQ ID NO: 9), the A-41G mutant cholera toxin promoter (SEQ ID NQ:10) or the mutant SulA promoter (PsuiA4os) (SEQ ID NO: 11).

In an embodiment the plasmid is LEXGFP or LEXBLUE. In another embodiment the plasmid is LEXGFP. In another embodiment the plasmid is LEXBLUE.

In another embodiment the plasmid is TOXGFP, TOXBLUE or TOXKAN. In another embodiment the plasmid is TOXGFP. In another embodiment the plasmid is TOXBLUE. In another embodiment the plasmid is TOXKAN.

In another aspect there is provided a library of plasmids, wherein the plasmids in the library are identical except for the part encoding the POI of the chimeric protein.

In yet another aspect is provided a bacterium transformed with the plasmid or the library of plasmids.

In another embodiment the bacterium is Escherichia coli such as E. coli DH5a.

In another aspect is provided the use of the plasmid for the qualitative or quantitative detection of homodimerization of the protein of interest (POI), said use comprising transforming said plasmid into a bacterium, cultivation of said bacterium and detection or quantification of the increase or decrease of the expression of the reporter gene. In another embodiment the use of the library of plasmids for the detection or quantification of homodimerization of a multitude of proteins of interest (POI), wherein said library of plasmids comprises plasmids having the genes encoding the multitude of POI, said use comprising transforming said library of plasmids into bacteria, cultivation of said bacteria, and detection or quantification of the increase or decrease of the expression of the reporter gene.

In another embodiment the use of the plasmid or library of plasmids comprises the detection or quantification is done by flow cytometry.

Applications of the developed systems.

In addition, TOXGFP and LEXGFP can easily be used for screening directly on agar plates, for example when screening mutant protein libraries. However, an appropriate imaging instrument is required. To make these systems more accessible for screening mutant protein libraries, we have developed new variants of both systems, namely LEXBLUE and TOXBLUE. These variants use p-galactosidase as the reporter, and the results of this study clearly demonstrate that the systems allow blue-white screening on solid agar plates, p-galactosidase activity can also be measured in cultures grown in multiwell plates using the Miller assay; however, the GFP variants of the systems allow for much less labor-intensive and more cost-effective detection of reporter expression. In addition, detection of fluorescence is much more sensitive, which is particularly advantageous for ToxR-based assays where induction of reporter expression is weak.

The TOXKAN system was developed to simplify the screening process by requiring only one agar plate to screen the entire library. In addition, TOXKAN can be used for screening in multiwell plates, for example, when screening libraries of potential stabilizers or destabilizers of homodimerization. Furthermore, TOXKAN can be used to discriminate between weak and strong homodimers if an appropriate kanamycin concentration is used.

Examples

Materials and methods.

Bacterial strains

E. coll DH5a was chosen as the chassis for the genetic circuit and was therefore used in all experiments.

PCR primers used for the construction and sequencing of plasmids for LexA- and ToxR-based assays are listed in Table 1.

Table 1. PCR primers.

Example 1 - Plasmid construction

To properly validate these screening assays, we prepared four control constructs for each system. The controls differ in the POI and thus their dissociation constant of homodimerization (K_o). We chose maltose binding protein (MBP) as the negative control for homodimerization and the preduplication ancestor of haemoglobin (AncHb, K_o ~ 9 pM ²⁴), the mutant dimerization domain of LexA (LexAC G14D, K_D ~ 0.6 pM ²⁵), and its wild-type counterpart (LexAC, K_D ~ 10 nM²⁵) for the positive controls. To avoid protein aggregation and false positives due to overexpression ¹⁵, we chose to express the chimeric transcription factors under the control of their native promoters, namely Pi_eXA and P_tOxR. The reporter proteins were expressed under the control of the cholera toxin promoter (P_rtx) in the ToxR-based assays and under control of the mutant SulA promoter ( PSUIAAOS) in the LexA-based assays.

Because E. coli DH5a was chosen as the chassis for screening, we combined the entire synthetic gene circuit on one plasmid. For the construction of LEXGFP, we used the plasmid pUCBB-eGFP, which was aquired from Addgene and already contains the coding sequence for enhanced GFP (eGFP) ²⁶. The sequences encoding the rest of the gene circuit, namely Pi_eXA, the LexA408 DBD, the flexible linker, the transcription terminator, and P_SUIA408, were synthesized as a single part (Twist Biosciences), digested with EcoR\ and Nde\ and ligated into the compatible EcoR\ and Nde\ restriction sites in pUCBB-eGFP using T4 DNA ligase. For the construction of TOXGFP, we used pccGFPKAN, which we acquired from Addgene ²⁷ This plasmid already contains the coding sequences for superfolder GFP (sfGFP) and ToxR DBD under the control of P_rtx and P_tOxR, respectively. The sequences encoding the rest of the gene circuit, namely the ToxR transmembrane region, the flexible linker and the transcription terminator, were synthesized as a single part (Twist Biosciences), digested with Nhel and Sacl and ligated into the compatible Nhe\ and Sacl restriction sites in pccGFPKAN using T4 DNA ligase.

To facilitate and standardize the cloning of POI coding sequences into the gene circuit, the above synthetic sequences for both LexA- and ToxR-based assays were designed to contain recognition sites for the restriction enzymes Xba\ and BamHI between the flexible linker and the terminator sequence. The coding sequences for MBP, Anca/P, LexAC G124D and LexAC were synthesised (Twist Biosciences) and the restriction enzyme recognition sites were included at the ends of the synthetic gene.

The TOXBLUE and LEXBLUE systems and their respective control constructs were prepared by replacing the GFP gene with the gene encoding the a-fragment of p-galactosidase (LacZa). In case of TOXBLUE, the reporter was replaced using the In Vivo Assembly method (IVA) ²⁸. In case of LEXBLUE, the cDNA coding for LacZa was amplified from pUC19 plasmid and the product was digested by Nde\ and Xho\ and then ligated into the compatible restriction sites in the cut LEXGFP vector.

The TOXKAN system was prepared by replacing the GFP gene with the gene encoding the kanamycin resistance using the In Vivo Assembly method. The kanamycin resistance cassette was amplified by PCR from the pET28 plasmid. Additionally, the cholera toxin promoter was weakened by the introduction of a point mutation A-41G ²³. By weakening the promoter we minimized the leakage of aminoglycoside phosphotransferase expression and thereby enabled the selection with TOXKAN at low kanamycin concentrations.

The oligonucleotide primers used for preparing the constructs are listed in Table 1.

Control constructs for ToxR and LexA-based assays were confirmed by sequencing (Eurofins Genomics). The plasmid maps of LEXGFP, LEXBLUE, TOXGFP, TOXBLUE and TOXKAN are shown in Figures 9-10.

Example 2 - TOXGFP and LEXGFP characterisation by spectral scanning E. coli DH5a were transformed with the control constructs of both systems. In addition, E. coli DH5a were transformed with a plasmid, lacking the POI coding sequence, which served as a control for promoter activity in absence of the homodimerizing moiety (the constructs were named LEXGFP- ctrl and TOXGFP-ctrl). Untransformed DH5a were also used as a control for background fluorescence in the chassis. Individual colonies were inoculated into 10 mL LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) and grown overnight at 37°C. 500 pL of each overnight culture medium was pelleted at 4000 ref for 5 minutes and then resuspended in 1 mL PBS (137 mM NaCI, 2.7 mM KCI, 10 mM Na2HPO4, pH 7.4) to an approximate OD600 of 0.5. 200 pL of each control sample was then transferred to a 96-well Greiner CELLSTAR® microplate (clear, F- bottom). Emission spectral scanning was performed in a BioTek Synergy Hl plate reader, using an excitation wavelength of 485 nm and measuring emission from 500 nm to 600 nm. The exact optical density at 600 nm was also measured for each sample. The fluorescence of PBS was subtracted from the recorded emission intensities, which were then normalized to the measured optical density of the sample.

Example 3 - TOXGFP and LEXGFP characterisation by endpoint fluorescence measurement

Single colonies of untranformed DH5a and DH5a transformed with the control constructs of both systems were inoculated into 10 mL LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) and grown overnight at 37°C. 2.5 pL of each control cell culture was inoculated into the wells of 48-well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 300 pL of fresh LB broth (containing 100 pg/mL ampicillin in the case of transformed cells). Cells were then grown overnight at 37 °C in a BioTek Synergy Hl plate reader, with OD600 and fluorescence intensity (at an excitation wavelength of 485 nm and an emission wavelength of 510 nm) measured every hour. The fluorescence of fresh LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) was subtracted from the recorded emission intensities, which were then normalized to the measured optical density of the sample.

Example 4 - Determination of the screening potential of TOXGFP and LEXGFP

E. coli DH5a transformed with the control constructs were plated on LB agar plates with 100 pg/mL ampicillin and incubated at 37°C for 18 hours. In addition, a mixture of E. coli DH5a transformed with the negative and the strongest positive control was plated on a LB agar plate containing 100 pg/mL ampicillin. After the initial incubation, the plates containing the TOXGFP controls were incubated for eight hours at room temperature and then overnight at 4°C. Plates containing individual colonies were then imaged using the Bio-Rad ChemiDoc Imaging System. The Alexa Fluor 488 filter set was used for imaging, and the exposure time was set to 0.347 and 0.529 seconds for LEXGFP and TOXGFP, respectively. Negatives of the images were made using Bio-Rad Image Lab 6.1.

Example 5 - TOXBLUE and LEXBLUE characterisation with the -galactosidase assay

LEXBLUE - Single colonies of untranformed DH5a and DH5a transformed with the control constructs were inoculated into 10 mL LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) and grown overnight at 37°C. 2.5 pL of each overnight culture was inoculated into the wells of a 48-well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 300 pL LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) with 0.5 mM IPTG and grown at 37 °C to an OD600 of approximately 0.6 (log phase of bacterial growth). The exact OD600 was determined for each sample, then 10 pL of the culture was transferred into the wells of a 96- well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 40 pL permeabilization solution (100 mM NajHPC , 20 mM KCI, 2 mM MgSO₄, 0.8 mg/mL CTAB, 0.4 mg/mL sodium deoxycholate, 5.4 pL/mL P-mercaptoethanol) and incubated for 10 minutes at room temperature. Cell lysates were transferred into the wells of 48-well Greiner CELLSTAR® multiwell plate (clear, F- bottom) containing 300 pL of prewarmed (37 °C) substrate solution (60 mM NajHPC , 40 mM NaHjPC , 1 mg/mL o-nitrophenyl-p-D-galactoside (ONPG), 2.7 pL/mL P-mercaptoethanol). The plate was incubated at 37 °C for 20 minutes, then 350 pL of stop solution (1 M NajCOs) was added. The absorbance at 420 nm was measured using a BioTek Synergy Hl plate reader.

TOXBLUE - Single colonies of untranformed DH5a and DH5a transformed with the control constructs were inoculated into 10 mL LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) and grown overnight at 37°C. 2.5 pL of each overnight culture was inoculated into the wells of a 48-well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 300 pL LB broth (containing 100 pg/mL ampicillin in the case of transformed cells) with 0.25 mM IPTG and grown overnight at 37 °C (stationary phase of bacterial growth). The exact OD600 was determined for each sample, then the cells were pelleted (5 minutes, 3000 ref) and resuspened in 60 pL of LB broth. 20 pL of the concentrated culture was transferred into the wells of a 96-well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 80 pL permeabilization solution (100 mM NajHPC , 20 mM KCI, 2 mM MgSO₄, 0.8 mg/mL CTAB, 0.4 mg/mL sodium deoxycholate, 5.4 pL/mL P-mercaptoethanol) and incubated for 10 minutes at room temperature. Cell lysates were transferred into the wells of 48-well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 300 pL of prewarmed (37 °C) substrate solution (60 mM Na₂HPO₄, 40 mM NaH₂PO₄, 1 mg/mL o- nitrophenyl-p-D-galactoside (ONPG), 2.7 pL/mL p-mercaptoethanol). The plate was incubated at 37 °C for 30 minutes, then 350 pL of stop solution (1 M Na₂COa) was added. The multiwell plate was then centrifuged (5 min, 3000 ref) and the supernatant was transferred into the wells of a new a 48-well Greiner CELLSTAR® multiwell plate (clear, F-bottom). The absorbance at 420 nm was measured using a BioTek Synergy Hl plate reader.

Calculation of Miller units - Miller units of p-galactosidase activity for both assays were calculated using the following formula:

MU = 1000

where A₄₂o is the absorbance of the sample; A₄₂o, blank is the absorbance of the blank solution (i.e., a mixture of permeabilization, substrate, and stop solution); ODgoo is the optical density of the cell culture before permeabilization; V is the volume of culture used in the assay; and t is the incubation time after addition of the cell lysate to the substrate solution.

Example 6 - Determination of the potential of TOXBLUE and LEXBLUE for blue-white screening

E. coli DH5a was transformed with the control plasmids and plated on LB agar plates containing 100 pg/mL ampicillin, 0.25 mM IPTG, and 40 pg/mL X-gal. The plates were incubated at 37 °C for 18 hours. Blue and white coloring of the control colonies was determined by naked eye.

Example 7 - Characterization of the TOXKAN system by selection with kanamycin

Single colonies of E. coli DH5a transformed with the control constructs of the TOXKAN system were inoculated into 10 mL LB broth and grown overnight at 37°C. 2.5 pL of each overnight culture was inoculated into the wells of a 48-well Greiner CELLSTAR® multiwell plate (clear, F-bottom) containing 300 pL LB broth with the following concentrations of kanamycin: 0, 50, 75, 100, 150, 200 and 300 pg/mL. Each combination of the control type and kanamycin concentration was measured in triplicate. The cells were grown overnight at 37 °C in a BioTek Synergy Hl plate reader, wherein OD600 was measured every hour.

Bacteria transformed with the control constructs were then plated on LB agar plates containing 100 pg/mL kanamycin and grown overnight at 37 °C. Example 8 - LEXGFP performance

LEXGFP was characterized by measuring GFP fluorescence in E. coli DH5a transformed with the control plasmids (Fig. 2). Wild-type LexAC, LexAC G124D, and AncHb were used as model proteins for homodimerization with different affinities, and MBP was used as a negative control for homodimerization.

As described above, LexA is a repressor and binds only in homodimeric form to its operator region- the 20 base-pair-long imperfect palindrome known as the SOS-box. Therefore, it was expected that the highest GFP expression would be detected in the case of MBP, which does not form homodimers. Previously, it was reported that the level of induction of SOS genes depends, among other factors, on the affinity of LexA homodimers for the SOS-box ²⁹. Therefore, it is possible that the extent of expression of the reporter protein in the LexA-based assays also depends on the homodimerization affinity of the chimera, as this determines the concentration of the homodimeric form of the repressor in the cell. Consequently, we expected GFP fluorescence to depend on the KD for homodimerization of the POL In contrast, the results of spectral scanning and end-point fluorescence measurements showed a binary, that is, an on-off relationship between the oligomeric state of the chimera and the reporter signal. Only the negative control and the control without the POI exhibited GFP fluorescence. The expression of GFP was almost completely suppressed in all three positive controls. It is important to note, however, that the fluorescence of the positive controls was significantly higher than the fluorescence of untransformed cells until the stationary phase of the growth curve was reached (Fig. 11). The reason for this initial deviation might be that GFP and the chimera are expressed simultaneously from the same plasmid and therefore there are not enough active chimeras to repress all transcription units when the cells are still in log phase. Fluorescence intensity should therefore be measured in the stationary phase of growth.

Moreover, GFP fluorescence in the absence of the POI was approximately 1.4-fold higher than the fluorescence of the negative control for homodimerization. This is surprising because LexA is known to bind the sulA promoter rather weakly, resulting in a higher induction ratio as well as a higher baseline level of protein expression ³⁰- ³¹. A possible explanation for this deviation from previous results could be the use of the mutant P_SUIA408, whose affinity for LexA408 remains to be determined. However, the affinity of LexA408 for a mutant recA, i.e. the recA408 promoter, has been determined previously and is about 10-fold lower than the affinity of wild-type LexA for the corresponding promoter ²². Moreover, the SOS-box of the mutant operator diverges even further from the consensus sequence than its wild-type counterpart, and therefore a lower affinity and consequently a higher baseline level of reporter expression would be expected ³¹. Another possible explanation is that the cellular concentration of the chimera was high enough that the homodimeric form predominated in all controls, which would also explain the complete repression of GFP expression in all three positive controls. High protein concentration could also lead to nonspecific interactions between the chimeras, which could explain the observed partial repression of GFP expression in the negative control.

Nevertheless, these results suggest that the LEXGFP system can be used to identify homodimeric proteins; however, it cannot be used to determine approximate homodimerization affinities when the value of K_o is ~ 10 pM or lower.

To determine whether LEXGFP can also be used as a method for screening, we plated the bacteria transformed with the control constructs ob solid LB agar plates and used Bio-Rad ChemiDoc Imaging System for fluorescence imaging (Fig. 3).

The results show that the difference in fluorescence intensity can be detected on agar plates with the use of an imaging system. Colonies transformed with the negative control and the control without the POI fluoresced intensely, whereas colonies transformed with the positive controls exhibited no fluorescence. The difference in fluorescence was evident when a mixture of the negative and the strongest positive control was plated on the same plate. This means that the LEXGFP system can be easily used for screening directly on agar plates.

Example 9 - LEXBLUE performance

The above results have shown that LEXGFP can be used to determine the oligomerization state of the POI, and although it can be used for screening mutant libraries, it requires special equipment (an imaging system for screening directly on agar plates). To make the system more accessible, we replaced GFP with the alpha fragment of p-galactosidase to enable blue-white screening. We named the new system LEXBLUE.

First, we quantitatively characterized the LEXBLUE system by measuring p-galactosidase activity in the crude cell lysates using the chromogenic substrate ONPG (Fig. 4A). As expected, enzyme activity was detected only in the negative control and in the control without the POI. The latter exhibited approximately 1.3-fold higher activity than the former, which corresponds well with the fluorescence measurements. In addition, basal expression of LacZa was not reported, which is also consistent with the LEXGFP system. The induction ratio was much lower than in the LEXGFP system, which is probably due to the inducible expression of p-galactosidase compared with the constitutive expression of GFP. However, the induction ratios are not comparable because the relative p- galactosidase activity is highly dependent on the incubation time in the substrate solution when the Miller assay is performed. Nevertheless, the Miller assay - analogous to the fluorescence measurements in the LEXGFP system - indicates that the LEXBLUE system provides only a binary response to protein oligomerization, at least in case K_o is lower than ~ 10 pM.

To test the screening potential of the LEXBLUE system, we performed a blue-white assay - bacteria transformed with the control constructs were plated on agar plates with IPTG and X-gal (Fig. 4B). The plates were incubated at 37 °C for approxiately 16 hours. As expected, colonies were intensely blue in the case of the negative control and white in the case of the positive controls, indicating that the LEXBLUE system is suitable for screening mutant libraries of the POL The white colonies of the positive controls did not develop a blue coloring neither after a prolonged incubation at 37 °C nor after being stored at 4 °C for a couple of days. This suggests that the probability of acquiring false negative results with the LEXBLUE system is low.

Example 10 - TOXGFP performance

TOXGFP was characterized by measuring GFP fluorescence in E. coli DH5a transformed with the control constructs. The model proteins used as controls for homodimerization were the same as in the LEXGFP system. Analogous to the LEXGFP system, the ratios between fluorescence intensities stabilized around the time the growth curve reached the stationary phase (Fig. 12).

In contrast to LEXGFP, the spectral scans and end-point fluorescence measurements (Fig. 5) all show a mostly continuous relationship between homodimerization affinity and reporter signal, which agrees well with the results of the TOXGREEN assay, developed by Claire R. Armstrong et al. ²⁷. GFP fluorescence was highest when the periplasmic domain was replaced by wild-type LexAC or LexAC G124D, approximately 1.6-fold lower when replaced by AncHb, and 3-fold lower when replaced by MBP. There are several possible explanations for the same level of reporter protein expression in the case of LexAC and LexAC G124D - it is possible that the upper limit of reporter protein expression in the TOXGFP system is reached when the K_o for dimerization is in the nanomolar range, implying that the system can be used to discriminate between weak and stronger interactions but not to discriminate between interactions with K_o for dimerization in the nanomolar range or below. Another possible explanation is that the K_o for dimerization of the LexAC G124D chimera is lower than that of the unfused LexAC mutant. This is because the chimera is inserted into the inner membrane of E. coli, which means that the rotational freedom of the POI is lower than the rotational freedom of an unfused, cytosolic POI. The loss of rotational freedom could theoretically lead to a decrease in K_o for homodimerization. Further studies are needed to determine the definitive cause of the apparent upper limit of detection of the TOXGFP system.

Low GFP expression was also detected in the negative control for dimerization (MBP as the POI), indicating a leaky expression of the reporter protein in the TOXGFP system. The control without the POI did not exhibit significant GFP fluorescence, which is not surprising since it was previously reported that the periplasmic domain is critical for ToxR stability ³². Thus, it is highly likely that the transcription factor does not fold properly without the periplasmic domain. Therefore, the leakage of the TOXGFP system is probably caused by the chimeric transcription factor, despite the inability of its periplasmic domain to form homodimers. One possible explanation for this contradiction is that the oligomerization of the ToxR protein is also greatly facilitated by its binding sites on DNA ³³. Moreover, under physiological conditions, the native periplasmic domain of ToxR allows homodimerization only under inducing conditions; however, it has been reported that the wildtype ToxR can form homodimers when overexpressed and activate transcription even in the absence of inducing conditions ¹⁵- ³³. The chimeric transcription factor in the TOXGFP system is transcribed under the native toxR promoter from a low-copy plasmid. However, the research group of Ottemann and Mekalanos reported that even slight changes in ToxR copy number can lead to drastic changes in its oligomeric state ¹⁵. Therefore, despite the use of a low-copy plasmid, the copy number of the chimera in the TOXGFP system could exceed the threshold for spontaneous oligomer formation. Therefore, the presence of multiple ToxR DNA binding sites together with multiple copies of the chimera could lead to weak homodimerization of the negative control in the cell and consequently to a weak background signal in the TOXGFP system.

In summary, these results suggest that the TOXGFP system can be used to detect homodimerizing interactions of soluble proteins and to discriminate between weak and stronger interactions. However, to avoid false-positive results due to leaky expression, it should always be used in parallel with the negative control for homodimerization.

To test the system's screening potential, we plated the controls on solid agar plates and used the Bio-Rad ChemiDoc Imaging System for fluorescence imaging (Fig. 6). Imaging results showed that there is noticeable difference in fluorescence intensity between TOXGFP controls. As can be seen on the plate with a mixture of the negative and the strongest positive control, the difference is not as clear as in the LEXGFP system. Consequently, caution should be exercised when screening directly on plates and screening should be performed in parallel with the negative and positive controls of the system.

Example 11 - TOXBLUE performance

Analogous to the LexA-based system, the ToxR system was adapted by replacing GFP with the a- fragment of p-galactosidase, which should theoretically allow blue-white screening on solid agar plates with IPTG and X-gal. The new system, designated TOXBLUE, was first characterised by the Miller assay (Fig. 7A).

Unlike the LEXBLUE system, which used cell cultures at the start of the stationary phase for the assay, the TOXBLUE system required cultures grown overnight (OD600 ~ 1) - no colour development could be detected when early stationary phase cultures were used. In addition, cells had to be pelleted prior to the assay for the colour change to be detectable. Together with the results of TOXGFP characterization, which showed much weaker fluorescence than the LEXGFP system, this suggests that the induction of reporter expression in the ToxR-based systems is much weaker than in the LexA-based systems. Nevertheless, the results of the Miller assay were reproducible in the stationary phase and were consistent with the fluorescence measurements - the p-galactosidase activity was highest when the periplasmic domain was replaced by wild-type LexAC or LexAC G124D, about 1.5-fold lower when replaced by AncHb, and 4-fold lower when replaced by MBP. No activity was detected in the control without the POI, which is also consistent with the fluorescence measurements. The background absorbance was much weaker in comparison with LEXBLUE - this is because the precipitate in the TOXBLUE samples was pelleted prior to absorbance measurements in order to further increase the sensibility of the assay. This was not required in case of LEXBLUE because the induction of expression was much higher than in the TOXBLUE system.

To confirm that TOXBLUE is suitable for blue-white screening despite its low reporter expression and leakiness, we plated E. coli DH5a on LB agar plates containing ampicillin, IPTG, and X-gal (Fig. 7B). The plates were incubated overnight at 37 °C. Blue staining of the positive controls was detectable after 19 hours of incubation. The colonies of the negative control did not turn blue even after prolonged incubation at 37 °C or after storage at 4 °C for several days. This indicates that there is no risk of false positive results despite the leakiness of the TOXBLUE system. However, it should be noted that the plates of the TOXBLUE system should be incubated at 37 °C for at least 19 hours, otherwise there is a risk of false negative results. As mentioned above, all three positive controls formed blue colonies, wherein the staining of the colonies of the control with the lowest affinity control (AncHb) was less intense than the staining of the controls with intermediate (LexAC G124D) and strong affinity (LexAC). This is consistent with the results of fluorescence measurements and the Miller assay; however, the blue-white screen should not be used to determine the approximate affinity for homodimerization because staining is a subjective matter and is also highly dependent on external conditions. If the approximate range of K_o is the goal of a study, the Miller assay or the fluorescence measurements of the TOXGFP system are more appropriate. However, when attempting to determine the approximate K_o with TOXBLUE or TOXGFP, one should always use the reference controls in parallel.

Example 12 - TOXKAN performance

The GFP and BLUE variants were found to be suitable for direct screening on agar plates, however, when screening large libraries, plating can be time and material consuming. To improve screening efficiency, we developed another variant of the ToxR-based system that uses aminoglycoside phosphotransferase as the reporter. This enzyme confers resistance to kanamycin, which means that theoretically only the positive controls should be able to grow in the presence of kanamycin. Consequently, the entire library can be plated on a single agar plate and only the colonies with homodimeric mutants should form colonies.

To determine the appropriate kanamycin concentration for screening, we cultured E. coli DH5a, transformed with the negative and positive controls in the presence of increasing kanamycin concentrations. The growth curves are shown in Fig. 8A.

Considering the results of the fluorescence measurements and the Miller assay, we assumed that due to leaky expression, the negative control would grow at low concentrations of kanamycin. To reduce the level of leakiness, we weakened the cholera toxin promoter by introducing a point mutation (A-41G). The results showed that the growth of the negative control was extremely weak in the presence of 50 and 75 pg/mL kanamycin and stopped growing completely at 100 pg/mL kanamycin. The growth of the weakest positive control was only slightly inhibited in the presence of 100 pg/mL kanamycin, and the growth of the strongest positive controls was not inhibited at all. The weakest positive control stopped growing completely in the presence of 300 pg/mL kanamycin, whereas the growth of the stronger positive controls was not disrupted. This implies that 300 pg/mL kanamycin can be used when only the detection of stronger interactions (nanomolar range) is desired.

To test the screening potential of the TOXKAN system, we plated the bacteria transformed with the controls on solid agar plates containing 100 pg/mL kanamycin (Fig. 8B). As expected, all three positive controls formed colonies, whereas no colonies were found on the plate containing the negative control. This indicates that 100 pg/mL kanamycin is a suitable concentration for screening.

References

1. M. Gross: Symmetry and complexity in protein oligomers. Curr. Biol. 2012, 22(6), pp. R175- R177.

2. D. S. Goodsell, A. J. Olson: Structural symmetry and protein function. Annu. Rev. Biophys. Biomol. Struct. 2000, 29, pp. 105-153.

3. H. Liu, M. Cao, Y. Wang, B. Lv, C. Li: Bioengineering oligomerization and monomerization of enzymes: learning from natural evolution to matching the demands for industrial applications, https://doi.org/10.1080/07388551.2019.17110142020, 40(2), pp. 231-246.

4. F. Rigoldi, S. Donini, A. Redaelli, E. Parisini, A. Gautieri: Review: Engineering of thermostable enzymes for industrial applications. APL Bioeng. 2018, 2(1).

5. A. Basso, S. Serban: Industrial applications of immobilized enzymes— A review. Mol. Cotai. 2019, 479, pp. 110607.

6. M. Vivoli, J. Pang, N. J. Harmer: A half-site multimeric enzyme achieves its cooperativity without conformational changes OPEN.

7. N. Iqbal, N. Iqbal: Human Epidermal Growth Factor Receptor 2 (HER2) in Cancers: Overexpression and Therapeutic Implications. Mol. Biol. Int. 2014, 2014, pp. 1-9.

8. B. Ma, R. Nussinov: Simulations as analytical tools to understand protein aggregation and predict amyloid conformation. Curr. Opin. Chem. Biol. 2006, 10(5), pp. 445-452.

9. B. O'Nuallain, D. B. Freir, A. J. Nicol I, E. Risse, N. Ferguson, C. E. Herron, J. Collinge, D. M. Walsh: Amyloid -Protein Dimers Rapidly Form Stable Synaptotoxic Protofibrils. J. Neurosci. 2010, 30(43), pp. 14411.

10. B. Goyal, D. Goyal: Targeting the Dimerization of the Main Protease of Coronaviruses: A Potential Broad-Spectrum Therapeutic Strategy. ACS Comb. Sci. 2020, 22(6), pp. 297-305.

11. E. Furuta, K. Yamamoto, D. Tatebe, K. Watabe, T. Kitayama, R. Utsumi: Targeting protein homodimerization: A novel drug discovery system. FEBS Lett. 2005, 579(10), pp. 2065- 2070.

12. I. V. Korendovych: Rational and Semirational Protein Design. Methods Mol. Biol. 2018, 1685, pp. 15.

13. B. Stynen, H. Tournu, J. Tavernier, P. Van Dijck: Diversity in Genetic In Vivo Methods for Protein-Protein Interaction Studies: from the Yeast Two-Hybrid System to the Mammalian Split-Luciferase System . Microbiol. Mol. Biol. Rev. 2012, 76(2), pp. 331-382.

14. M. Butala, D. Z “ Gur-Bertok, S. J. W. Busby: The bacterial LexA transcriptional repressor.

15. K. M. Ottemann, J. J. Mekalanos: The ToxR Protein of Vibrio cholerae Forms Homodimers and Heterodimers. J. Bacterial. 1996, 178(1), pp. 156-162.

16. N. Gubensak, E. Schrank, C. Hartlmuller, C. Gobi, F. S. Falsone, W. Becker, G. E. Wagner, S. Pulido, N. H. Meyer, ... K. Zangger: Structural and DNA-binding properties of the cytoplasmic domain of Vibrio cholerae transcription factor ToxR. J. Biol. Chem. 2021, 297(4), pp. 101167.

17. W. P. Russ, D. M. Engelman: TOXCAT : A measure of transmembrane helix association in a biological membrane. Proc. Natl. Acad. Sci. U. S. A. 1999, 96(3), pp. 863.

18. D. Schneider, D. M. Engelman: GALLEX, a Measurement of Heterologous Association of Transmembrane Helices in a Biological Membrane*. 2002.

19. M. Dmitrova, G. Younes-Cauet, P. Oertel-Buchheit, D. Porte, M. Schnarr, M. Granger- Schnarr: A new LexA-based genetic system for monitoring and analyzing protein heterodimerization in Escherichia coli. Mol. Gen. Genet. MGG 199825721998, 257(2), pp. 205-212.

20. B. Manta, D. Boyd, M. Berkmen: Disulfide Bond Formation in the Periplasm of Escherichia coli. EcoSal Plus 2019, 8(2).

21. F. Hennecke, A. Muller, R. Meister, A. Strelow, S. Behrens: A ToxR-based two-hybrid system for the detection of periplasmic and cytoplasmic protein-protein interactions in Escherichia coli: minimal requirements for specific DNA binding and transcriptional activation. Protein Eng. Des. Sei. 2005, 18(10), pp. 477-486.

22. A. T. Thliveris, J. W. Little, D. W. Mount: Repression of the E coli recA gene requires at least two LexA protein monomers. Biochimie 1991, 73(4), pp. 449-456. 23. J. D. Pfau, R. K. Taylor: Genetic footprint of the ToxR-binding site in the promoter for cholera toxin. Mol. Microbiol. 1996, 20(1), pp. 213-222.

24. A. S. Pillai, S. A. Chandler, Y. Liu, A. V. Signore, C. R. Cortez-Romero, J. L. P. Benesch, A. Laganowsky, J. F. Storz, G. K. A. Hochberg, J. W. Thornton: Origin of complexity in haemoglobin evolution. Nat. 202058178092020, 581(7809), pp. 480-485.

25. K. C. Giese, C. B. Michalowski, J. W. Little: RecA-Dependent Cleavage of LexA Dimers. J. Mol. Biol. 2008, 377(1), pp. 148.

26. J. E. Vick, E. T. Johnson, S. Choudhary, S. E. Bloch, F. Lopez-Gallego, P. Srivastava, I. B. Tikh, G. T. Wawrzyn, C. Schmidt-Dannert: Optimized compatible set of BioBrick™ vectors for metabolic pathway engineering. Appl. Microbiol. Biotechnol. 2011, 92(6), pp. 1275-1286.

27. C. R. Armstrong, A. Senes: Screening for transmembrane association in divisome proteins using TOXGREEN, a high-throughput variant of the TOXCAT assay. Biochim. Biophys. Acta - Biomembr. 2016, 1858(11), pp. 2573-2583.

28. J. Garci'a-Nafn'a, J. F. Watson, I. H. Greger: IVA cloning: A single-tube universal cloning system exploiting bacterial In Vivo Assembly. Sci. Reports 2016 61 2016, 6(1), pp. 1-12.

29. O. Gillor, J. A. C. Vriezen, M. A. Riley: The role of SOS boxes in enteric bacteriocin regulation. Microbiology 2008, 154(Pt 6), pp. 1783.

30. J. Dreier, E. B. Breitmaier, E. Goeke, C. M. Apfel, M. G. P. Page: Direct influence of S9 liver homogenate on fluorescence signals: impact on practical applications in a bacterial genotoxicity assay. Mutat. Res. Toxicol. Environ. Mutagen. 2002, 513(1-2), pp. 169-182.

31. A. Norman, L. H. Hansen, S. J. Sprensen: Construction of a ColD eda promoter-based SOS- green fluorescent protein whole-cell biosensor with higher sensitivity toward genotoxic compounds than constructs based on recA, umuDC, or sulA promoters. Appl. Environ. Microbiol. 2005, 71(5), pp. 2338-2346.

32. N. Gubensak, G. E. Wagner, E. Schrank, F. S. Falsone, T. M. I. Berger, T. Pavkov-Keller, J. Reidl, K. Zangger: The periplasmic domains of Vibriocholerae ToxR and ToxS are forming a strong heterodimeric complex independent on the redox state of ToxR cysteines. Mol. Microbiol. 2021, 115(6), pp. 1277. M. Lembke, T. Hbfler, A. N. Walter, S. Tutz, V. Fengler, S. Schild, J. Reidl: Host stimuli and operator binding sites controlling protein interactions between virulence master regulator ToxR and ToxS in Vibrio cholerae. Mol. Microbiol. 2020, 114(2), pp. 262-278.

Claims

- an origin of replication, and

- a selection system.

2. The plasmid according to claim 1, wherein the DBD is from the bacterial transcription factor

LexA.

3. The plasmid according to claim 1 or 2, wherein the DBD is the LexA mutant LexA408 (SEQ ID

NO:1).

4. The plasmid according to any one of claims 1-3, wherein said chimeric protein has the structure LexA408 DBD + linker + POI.

5. The plasmid according to claim 4, wherein the part of the chimeric protein having the structure LexA408 DBD + linker is SEQ ID NO:2

6. The plasmid according to claim 1, wherein the DBD is from the bacterial transcription factor

ToxR, such as SEQ ID NO:3

7. The plasmid according to any one of claims 1-4, wherein the expression of the chimeric protein is under control of the native promoter for the DBD.

8. The plasmid according to claims 6 or 7, wherein the expression of the chimeric protein is under control of the ToxR promoter (SEQ ID NO:4).ToxR promoter (SEQ ID NO:4).

9. The plasmid according to any one of claims 4-6, wherein said chimeric protein has the structure ToxR DBD + ToxR transmembrane region + linker + POL

10. The plasmid according to claim 7, wherein the ToxR transmembrane region is SEQ ID NO:5.

11. The plasmid according to claim 7, wherein the part of the chimeric protein being ToxR DBD + ToxR transmembrane region + linker is SEQ ID NO:6.

12. The plasmid according to any one of claims 1-5, wherein the flexible linker is a G/S rich linker, such as having more than about 50% amino acid residues being G or S, or having more than 75% amino acid residues being G or S.

13. The plasmid according to any one of claims 1-6, wherein the flexible linker has a length of about 6 to 25 amino acid residues, of about 6 to 18 amino acid residues or of about 10 to about 15 amino acid residues.

14. The plasmid according to any one of claims 1-7, wherein the flexible linker is EASSGGGSGGGSSR (SEQ ID NO: 7) or ASSGGGSGGGSSR (SEQ ID NO:8).

15. The plasmid according to any one of claims 1-8, wherein the reporter protein is green fluorescent protein (GFP), p-galactosidase ( -gal) or kanamycin resistance (KanR).

16. The plasmid according to claim 15, wherein the reporter protein is green fluorescent protein (GFP).

17. The plasmid according to claim 15, wherein the reporter protein is p-galactosidase (P-gal).

18. The plasmid according to claim 15, wherein the reporter protein is kanamycin resistance (KanR).

19. The plasmid according to any one of claims 1-9, wherein the reporter protein is expressed under control of the cholera toxin promoter (P_rtx) (SEQ ID NO: 9), the A-41G mutant cholera toxin promoter (SEQ ID NQ:10) or the mutant SulA promoter (PsuiA4os) (SEQ ID NO: 11).

20. The plasmid according to any one of claims 1-5 and 6-15 which is LEXGFP or LEXBLUE.

21. The plasmid according to any one of claims 1-5 and 6-15 which is LEXGFP.

22. The plasmid according to any one of claims 1-5 and 6-15 which is LEXBLUE.

23. The plasmid according to any one of claims 1 and 6-15, which is TOXGFP, TOXBLUE or

TOXKAN.

24. The plasmid according to claim 23, which is TOXGFP.

25. The plasmid according to claim 23, which is TOXBLUE.

26. The plasmid according to claim 23, which is TOXKAN.

27. Library of plasmids as defined in any one of claims 1-26, wherein the plasmids in the library are identical except for the part encoding the POI of the chimeric protein.

28. Bacterium transformed with the plasmid as defined in any one of claims 1-26 or the library of plasmids as defined in claim 27.

29. The bacterium according to claim 28, which is Escherichia coli such as E. coli DH5a.

30. Use of the plasmid as defined in any one of claims 1-26 for the qualitative or quantitative detection of homodimerization of the protein of interest (POI), said use comprising transforming said plasmid into a bacterium, cultivation of said bacterium and detection or quantification of the increase or decrease of the expression of the reporter gene.

31. Use of the library of plasmids as defined in claim 27 for the detection or quantification of homodimerization of a multitude of proteins of interest (POI), wherein said library of plasmids comprises plasmids having the genes encoding the multitude of POI, said use comprising transforming said library of plasmids into bacteria, cultivation of said bacteria, and detection or quantification of the increase or decrease of the expression of the reporter gene.

32. The use according to claim 30 or 31, wherein the detection or quantification is done by flow cytometry.