Vector construct

Abstract

The present invention provides a vector construct comprising the following components: (i) a sequence encoding a signal peptide which directs proteins into the Tat secretory pathway; and (ii) a sequence encoding a fluorophore fused to a sequence encoding a pVIII phage coat protein. Nucleic acid molecules comprising components (i) and (ii) are also provided, together with phage particles comprising such vectors or nucleic acid molecules and expressing a fluorophore-pVIII fusion protein on the surface. Methods for producing such fluorescent phage particles are also provided.

Claims

1. A vector construct comprising: (i) a polynucleotide sequence encoding a signal peptide which directs proteins into the Tat secretory pathway; and (ii) a polynucleotide sequence encoding a fluorophore fused to a polynucleotide sequence encoding a pVIII phage coat protein.

2. The vector construct of claim 1, wherein the fluorophore is a fluorophore having or comprising a β-barrel structure or architecture.

3. The vector construct of claim 2, wherein the fluorophore is green fluorescent protein (GFP), a GFP-derivative, or has GFP architecture.

4. The vector construct of claim 3, wherein the fluorophore is mNeonGreen comprising SEQ ID NO:4 or a variant sequence thereof having at least 70% identity to SEQ ID NO:4.

5. The vector construct of claim 1, wherein the signal peptide is TorA comprising SEQ ID NO:32 or a variant sequence thereof having at least 70% identity to SEQ ID NO:32.

6. The vector construct of claim 1, wherein the signal peptide is Tor AB7 comprising SEQ ID NO:2 or a variant sequence thereof having at least 70% identity to SEQ ID NO:2.

7. The vector construct of claim 1, wherein the pVIII phage coat protein corresponds to the pVIII coat protein from M13, fd or f1 phage or a variant thereof.

8. The vector construct of claim 1, wherein the pVIII phage coat protein comprises SEQ ID NO:8, or a variant sequence thereof having at least 70% identity to SEQ ID NO:8.

9. The vector construct of claim 8, wherein the variant sequence has a valine to isoleucine mutation at position 33 of SEQ ID NO:8 or a corresponding position.

10. The vector construct of claim 9, wherein the nucleic acid sequence encoding said isoleucine residue comprises the codon ATA.

11. The vector construct of claim 9, wherein the pVIII phage coat protein comprises SEQ ID NO:61, or a variant sequence thereof having at least 70% identity to SEQ ID NO:61, or is encoded by a nucleic acid sequence comprising SEQ ID NO:60 or a variant sequence thereof having at least 70% identity to SEQ ID NO:60.

12. The vector construct of claim 1, wherein the pVIII phage coat protein comprises SEQ ID NO:25, or a variant sequence thereof having at least 70% identity to SEQ ID NO:25.

13. The vector construct of claim 1, wherein the vector further comprises a polynucleotide sequence encoding a linker between the fluorophore and the pVIII phage coat protein.

14. The vector construct of claim 13, wherein the linker comprises SEQ ID NO:6 or a variant sequence thereof having at least 70% identity to SEQ ID NO:6.

15. The vector construct of claim 1, wherein the vector further comprises a polynucleotide sequence encoding a detectable tag.

16. The vector construct of claim 15, wherein the tag is located between the signal peptide and the fluorophore.

17. The vector construct of claim 15, wherein the tag is a FLAG tag or a derivative thereof, or another negatively charged tag.

18. The vector construct of claim 1, wherein the vector further comprises one or more ribosome binding sites (RBS).

19. The vector construct of claim 18, wherein the RBS driving translation of the fluorophore-pVIII fusion protein is a weak RBS.

20. The vector construct of claim 1, wherein the vector construct encodes a polypeptide sequence comprising SEQ ID NO:16: or a variant sequence thereof having at least 70% identity to SEQ ID NO:16.

21. The vector construct of claim 1, wherein the vector construct encodes a polypeptide sequence comprising: (i) SEQ ID NO:18 or a variant sequence thereof having at least 70% identity to SEQ ID NO:18; (ii) SEQ ID NO:20 or a variant sequence thereof having at least 70% identity to SEQ ID NO:20; or (iii) SEQ ID NO:22 or a variant sequence thereof having at least 70% identity to SEQ ID NO:22.

22. The vector construct of claim 21, wherein the pVIII phage coat protein comprises a variant sequence having a valine to isoleucine mutation at the position corresponding to position 33 of SEQ ID NO:8.

23. The vector construct of claim 22, wherein the nucleic acid sequence encoding said isoleucine residue comprises the codon ATA.

24. The vector construct of claim 1, wherein the vector construct further comprises a polynucleotide sequence encoding a protein of interest fused to a polynucleotide sequence encoding a non-pVIII phage coat protein.

25. The vector construct of claim 1, wherein the vector is a phagemid or a phage vector.

26. A nucleic acid molecule comprising a sequence encoding the pVIII phage coat protein variant sequence as defined in claim 9.

27. Phage particles comprising the vector construct of claim 1 and expressing a fluorophore-pVIII fusion protein on the surface.

28. The phage particles of claim 27, wherein said phage particles are particles of a filamentous phage.

29. A library of phage particles, wherein the phage particles are as defined in claim 27, and wherein multiple different proteins of interest are expressed on the surface of the phage particles.

30. A phage display system comprising a vector construct as defined in claim 1.

31. The phage display system of claim 30, further comprising an E. coli host cell which over expresses the proteins Tat A, Tat B and Tat C.

32. A method for producing fluorescent phage particles comprising: introducing the vector construct as defined in claim 1 into a bacterial host cell.

33. The method according to claim 32, wherein the bacterial host cell is an E. coli host cell.

34. The method according to claim 33, wherein the E. coli host cell over expresses the proteins Tat A, Tat B and Tat C.

Description

(1) The invention will be further described with reference to the following non-limiting Examples with reference to the following drawings in which:

(2) FIG. 1A shows a schematic overview illustrating some of the components of the fluorophore-pVIII phagemid constructs described herein. SD=Shine Dalgarno sequence (which can be weak or strong); Sig Sequence=Signal Sequence (shown here as TorAB7 but can be any signal sequence that directs the protein to the TAT secretory pathway); FLAG=FLAG tag; pVIII=phage pVIII coat protein. The boxes show variants of the various components in the SD, Sig Seq and Fluorophore parts of the vector which were made and tested. All variants were identical in sequence apart from the indicated variations. FIG. 1B shows a cartoon of a fluorophore-pVIII display Fluorophage particle. The phage particles will have multiple copies of the fluorophore randomly distributed along the length of the phage particle.

(3) FIG. 2A shows a schematic overview illustrating some of the components of a dual display phagemid construct described herein. Lac=lac promoter; SD=Shine Dalgarno sequence; POI=protein of interest; pIX=phage pIX coat protein; Sig Sequence=Signal Sequence; FLAG=FLAG tag; pVIII=phage pVIII coat protein. The boxes show the variants of the various components in the SD, Sig Seq and Fluorophore parts of the vector which were made and tested. All variants were identical in sequence apart from the indicated variations. FIG. 2B shows a cartoon of a dual display Fluorophage particle. The phage particles will have one or multiple copies of the POI displayed on pIX depending on helper phage, and multiple copies of the fluorophore randomly distributed along the length of the phage particle.

(4) FIG. 3 shows E. coli staining with phagemid constructs containing the fluorophores mNeonGreen, mCherry and mGFPmut2 fused to pVIII. Transformed E. coli cells were grown overnight at 30° C. to allow for expression of the fluorophore. Cell cultures were added to glass slides and fluorescence was visualized using a confocal microscope. mNeongreen and eGFPmut2 show clear and bright fluorescence in most cells. mCherry shows weak but detectable fluorescence.

(5) FIG. 4 shows an assessment of Fluorophage in phagemid rescue. A) Six variants of the Fluorophage phagemids containing weak or strong SD sequences, Tor AB7 or gp8 signal sequences, and mNeongreen or mGFPmut2 fluorophores, were propagated in 100 ml cultures of E. coli TOP10F′ followed by rescue by either M13K07 or Deltaphage helper phages. Phages were concentrated by PEG precipitation followed by infectious titration, and the results given as cfu.sup.ampR/ml. B) Phagemid to helper phage ratios were determined by cfu.sup.ampR/cfu.sup.kanR. C) Rescued Fluorophage particles were normalized to 1×10.sup.12 cfu.sup.ampR and tested for fluorescence in a Victor.sup.3 multilabel plate reader using a FITC filter. Fluorescence intensity is given as arbitrary units.

(6) FIG. 5 shows an assessment of the effect on fluorescence intensity by use of different E. coli strains. A) Four different E. coli strains were transduced with the Fluorophage with or without a FLAG-tag and grown overnight and normalized on OD.sub.600nm. Bacterial fluorescence was measured with the varioskan multimode plate reader and fluorescence intensity is given as arbitrary units. B) Normalized Fluorophage samples (Fluorophage particles) rescued using DeltaPhage from four different strains were analyzed for fluorescent intensity in the varioskan multimode plate reader. Fluorescence intensity is given as arbitrary units.

(7) FIG. 6 shows evaluation of Fluorophage performance in ELISA and FLISA. A) Normalized phage samples were analyzed for fluorescence intensity on the Victor.sup.3 plate reader and intensity is given as arbitrary units. B) A dilution series of normalized phage samples were added to wells either coated with 1 μg/ml phOx-BSA or milk block. Captured phages were detected with an anti-M13 HRP mAb and developed with soluble TMB. C) The same dilution series were analyzed for fluorescent signal in the Victor.sup.3 plate reader in parallel.

(8) FIG. 7 shows fluorophage staining of BW TCR 4B2A1 cells by flow cytometry. A normalized phage input of 5×10.sup.11 cfu/ml to 2×10.sup.5 cells was used throughout. A) Staining of transfected BW 4B2A1 cells. Bound phages were detected with PE-conjugated anti-FD IgY. B) Staining of untransfected BW cells. Bound phages were detected with PE-conjugated anti-FD IgY. C) Staining of BW 4B2A1 cells. Bound phages were detected by intrinsic phage fluorescence. D) Normalized phage samples were analyzed for fluorescence intensity on the Victor.sup.3 plate reader and intensity is given as arbitrary units.

(9) FIG. 8 shows an overview of the sorting/gating strategy used with the mutant library. Round 1 was a combination of cell based sorting and phage based enrichment of clones with high display level. Round 2 focused on clones with mNeonGreen signal and high levels of surface FLAG (double positive cells). Single clones were analysed using Sanger sequencing, Flow cytometry and fluorescence measurement.

(10) FIG. 9 shows the sequence of FLAG_mNG_pVIII vector (which is a preferred embodiment of the invention); the nucleotide sequence is provided as SEQ ID NO:54 and the amino acid sequence as SEQ ID NO:55. The TorAB7, FLAG, mNG (mNeonGreen), linker and pVIII components of this vector are all shown.

(11) FIG. 10 shows the comparison of the original Fluorophage (mNGwt) with the lead candidate from the screening (mNGF03). (A) XL1-blue cells containing the different phagemids were cultured for 16 hours at 37 deg C. in the absence of glucose to allow expression of the fusion protein. Cells were washed 3× times in PBS. The optical density of the cells was measured, and all samples were adjusted based on the optical density, to an A600 value of 1. A serial dilution with 1:2 steps was prepared. (B) Phages were packaged following the standard protocol using Deltaphage. Sample concentrations were determined by infectious titration and all samples were normalized to 5.0e+12 CFU.sup.AMP/mL. Fluorescence measurement was done in the Varioskan LUX at 488ex/517em.

(12) FIG. 11: Assessment of pVIII-mNGF03 clone in the context of the phOx-pIX fusion system. (A) Cells were taken during the packaging process. The fluorescence from the starting cultures (containing glucose) was measured to establish the baseline for the fluorescence measurement (Glu). The fluorescence of the same cells was measured after overnight packaging (absence of glucose). All samples were normalized based on the optical density, to an A600 value of 1. A serial dilution in 1:2 steps was prepared. (B) the phage concentration was determined based on the DNA amount present in the sample. Raw data was adjusted to the lowest value, 4,18E+12.

(13) FIG. 12 shows the Fluorophage staining of SKW-3 R12-C9 cells by flow cytometry. A normalized phage input of 5×10.sup.12 cfu/ml to 5×10.sup.4 cells was used throughout. A) Staining of transfected SKW-3 R12-C9 cells. Bound phages were detected with APC-conjugated anti-FD IgY. B) Staining of untransfected SKW-3 cells. Bound phages were detected with APC-conjugated anti-FD IgY. C) Staining of SKW-3 R12-C9 cells. Bound phages were detected by intrinsic phage fluorescence. D) Normalized phage samples were analyzed for fluorescence intensity on the Varioskan LUX plate reader and intensity is given as arbitrary units.

(14) FIG. 13 shows the evaluation of Fluorophage performance in ELISA and FLISA. A,B) A dilution series of normalized phage samples (the different titers used in pNGF03, TAT high are indicated on the x-axis) were added to wells either coated with 1 μg/ml phOx-BSA or milk block. Captured phages were analyzed for fluorescent signal in the Varioskan LUX plate reader. C) The same samples were detected with an anti-M13 HRP mAb and developed with soluble TMB. (D) Normalized phage (5.0E+12 CFU.sup.AMP/mL) samples were analyzed for fluorescence intensity on the Varioskan LUX plate reader and intensity is given as arbitrary units.

(15) FIG. 14 shows the comparison of bacterial fluorescence level. XL1-blue cells containing expression vectors for the TAT AC and B genes in addition to the mNGF03 phagemid were cultured for 16 hours at 37 deg C. in the absence of glucose to allow expression of the fusion protein. Cells without phagemids were used as negative controls. The cells were washed 3× times in PBS. The optical density of the cells was measured, and all samples were adjusted based on the optical density, to an A600 value of 1. A serial dilution with 1:2 steps was prepared. Fluorescence measurement was done in the Varioskan LUX at 488ex/517em.

EXAMPLE 1—EFFICIENT M13 DISPLAY OF A FUNCTIONAL BIOLOGICAL FLUOROPHORE ON THE MAJOR CAPSID PROTEIN PVIII BY USING PERIPLASMIC TARGETING

(16) Materials and Methods

(17) Construction of the Dual Display Phagemid

(18) All fluorophore-pVIII constructs were assembled in silico and ordered from Genscript (China). The fragments were inserted in the NheI cloning site in the pGALD9ΔLFN phagemid (Løset, G. Å., et al., 2011, PLoS ONE, 6, e17433, details of the vector also in Genbank HQ528250) carrying the scFv GB113 (Nilssen, et al., 2012, Nucleic acids research, 40, e120). Cloning procedure was confirmed by sequencing (GATC, Germany).

(19) A gene fragment containing the FLAG-tag and N-terminal portion of mNeongreen was ordered from Genscript and inserted using the SnabI and BsgI restriction sites in order to create FLAG-mNeongreen. The general structure of this phagemid is shown in the schematic in FIG. 2A. Part of this vector sequence is shown in FIG. 9. Phagemids encoding the mCherry and the mGFPmut 2 fluorophores were made in an analogous way.

(20) Phage Production

(21) Five ml of 2×YT medium containing 100 μg/ml Ampicilline, 20 μg/ml tetracycline and 0.1 glucose (YT-TAG) were inoculated with cells from a glycerol stock of E. coli XL1-Blue containing the respective phagemid and grown overnight at 37° C. on an orbital shaker. 200 ml of YT-TAG were inoculated with the pre-culture at an OD.sub.600 of 0.025 and grown at 37° C. with shaking. Cultures were infected with M0110 of helper phage DeltaPhage (Nilssen et al., 2012, supra), which is a helper phage with a conditional knockdown of its cognate pIX, allowing for high valence display of the pIX-POI, at an OD.sub.600 of 0.3-0.4. After 60 min incubation the cultures were pelleted and resuspended in 2×YT supplemented with 100 μg/ml ampicillin and 50 μg/ml Kanamycin, and shaking was continued at 28° C. for 16 hrs. Bacterial cells were removed by centrifugation and filtering through 0.2 μm vacuum driven filters (Millipore), and mixed 1:5 with polyethylene glycol (PEG)/NaCl solution (20% PEG 8000, 2.5 M NaCl). After overnight incubation on ice, the sample was centrifuged (5000×g, 45 min, 4° C.) and the pellets were dissolved in 25 ml Phosphate-buffered saline (PBS). The samples were mixed 1:5 with PEG/NaCl and incubated for 4 h on ice followed by centrifugation as before, and pellets were resuspended in 1 ml PBS. Virion titers were assessed by infectious spot titration as described in Koch, J., et al., 2000, Biotechniques, 29, 1196-1198, 2002.

(22) Fluorescence Measurements

(23) Excitation/emission spectra were determined using an FP-8500 Spectrofluorometer (Jasco). In balancing the cross-talk versus sensitivity of the instrument we set the separation of excitation and emission to 15 nm while allowing the emission filter to accept ±2.5 nm of variation. The measurements were done with 5 nm intervals for both excitation and emission.

(24) Fluorescence Measurement of Bacteria

(25) Single colonies of E. coli harboring Fluorophage mNeongreen or Fluorophage mNeongreen-FLAG phagemids were inoculated in 5 ml 2×YT supplemented with 100 μg/ml Ampicillin and incubated over night at 37° C. The cells were pelleted, and resuspended in 1×PBS followed by normalization based on OD.sub.600nm. The samples were prepared in 1:2 dilution series, and fluorescence was measured with the Varioskan multimode plate reader (Thermo Fischer) with appropriate ex/em setting for mNeongreen within the limitations of the instrument (500/525).

(26) Confocal Microscopy of Bacteria on Glass Slides

(27) Single colonies of E. coli harboring all three variants of mNeongreen, mGFPmut2 and mCherry Fluorophage phagemids were inoculated in 5 ml 2×YT supplemented with 100 μg/ml Ampicillin and incubated over night at 37° C. 3 μl of bacterial cultures were pipetted onto glass slides and glass cover slips were placed on top. The bacteria were visualized using a FV1000 Confocal Laser Scanning Microscope (Olympus) with appropriate available laser wave lengths for each fluorophore (488 nm: mNeongreen and mGFPmut2, 543 nm: mCherry).

(28) Fluorescence Measurement of Phage

(29) Fluorescence intensity measurements of phage were done by normalizing phage samples (diluted in 1×PBS) and measuring fluorescence with either the Victor.sup.3 multilabel reader (PerkinElmer) with FITC-filter (488/510) at 1 s. excitation, or with the Varioskan multimode plate reader (Thermo Fischer) with appropriate ex/em setting for mNeongreen within the limitations of the instrument (500/525).

(30) Phage Capture ELISA

(31) Microtiter plates were coated with 1 μg/ml phOx-BSA over night at 4° C. and blocked with PBS with 0.1% Tween (X) and 4% non-fat skimmed milk powder (PBSTM). 1:4 serial dilutions of phage samples starting at 2.5×10.sup.12 cfu/ml diluted in PBSTM were added in triplicate. Bound phage were detected by either anti-m13 HRP (Amersham Biosciences) followed by development with TMB solution and absorbance reading at 610 nm, or by fluorescence measurement in the Victor.sup.3 multilabel reader.

(32) Flow Cytometry

(33) Aliquots of 2×10.sup.5 BW 4B2A1 TCR transfectant and untransfected BW cells (negative control) were distributed into a V-shaped 96-well dish (NUNC). The total volumes were adjusted to 250 μl/well with 5% w/v FCS/PBS (pH 7.4). The plate was centrifuged at 300 g/5 min at RT and the supernatants discarded. Aliquots of 50 μl/well of phages with normalized titers of 5×10.sup.11 cfu.sup.ampR/ml pre-blocked in 5% FCS/PBS were added (Control wells received 5% FCS/PBS only) and the plate incubated for 1 h at 4° C. The cells were washed by adding 200 μl/well with 5% FCS/PBS, the cells pelleted by centrifuged at 300 g/5 min/RT and the supernatants discarded. PE-conjugated chicken Anti-fd (Norwegian Antibodies) in 5% FCS/PBS was added to the appropriate wells, and unstained samples received 5% FCS/PBS, followed by a 30 min incubation at 4° C. The wells were washed as above and fixed with 200 μl/well 2% PFA and kept in the dark until analysis on a FACScalibur (BD Biosciences). Data analysis was done using the FlowJo (v10.2) software.

(34) Results and Discussion

(35) Construction

(36) A number of phagemid variants were designed based on the pGALD9ΔLFN (Løset, G. Å., et al., 2011, supra) phagemid vector (see FIGS. 1 and 2). In the case of the dual display phagemids as depicted in FIG. 2, such phagemids encode a protein of interest (POI) (herein an antibody fragment, scFv of the antibody GB113) fused to pIX. This system was chosen as it has previously been found that display on pIX performs highly favorably in e.g. antibody selection (Hoydahl, et al., 2016, Scientific reports, 6, 39066; Loset, G. A., et al., 2011, PLoS One, 6, e14702). The phagemid also encodes the periplamic chaperone FkpA for enhanced folding efficiency (Gunnarsen, et al., 2010, BMC Biotechnol, 10, 8). In order to display a targeting protein and a fluorophore on two different coat proteins simultaneously, the pVIII fusion was inserted directly after the termination of pIX in the phagemid. This allows the new ORF to be regulated by the same promoter (lac), thus allowing production of dual display phage particles with fusions to both pIX and pVIII (FIGS. 2A and 2B).

(37) Some variations were included in the constructs (as shown in FIG. 2A). To investigate the effect of variations in translation intensity, we included either a weak SD (comprising the sequence AGGAGA, with the upstream region from the ATG start site of the Tat signal peptide having the sequence AAGGAGACAGTCATA) or a strong SD (the T7g10 TIR was used, which also includes an Epsilon sequence, TTAACTTTA, with the upstream region from the ATG start site of the Tat signal peptide having the sequence TTAACTTTAAGAAGGAGATATACAT). A stronger translation could force a greater amount of pVIII-fusions into the growing particle. To investigate the importance of translocation pathways, we introduced either gp8ss or TorAB7 as periplasmic targeting leader peptides. The leader peptide gp8ss provides targeting of the fusion to the Sec pathway of the secretory system, through which the protein is translocated in an unfolded state and allows folding to occur in the oxidizing environment of the periplasm (Manting, E. H. and Driessen, A. J., 2000, Mol Microbiol, 37, 226-238). The TorAB7 leader peptide has been engineered for improved periplasmic targeting of GFP (DeLisa et al., 2002, supra), and targets the fusion to the Twin Arginine Transport (TAT) pathway where the protein is folded in the reducing conditions of the cytoplasm prior to translocation. The fully folded protein is then translocated to the periplasm (Berks, B. C., et al., 2005, Curr Opin Microbiol, 8, 174-181).

(38) To investigate how various fluorescent proteins perform as fusions to pVIII, three different fluorescent proteins were assessed: GFPmut2 (Cormack, B. P., et al., 1996, Gene, 173, 33-38.) which is a FACS-optimized version of GFP that was selected for a red-shift for optimal detection in FACS selection using a standard FITC filter from an E. coli library. The mutant GFP also proved to be an excellent folder when expressed in E. coli. Another green fluorescent protein, mNeongreen (Shaner, N.C., et al., 2013, Nat Methods, 10, 407-409) was chosen. We also chose to include mCherry (Shaner, N.C., et al., 2004, Nat Biotechnol, 22, 1567-1572.) based on the results of a previous study done with display of biological fluorophores (Speck et al., 2011, supra), where it was shown to be functionally displayed regardless of translocation route. Unless otherwise specified, the pIX-fusion was scFv GB113, an antibody fragment that exclusively reacts with the murine T cell receptor (TCR) 4B2A1 (Bogen, B., et al., 1990, Eur J Immunol, 20, 2359-2362.), and has previously been shown to perform well in flow cytometry when displayed on phage (Nilssen, N. R., et al., 2012, Nucleic acids research, 40, e120).

(39) Phage Production and Validation

(40) In order to confirm that the pVIII-fluorophore fusion did indeed produce fluorescent proteins, we inspected cultures of transformed E. coli cells using fluorescence microscopy. In preparation for this, we discovered that bacteria transformed with a phagemid harboring the combination of strong SD and SEC-targeted fluorescent protein did not grow in culture when the glucose repressor was removed, and thus this phagemid version was omitted from further experiments (data not shown).

(41) Phagemid variants with the fluorophores mNeongreen, mCherry and mGFPmut2 (FIG. 2A) were transformed into E. coli which were then grown overnight at 30° C. to allow for expression of the fluorophore. These phagemid variants all contained a weak SD sequence and a TorAB7 signal sequence. Cell cultures were added to glass slides and fluorescence was visualised using a confocal microscope. The results are shown in FIG. 3 where it can be seen that mNeongreen and mGFPmut2 show clear and bright fluorescence and mCherry shows weak fluorescence. The presence of detectable levels of fluorescence with all three fluorophores demonstrates that the fluorophores are expressed and correctly folded. This fluorophore expression was confirmed by western blotting (data not shown). Thus, the above-described phagemid system is working to display functional fluorophore-pVIII fusion proteins.

(42) As mNeongreen and mGFPmut2 showed better detectable fluorescence than mCherry further experiments were carried out with these constructs.

(43) To evaluate whether the addition of a second coat protein fusion to the system would affect phage production and/or phagemid packaging, we produced all Fluorophage variants carrying the Green fluorescent proteins (mGFPmut2 and mNeongreen), and measured phage production and phagemid packaging by infectious titration (FIG. 4A). All phage samples reached end titers comparable to those previously observed in our pIX system (Loset et al., 2011, e17433, supra, and Nilssen et al., 2012, supra). Furthermore, as previously observed, phagemid rescue with DeltaPhage, a helper phage with a conditional knockdown of its cognate pIX, allowing for high valence display of the pIX-POI, did not appear to affect phage production (Nilssen et al., 2012, supra). The combination of weak SD and SEC-targeting caused preferential packaging of helper phage genome for mNeongreen, and the same effect was observed for strong SD and TAT targeting for mGFPmut2 (FIG. 4B). However, these results show that the dual display phagemids containing two coat protein fusions are packaged and allow phage production at equivalent levels to that seen when only a pIX fusion is present.

(44) In order to assess the levels of functional fluorescent proteins presented on the Fluorophage particles, normalized Fluorophage samples were analyzed in the victor.sup.3 multilabel reader using a FITC filter (488 nm/510 nm) (FIG. 4C). The combination of weak SD and TAT-targeted mNeongreen stood out as it exceeded background fluorescence by at least three-fold. However, the Fluorophage with a strong SD and TAT targeting also showed approximately two-fold higher intensity than background, while the intensity of, for example, the mGFPmut2 Fluorophage with a weak SD and TAT targeting was slightly above the background signal. Variations in display valence of the pIX POI did not appear to affect functional display levels of the fluorescent protein on pVIII. Reaching high end titers is highly important for downstream applications such as library selections, and the Fluorophages were not produced in lower numbers than previously observed in the pIX system. The Fluorophage construct did indeed produce fluorescent phage particles, and one of the variations, namely TAT-targeted mNeongreen with a weak SD reached a satisfactory signal to noise ratio. Worth noting here is that mNeongreen possesses a sharp excitation peak at 506 nm and a likewise sharp emission peak at 517 nm which does not match the fluorescein filter used in the Victor.sup.3 (485 nm/535 nm). The intensity measured using this filter is likely to be highly underestimated. Regardless, the combination of weak SD and TAT targeted mNeongreen stood out as the best performing construct and was thus chosen for downstream characterization.

(45) In order to investigate and verify the excitation and emission spectra of mNeongreen when displayed on pVIII, 2×10.sup.11 phage particles/ml were analyzed on a Spectrofluorometer (data not shown). As a positive control we included 2 μg/ml of soluble eGFP which showed a clear and defined peak at the expected wavelengths (490 nm/510 nm). As a negative control we analyzed phage particles displaying the scFv GB113, but no fluorophore. Here we observed a sharp peak at 440 nm/455 nm. When analyzing the phage samples with display of mNeongreen, the same sharp peak appeared in addition to the peak that was expected from correctly folded mNeongreen at (505 nm/520 nm) (Shaner et al., 2013, supra).

(46) Optimization of Functional Display

(47) We wanted to investigate whether there were any differences in expression and folding in the standard lab strains of E. coli and whether any differences would translate into functional display levels. In addition, we wanted to include a detection tag for fluorescent protein expression, and added a FLAG tag to the N-terminus of mNeongreen. Four E. coli strains, XL1-Blue, AVB100F′, SS320 and Top10F′ were transformed with the Fluorophage±FLAG.

(48) Fluorescence intensity was then measured directly in live E. coli cells (FIG. 5A). There was a clear difference in fluorescence between strains. Proteins expressed in XL1-blue gave the highest intensity, closely followed by AVB100F′. These two strains showed markedly higher fluorescence intensity than SS320 and TOP10F′, although functional expression and folding was still clearly observed in these strains. To our surprise, we observed that the addition of the FLAG appeared to translate into higher overall fluorescence intensities. These FLAG containing constructs were then used to measure the fluorescence intensity from phage samples produced in the four individual strains (FIG. 5B), and again comparing the fluorescence of mNeongreen-FLAG from Fluorophage packaged in the four different strains (FIG. 5B), it became clear that XL1-blue was the superior strain, followed by AVB100F′. Phages packaged in SS320 and Top10F′ gave poorer fluorescence intensities.

(49) Even though they are closely related, the different strains showed quite large differences in functional expression and folding of the fluorophores. Moreover, although XL1-Blue and AVB100 were very similar in bacterial fluorescence when producing the pVIII-fluorophore fusion alone, Fluorophage particles produced in XL1-blue gave almost two-fold higher fluorescence intensity. These data suggest that the favourable ability to fold the fluorophore functionally in the cytosol is shared between XL1-Blue and AVB100, which contrast SS320 and TOP10F′. However, XL1-Blue appears to have a superior ability also compared with AVB100 to transport the folded fluorophore from cytosol to the periplasm and integrate it into the virion. Thus, although any of these strains (or indeed other E coli strains) can be used, XL1-blue is preferred in some cases.

(50) Enzyme-Linked Immunosorbent Assay (ELISA) and Fluorescent-Linked Immunosorbent Assay (FLISA)

(51) To assess whether or not we could combine the ability to display both a functional fluorophore-pVIII fusion and a different functional fusion on a different capsid than pVIII on the same phage particle, we chose to insert a well-described anti-phOx scFv (Marks et al., 1992, Biotechnology, 10:779-783) into the fluorophore-pVIII expressing phagemid as a pIX fusion. We then prepared phages and tested them for phox binding in classical ELISA as well as FLISA as described. For comparison, we prepared anti-phox scFv-pIX displaying phages without the fluorophore-pVIII fusion and tested in parallel.

(52) The fluorescence intensity of normalized amounts of phage particles was measured in the Victor.sup.3 plate reader (FIG. 6A) and showed very high fluorescence intensity, and only from the phages displaying the fluorophore-pVIII fusion. Importantly, when testing the same phage samples for scFv target binding in standard ELISA, specific and identical binding was observed independent of presence or absence of the fluorophore-pVIII fusion (FIG. 6B). Thus, the data demonstrate that display of a fluorophore on pVIII did not affect antigen binding mediated by the targeting moiety in the pIX fusion on dual display phage. When the same phage samples again were tested for target binding using FLISA employing the Victor.sup.3 plate reader, we could not observe a fluorescent signal from any sample (FIG. 6C). Thus, the data show that measuring the fluorescence of all phages applied in the well is readily done, whereas in the current set-up we did not have the sensitivity to specifically detect the reduced amount of phages retained on target after washing in the FLISA. Importantly, the data irrevocably shows that dual display of the fluorophore-pVIII fusion and scFv-pIX fusion on the same phage particle does not affect the phenotype of the scFv-pIX fusion.

(53) Detection of Cell Surface Expressed Proteins by Flow Cytometry

(54) A powerful application of the Fluorophage would be to use it in real time selection on FACS without the need for staining antibodies, but only relying on the inherent fluorescence of the fluorophore-pVIII fusion for detection. To assess this, we tested the GB113 scFv-pIX displaying Fluorophage for specific binding to the T-cell receptor expressed on murine 4B2A1 T cell hybridoma cells using flow cytometry. For comparison, we prepared GB113 scFv-pIX displaying phages without the fluorophore-pVIII fusion and tested in parallel.

(55) In order to verify specific binding to the cells expressing the TCR 4B2A1 using standard methodology, we used a PE-conjugated anti-fd antibody to detect bound phages. This also allowed a side-by-side comparison of the Fluorophage and the regular phage bound to the TCR (FIG. 7A), and neither bound to the TCR negative murine BW58 cells (PMID: 2558022) (FIG. 7B). When attempting to stain cells without the addition of a detection mAb, we observed no increase in fluorescence intensity compared to the controls (FIG. 7C), despite the strong fluorescence signal observed from the Fluorophage scFv GB113 when in solution (FIG. 7D). Thus, even though the Fluorophage exhibited strong fluorescent signal in solution, we were unable to detect it in binding to surface bound antigen using flow cytometry. This means that although this Fluorophage is highly useful and advantageous in many assays, the Fluorophage in this form does not yet reach the brightness per particle needed to allow direct staining and FACS sorting of cells.

(56) Concluding Remarks

(57) Previous efforts that attempt to build on the combined strengths of phage and cellular technologies have been reported, but have yet to see regular use. Functional display of fluorescent proteins on phage particles has been achieved previously, but has so far only shown limited success and has not been achieved on the pVIII coat protein. At this point, in vitro indirect coupling of a fluorophore to sortase tagged pVIII has shown the greatest promise (Hess et al., 2012, supra), but this procedure is inefficient, laborious and expensive.

(58) We here show the generation of a phage particle with both intrinsic fluorescence capability and target binding activity. mNeongreen proved to be the superior fluorophore of those investigated for display on pVIII. We found that the optimal conditions for functional display of biological fluorophores on pVIII involved translocation of the fusion through the Tat pathway for cytosolic folding. We surprisingly also found that a weak SD gave higher functional display. The exact cause of this is unknown, but one might speculate that a form of overload of the E. coli folding machinery may occur. XL1-Blue was shown to give the highest fluorescence intensity to the produced phage particles.

(59) Potential end use of this reagent is not limited to selection and screening of antibodies. Any protein that can be displayed on a phage can be inserted into Fluorophage phagemid to produce a single layer detection reagent.

(60) Even though the Fluorophage exhibited strong fluorescence and will have advantageous utility in many assays due to its intrinsic fluorescence, thereby avoiding the use of labelled or staining reagents/antibodies, combined with its ability to bind to a target molecule via the second coat protein fusion to a POI, it was still not bright enough to be detected when bound to antigen. A possible way to achieve this might be to increase the brightness of the Fluorophage particles. This might theoretically be achieved by: 1) increasing the number of pVIII-mNeongreen per particle, or 2) improving functional display of mNeongreen by improving folding, or 3) increasing the intrinsic brightness of mNeongreen, or a combination of all.

(61) To this end, we have advantageously been able to develop variants of the Fluorophage constructs which show improved efficacy (improved fluorescence properties).

EXAMPLE 2—GENERATION OF FLUOROPHAGE VARIANTS

(62) Random mutation of amino acids in the sequence of the original (parent) mNeonGreen protein as described herein (SEQ ID NO:3 and 4) was carried out to assess whether it was possible to produce a version that is either brighter than the existing (parent) protein or shows an improved display as a pVIII-fusion in the phage coat. Ideally, we would identify a mutant that shows improvement in both categories. However, we were not sure that this would be possible.

(63) We constructed a library that consists of Fluorophage vector variants carrying mutations in the mNeonGreen gene, the linker and the first 90 nucleotides of the pVIII gene.

(64) The part of the construct which was subjected to mutagenesis is shown below (these three parts are found as consecutive sequences in the vector but have been shown separately here for clarity):

(65) TABLE-US-00012 (SEQ ID NO: 56) GTTTCTAAAGGTGAAGAAGACAACATGGCTTCTCTGCCGGCTACCCACGA ACTGCACATCTTCGGTTCTATCAACGGTGTTGACTTCGACATGGTTGGTC AGGGTACCGGTAACCCGAACGACGGTTACGAAGAACTGAACCTGAAATCT ACCAAAGGTGACCTGCAGTTCTCTCCGTGGATCTTAGTTCCGCACATCGG TTACGGTTTCCACCAGTACCTGCCGTACCCGGACGGTATGTCTCCGTTCC AGGCTGCTATGGTTGACGGTTCTGGTTACCAGGTTCACCGTACCATGCAG TTCGAAGACGGTGCTTCTCTGACCGTTAACTACCGTTACACCTACGAAGG TTCTCACATCAAAGGTGAAGCTCAGGTTAAAGGTACCGGTTTCCCGGCTG ACGGTCCGGTTATGACCAACTCTCTGACCGCTGCTGACTGGTGCCGTTCT AAAAAAACCTACCCGAACGACAAAACCATCATCTCTACCTTCAAATGGTC TTACACCACCGGTAACGGTAAACGTTACCGTTCTACCGCTCGTACCACCT ACACCTTCGCTAAACCGATGGCTGCTAACTACCTGAAAAACCAGCCGATG TACGTTTTCCGTAAAACCGAACTGAAACACTCTAAAACCGAACTGAACTT CAAAGAATGGCAGAAAGCTTTCACCGACGTTATGGGTATGGACGAACTGT ACAAA (mNeonGreen) GGCGGTGGCAGCGGCGGTGGCAGC (linker) GCTGAGGGTGACGATCCCGCAAAAGCGGCCTTTAACTCCCTGCAAGCCTC AGCGACCGAATATATCGGTTATGCGTGGGCGATGGTTGTT pVIII (partially, first 90 nucleotides),
Library Construction

(66) We carried out incorporation of dNTP analogues by PCR to introduce random mutations into the phagemid DNA. Based on a paper by Zaccolo et al. (1996, J Mol Biol 255(4):589-603), Jena Biosciences produces and sells a PCR-based mutagenesis kit that was used during library generation. As a cloning tool we used the NEBuilder® HiFi DNA Assembly Cloning Kit. This kit will ligate any fragments with sequence overlaps between 15-80 nts.

(67) Mutagenesis:

(68) We created and characterized libraries of one fragment with different mutational loads based on the number of rounds/cycles used to incorporate dNTP analogues during PCR. We choose a fragment that contains the complete fluorophage pVIII-fusion gene (including all 5 functional domains, i.e. TorAB7, FLAG, mNG, linker and full length pVIII) and its flanking regions (see sequence below). Using this approach, we have the possibility to go back and amplify different regions to construct new libraries in the future without having to go through the mutagenesis process all over again:

(69) The individual parts of this construct are shown below (these parts are found as consecutive sequences in the vector but have been shown separately here for clarity):

(70) TABLE-US-00013 (SEQ ID NO: 57) TCTTTCGTTTTAGGTTGGTGCCTTCG TAGTGGCATTACGTATTTTACCCGTTTAATGGAAACTTCCTCATGATAAG CTAGCAAGCAAGGAGACAGTCATA (flanking region) ATGAACAATAACGATCTCTTTCAGACATCACGTCAGCGTTTTTTGGCACA ACTCGGCGGCTTAACCGTCGCCGGGATGCTGGGGCCGTCATTGTTAACGC CGCGACGTGCGACT (Tor AB7) GCGGCG (amino acids AA) Gattacaaggatgacgatgacaag (FLAG) GGC (amino acid G) GTTTCTAAAGGTGAAGAAGACAACATGGCTTCTCTGCCGGCTACCCACGA ACTGCACATCTTCGGTTCTATCAACGGTGTTGACTTCGACATGGTTGGTC AGGGTACCGGTAACCCGAACGACGGTTACGAAGAACTGAACCTGAAATCT ACCAAAGGTGACCTGCAGTTCTCTCCGTGGATCTTAGTTCCGCACATCGG TTACGGTTTCCACCAGTACCTGCCGTACCCGGACGGTATGTCTCCGTTCC AGGCTGCTATGGTTGACGGTTCTGGTTACCAGGTTCACCGTACCATGCAG TTCGAAGACGGTGCTTCTCTGACCGTTAACTACCGTTACACCTACGAAGG TTCTCACATCAAAGGTGAAGCTCAGGTTAAAGGTACCGGTTTCCCGGCTG ACGGTCCGGTTATGACCAACTCTCTGACCGCTGCTGACTGGTGCCGTTCT AAAAAAACCTACCCGAACGACAAAACCATCATCTCTACCTTCAAATGGTC TTACACCACCGGTAACGGTAAACGTTACCGTTCTACCGCTCGTACCACCT ACACCTTCGCTAAACCGATGGCTGCTAACTACCTGAAAAACCAGCCGATG TACGTTTTCCGTAAAACCGAACTGAAACACTCTAAAACCGAACTGAACTT CAAAGAATGGCAGAAAGCTTTCACCGACGTTATGGGTATGGACGAACTGT ACAAA (mNeonGreen) GGCGGTGGCAGCGGCGGTGGCAGC (Linker) GCTGAGGGTGACGATCCCGCAAAAGCGGCCTTTAACTCCCTGCAAGCCTC AGCGACCGAATATATCGGTTATGCGTGGGCGATGGTTGTTGTCATTGTCG GCGCAACTATCGGTATCAAGCTGTTTAAGAAATTCACCTCGAAAGCAAGC (pVIII full) TGATAAGCTAGCTTGAGGCATCAATAAAACGAAAGGCTCAGTCGAAAGAC TGGGCCTTTCATTTTATCTGTTGTTTGTCGGTTAACGCTTGTCGTCATCG TCCTTGTAGTCTTTTTTAGCAGAATCTGCGGCTTTCGCATC  (flanking region), 

(71) This sequence was produced with following primers:

(72) TABLE-US-00014 (SEQ ID NO: 58) pVIII_fwd: AGTGTTTTAGTGTATTCTTTCGCC (SEQ ID NO: 59) FkpA_R: GCTGATGCAAAGCCGG

(73) Sequences were analyzed using Sanger sequencing service by GATC (Germany). Up to 28 rounds (PCR cycles) of mutation were carried out which resulted in around 25% mutation frequency at the amino acid level (around 10% at the nucleotide level), data not shown.

(74) The mutagenized fragments were cloned back into the vector as shown in FIG. 2 comprising a TorAB7 signal peptide and a FLAG tag. The sequence of the TorAB7, FLAG, mNG, Linker and pVIII components of this vector are as shown in FIG. 9.

(75) Cloning of the Library:

(76) The cloning process is straight forward and we strictly followed the recommendations of the NEBuilder® HiFi DNA Assembly Cloning Kit.

(77) The complete ligation reaction was transformed into electro-competent SS320 to produce a library with high diversity. Titration based calculation of the library was 1.0E+06. We then prepared DNA from the SS320 cells and electroporated XL1-blue cells, the standard packaging strain for the Fluorophages.

(78) The library was packaged with Hyperphage following standard packaging protocols and the mutant library members were screened by FACS.

(79) Gating Strategy on the FACS ARIA IIu

(80) An overview of the complete sorting strategy is shown in FIG. 8. We gated the library based on two characteristics: mNeonGreen performance (green; Alexa488) and display level of fusion protein (red; APC). (For the second characteristic, the level of surface bound mNeonGreen-pVIII fusion protein can be quantified using an anti-DYKDDDDK-APC conjugated antibody, i.e. a fluorescently labelled anti-FLAG tag antibody). The gating strategy in the first round was liberal regarding the signal strength of both channels. The sorted cells were packaged to produce the corresponding phage library. An anti-FLAG®-M2 coated immunotube was then used to enrich clones with fusion proteins integrated into the phage coat. Fresh XL1-blue cells were transduced with the output of this enrichment and the resulting cells were prepared for sorting as in round 1. The gating in round 2 was much more stringent and intended to sort double positive clones only.

(81) It can be seen from the FACS profiles shown in FIG. 8 that library members with increased numbers of fluorophores on the surface and/or clones with increased mNeonGreen signal have been generated. See the shift of clones from the bottom left quadrant of the FACS panel on the left-hand side of FIG. 8 (which shows results before sorting), to the bottom right quadrant (high levels of surface FLAG, APC signal), the top left quadrant (increased mNeonGreen signal) or the top right quadrant (clones with both high levels of surface FLAG, APC signal, and increased mNeonGreen signal) seen in the Round 2 FACS profiles of FIG. 8 (i.e. after two rounds of sorting).

(82) Such results show that obtaining variant fluorophages which have significantly increased and improved fluorescence from that observed when the parent mNeonGreen sequence was used can be achieved, and single clones can now be further characterised in order to select highly advantageous mutants.

EXAMPLE 3—CHARACTERIZATION OF A MUTANT FLUOROPHORE

(83) The libraries prepared as described above can be screened for mutants showing improved fluorescence properties. One of the improved clones identified by these screenings is the so called F03 mutant as described herein. This F03 mutant was identified from one of the screenings which contained libraries with mutations in the full length pVIII. When sequenced the pVIII part of the F03 mutant had the nucleotide sequence:

(84) TABLE-US-00015 (SEQ ID NO: 60) GCTGAGGGTGACGATCCCGCAAAAGCGGCCTTTAACTCCCTGCAAGCCTC AGCGACCGAATATATCGGTTATGCGTGGGCGATGGTTGTTGTCATTATAG GCGCCACTATCGGTATCAAGCTGTTTAAGAAATTCACCTCGAAAGCAAGC

(85) The amino acid sequence was:

(86) TABLE-US-00016 (SEQ ID NO: 61) AEGDDP AKAAFNSLQASATEYIGYAW AMVVVIIGAT IGIKLFKKFT SKAS

(87) The F03 clone contained only a single amino acid mutation in the mNeonGreen-pVIII fusion protein as compared with the parent clone, which was a V to I mutation at position 33 of the pVIII protein. The corresponding codon was GTC in the parent clone, which was ATA in the F03 clone.

(88) XL1-blue cells containing the different phagemids (either the mNG-pVIII wild type/parent fusion protein in the context of the GB113 fusion system described above, i.e. mNGwt, or the F03 clone with the mNG-mutated pVIII sequence, mNGF03) were cultured for 16 hours at 37 deg C. in the absence of glucose to allow expression of the fusion protein. Cells were washed 3× times in PBS. The optical density of the cells was measured, and all samples were adjusted based on the optical density, to an A600 value of 1. A serial dilution with 1:2 steps was prepared.

(89) The results are shown in FIG. 10. Panel 10A shows the comparison of fluorescence intensity measured directly in live XL-1 blue cells expressing the original Fluorophage (mNGwt) with the lead candidate from the screening (mNGF03). Panel 10B shows the same comparison by monitoring the fluorescence intensity of phage particles which had been packaged in the strains following the standard protocol using Deltaphage as described above. Sample concentrations were determined by infectious titration and all samples were normalized to 5.0e+12 CFU.sup.AMP/mL. Fluorescence measurement was done in the Varioskan LUX at 488ex/517em.

(90) It can be seen that the F03 clone shows increased bacterial and phage particle fluorescence which is approximately 2 fold higher than the fluorescence seen with the parent clone.

(91) The mNGF03-pVIII fusion cassette was cloned into the phOx-pIX fusion system as described above to form a dual expression construct. Bacterial cells expressing both the phOx-mNGwt as described above and the phOx-mNGF03 were assessed for fluorescence intensity (FIG. 11A), as were phage particles (FIG. 11B).

(92) In FIG. 11A cells were taken during the packaging process. The fluorescence from the starting cultures (containing glucose) was measured to establish the baseline for the fluorescence measurement (Glu). The fluorescence of the same cells was measured after overnight packaging (absence of glucose). All samples were normalized based on the optical density, to an A600 value of 1. A serial dilution in 1:2 steps was prepared. In FIG. 11B the phage concentration was determined based on the DNA amount present in the sample. Raw data was adjusted to the lowest value, 4.18E+12.

(93) It can be seen from FIGS. 11A and B that in the phOx system the mNGwt produces good fluorescence as measured on both bacterial cells (FIG. 11A) and phage particles (FIG. 11B). FIGS. 11A and B show that the mNGF03 also produces fluorescence in the phOx system. FIG. 11B also shows that, once adjusted for titre, the phOx-mNGF03 phage show a better fluorescence signal per phage than the mNGwt.

(94) A further dual display vector was constructed in which the H57 anti-mouse TCR scFv was fused to pIX.

(95) Construction of the pGALD9FN

(96) The H57 anti-mouse TCR scFv antibody sequence, as described by Huppa et al, 2010 (Nature 463:963-7) fused to pIX was generated by gene synthesis (GS) and standard molecular cloning into the pGALD9ΔLFN phagemid on the compatible NcoI/NotI RE sites (Huppa et al. 2010, supra; Løset et al. 2011, PLosONE, 6: e17433 supra) was carried out. The GS, cloning procedure and confirmation by sequencing was performed by Eurofins.

(97) Construction of the H57-mNGF03

(98) The cassette containing the complete mNGF03-pVIII cassette was sub-cloned into the pGALD9ΔLFN-H57 phagemid using the flanking NheI. The correct orientation was confirmed by sequencing by Eurofins.

(99) Construction of Retroviral Vector for TCR Expression

(100) A gene containing the Vα and Vβ of the R12-C9 TCR (Straetemans, T., et al., 2012) grafted onto murine Cα and Cβ, respectively, with alpha and beta coupled by P2A cleavable peptide (Eurofins Genomics) were cloned into pMSCV-Neo (Clontech) with a re-organized MCS on EcoRI and XhoI restriction sites essentially as described (Hoist, J. et al. Generation of T-cell receptor retrogenic mice. Nat. Protocols 1, 406-417 (2006).

(101) Construction of SKW-3 R12-C9 Cell Line

(102) GP2-293 cells (Clontech) were co-transfected with pMSCV-Neo R12-09 and pEco (Clontech). Viral supernatant was transduced into SKW-3 cells (CLS Cell Line Services, Germany). A transduced pool of SKW-3 R12-C9 cells were co-stained with PE α-hCD3 mAb (BD Biosciences) and mCβ specific H57-A647 mAb (LifeTech) and positive transductants were isolated on a FACSAria Hu (BD).

(103) Flow Cytometry

(104) Aliquots of 5×10.sup.4 SKW-3 R12-C9 TCR transfectant and untransfected SKW-3 cells (negative control) were distributed into a V-shaped 96-well dish (Costar). The total volumes were adjusted to 250 μl/well with 0.5% w/v BSA/PBS (pH 7.4). The plate was centrifuged at 300 g/5 min at RT and the supernatants discarded. Aliquots of 25 μl/well of phages with normalized titers of 5×10′.sup.12 cfu.sup.ampR/ml pre-blocked in 0.5% BSA/PBS were added (Control wells received 0.5% BSA/PBS only) and the plate incubated for 30 min at RT. The cells were washed by adding 200 μl/well with 0.5% BSA/PBS, the cells pelleted by centrifuged at 300 g/5 min/RT and the supernatants discarded. APC-conjugated chicken Anti-fd (Norwegian Antibodies) in 0.5% BSA/PBS was added to the appropriate wells, followed by a 30 min incubation at 4° C. The wells were washed twice as above and immediately analysed on an Accuri C6 (BD Biosciences). Data analysis was done using the FlowJo (v10.5) software.

(105) Detection of Cell Surface Expressed Proteins by Flow Cytometry

(106) A powerful application of the Fluorophage would be to use it in real time selection on FACS without the need for staining antibodies, but only relying on the inherent fluorescence of the fluorophore-pVIII fusion for detection. To assess this, we tested the H57 scFv-pIX displaying Fluorophage for specific binding to the T-cell receptor expressed on human SKW-3 R12-C9 T cell hybridoma cells using flow cytometry. For comparison, we prepared H57 scFv-pIX displaying phages without the fluorophore-pVIII fusion and tested in parallel.

(107) In order to verify specific binding to the cells expressing the TCR R12-C9 using standard methodology, we used a APC-conjugated anti-fd antibody to detect bound phages. This also allowed a side-by-side comparison of the Fluorophage and the regular phage bound to the TCR (FIG. 12A), and neither bound to the TCR negative human SKW-3 (FIG. 12B). When attempting to stain cells without the addition of a detection mAb, we observed no increase in fluorescence intensity compared to the controls (FIG. 12C), despite the strong fluorescence signal observed from the Fluorophage scFv H57 when in solution (H57-mNGF03, FIG. 12D). Thus, even though the Fluorophage exhibited strong fluorescent signal in solution, we were unable to detect it in binding to surface bound antigen using flow cytometry. This means that although this Fluorophage is highly useful and advantageous in many assays, e.g. FLISA as described below which shares many similarities and components with a FACS assay, the Fluorophage in this form does not yet reach the brightness per particle needed to allow direct staining and FACS sorting of cells.

EXAMPLE 4—ENZYME-LINKED IMMUNOSORBENT ASSAY (ELISA) AND FLUORESCENT-LINKED IMMUNOSORBENT ASSAY (FLISA)

(108) Materials and Methods

(109) Construction of TAT-ABC Expression Vectors

(110) The vectors containing the Tat AC/B genes were designed using the Gensmart design tool (Genscript Ltd.). All constructs were produced and quality controlled by Genscript Ltd. The TatB sequence source is NC_000913.3: 4022218-4022733 (NCBI) and the TatAC sequence source is #X73888.1 (European Nucleotide Archive). For detection purposes TatB was fused to an C-terminal 6×HIS tag and TatC to an C-terminal FLAG tag.

(111) Construction of TaT-AC/B Over Expressing XL1-Blue Cells Containing the phOx-mNGF03 Phagemid

(112) Electrocompetent XL1-blue cells were transformed with 1 ng of each vector using the standard protocol for the ECM399 Electroporation System. XL1-blue cells were grown an LB-agar containing 30 μg/mL Chloramphenicol and 50 μg/mL Zeocin to select cells with both expression vectors. Expression of TatC and TatB were confirmed by standard Western Blot analysis using horse radish peroxidase conjugated anti-FLAG-M2 (1:5000 dilution) and anti-HIS (1:2000 dilution) mAB on lysates from double positive clones. 3 cell clones were chosen based on their expression levels. 10 mL from all three cell clones were transduced with phages carrying a phagemid containing the phOx-pIX and the mNeonGreen-F03-pVIII fusion. 10 μl culture were incubated on LB-agar containing 30 μg/mL Chloramphenicol, 50 μg/mL Zeocin, 500 mM glucose and 100 μg/mL Ampicillin. Resulting clones were tested based on their ability to fluoresce.

(113) In this Example, in some experiments XL-1 host cells were used which had been engineered to overexpress the Tat transporter as described above, as well as expressing the phOx-mNGF03 phagemid in order to see if fluorescence levels could be improved by overexpressing the Tat transporter.

(114) Some exemplary results are shown in FIG. 14, in which an overnight culture of bacterial cells in the absence of glucose shows a high increase in the fluorescence intensity of bacterial cells containing the phOx-mNGF03 phagemid when the Tat transporter is overexpressed (XL1-blue TAT; phOx-mNGF03) as compared to bacterial cells with no Tat transporter overexpression (XL1-blue; phOx-mNGF03).

(115) In addition, to assess whether or not we could combine the ability to display both a functional fluorophore-pVIII fusion and a different functional fusion on a different capsid than pVIII on the same phage particle, we chose to insert a well-described anti-phOx scFv (Marks et al., 1992, Biotechnology, 10:779-783) into the fluorophore-pVIII expressing phagemid as a pIX fusion (this anti-phOX phagemid was also used in some experiments above). We then prepared phages and tested them for phOx binding in classical ELISA as well as FLISA as described. For comparison, we prepared anti-phox scFv-pIX displaying phages without the fluorophore-pVIII fusion and tested in parallel.

(116) The fluorescence intensity of normalized amounts of phage particles was measured in the Varioskan LUX plate reader (FIG. 13D) and showed very high fluorescence intensity, and only from the phages displaying the fluorophore-pVIII fusion. Again when assessed on the phage particles the F03 clone shows improved fluorescence per phage particle when compared to the parent (wt) clone.

(117) Importantly, when testing the same phage samples for scFv target binding in standard ELISA, specific binding was observed independent of presence or absence of the fluorophore-pVIII fusion (FIG. 13C). Thus, the data demonstrate that display of a fluorophore on pVIII did not affect antigen binding mediated by the targeting moiety in the pIX fusion on dual display phage. When the same phage samples again were tested for target binding using FLISA employing the Varioskan LUX plate reader, we could observe fluorescent signal from phage particles that were displaying the mNGwt-pVIII (FIG. 13A, phOx-mNGwt) and the mNGF03-pVIII fusion protein (FIG. 13B), in particular when a high titre of the Tat overexpressing XL-1 bacterial host cells were used (TAT high). Thus, the data show that the intrinsic fluorescence of these phage particles is high enough to provide a sensitivity to specifically detect the reduced amount of phages retained on target after washing in the FLISA and that these phage particles find advantageous utility in such assays, for example enable detection when bound to antigen. Importantly, the data irrevocably shows that dual display of the fluorophore-pVIII fusion and scFv-pIX fusion on the same phage particle does not affect the phenotype of the scFv-pIX fusion.