Modified pix vector construct

Abstract

The present invention provides a vector construct comprising an open reading frame comprising a nucleic acid sequence encoding a modified pIX filamentous phage coat protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue. Phage particles or a phage display system comprising said vector constructs are also provided, together with methods of phage display and kits.

Description

[0001] Modified pIX Vector construct

[0002] The present invention relates to modified pIX filamentous phage coat proteins for phage display and phage particles displaying fusion proteins comprising such modified pIX proteins. More particularly, the present invention relates to vector constructs and nucleic acid molecules encoding said modified pIX phage coat proteins, preferably fused to proteins of interest, in order to display said proteins of interest on the surface of the phage as a pIX fusion protein.

[0003] Over the last three decades, phage display has become a powerful and efficient method for discovery and evolution of novel binding proteins. The principle of combinatorial phage display technology is based on the genotype - phenotype linkage offered by the property that each virion will only display on its surface the very same proteins that are encoded by the genome encapsulated by its protein coat. The phage particle itself is highly resistant to a variety of physiochemical conditions; hence phage display offers superior versatility in many selection regimes as compared to competing combinatorial technologies. Thus, although competing combinatorial technologies exist, none show the high degree of versatility combined with the ease of use. Nonetheless, it is still challenging to screen for the desired phenotype following phage panning, and it is not given that the optimal binders are identified.

[0004] Phage display of heterologous polypeptides has been achieved using all five structural proteins of the filamentous phage coat, but pill- has gained the most widespread use. pIX display has also been described and has been shown to have certain advantages over standard pill display in terms of efficacy in identifying desired target specific antibody candidates as well as yielding clones with superior biophysical properties (Hoydahl et al., 2016, Sci.Rep. 6, 39066).

[0005] The present inventors have now developed an improved pIX phage display system that utilizes an improved pIX expressing vector construct. The improvement is in the form of providing a vector encoding a modified pIX phage coat protein in which rather than the full-length wild type pIX protein being provided, the vector encodes a modified pIX phage coat protein in which the methionine (M) residue at position 1 is replaced by an alternative amino acid residue. Surprisingly, the inventors have shown that such modified pIX vectors can give rise to significant improvements in phage display in terms of functional protein display (e.g. in the form of improved antigen or target binding / reactivity or improved fusion protein display or functionality), or infectious phage particle production (infectious phage titre), or both, when compared to phage display using the wild-type pIX protein. Such improvements are highly advantageous for phage display either in terms of the ability to successfully select and isolate binders at all, the quality and / or quantity of binders that can be selected, or in terms of the size of the phage display library that can be produced. For example, the most effective use of phage display as an engineering and discovery tool requires both the highest possible functional display in combination with the highest possible infective virion / phage particle production. This will ensure that the largest heterologous fusion protein library pool can be screened for the desired variant properties, thereby improving the prospects of identifying a protein with the desired properties. Advantageously, the present invention manages to achieve this combination of properties.

[0006] The advantages of the present invention are further enhanced in phage display systems that enable high valency (HV) display, i.e. systems designed to maximise the number of pIX fusion proteins displayed on the surface of the phage when compared to the number of wild type (or non-fusion) pIX proteins displayed on the surface.

[0007] Thus, in one aspect the present invention provides a vector construct comprising an open reading frame comprising a nucleic acid sequence encoding a modified pIX filamentous phage coat protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue.

[0008] Viewed alternatively, the present invention provides a vector construct comprising a nucleic acid sequence encoding a pIX phage coat protein (a modified pIX phage coat protein) in which the methionine (M) residue at position 1 is replaced by an alternative amino acid residue.

[0009] The vectors of the invention are expression vectors or expression constructs, i.e. are generally comprised of nucleic acid sequences which enable the expression (protein synthesis) of desired encoded protein components in an appropriate host cell.

[0010] As outlined above, this invention is in the field of phage display on the pIX phage coat protein. Thus, the vectors of the invention can be phage vectors or phagemid vectors (plasmids) the basic construction and components of which will be well known to a person skilled in the art and selected in order to achieve expression of phage proteins and packaging of phage particles in an appropriate host cell such that the heterologous or exogenous proteins (proteins of interest, POIs) fused to the modified pIX phage coat proteins of the invention are displayed on the surface of the phage particle.

[0011] Thus, when the vectors of the invention are used to transform an appropriate host cell, e.g. an appropriate prokaryotic host cell, such as an appropriate E. coli strain, phage particles are produced which contain desired POIs fused to the modified pIX phage coat protein of the invention and displayed on the surface of the phage particle, with the vector sequences or other nucleic acid sequences encoding the various phage components of the phage genome and also encoding the POI contained within the phage particle.

[0012] In the modified pIX phage coat proteins of the invention, the methionine (M) residue at position 1 of the pIX phage coat protein, e.g. a wild-type or native pIX phage coat protein, can be replaced or exchanged by any alternative amino acid residue, i.e. any amino acid residue that is not methionine (M). Put another way, the N-terminal methionine (M) residue of the pIX phage coat protein, e.g. a wild-type or native pIX phage coat protein, can be replaced or exchanged by any alternative amino acid residue, i.e. any amino acid residue that is not methionine (M). These pIX phage coat proteins as used in the invention or as encoded by the constructs of the invention are referred to herein as modified pIX phage coat proteins or non-wild type pIX phage coat proteins of the invention. Thus, the modified pIX phage coat proteins of the invention do not correspond to wild-type pIX phage coat proteins in terms of amino acid sequence.

[0013] As having a genetic fusion between the POI and the modified pIX phage coat protein of the invention in a single ORF is advantageous for the present invention, conveniently appropriate alternative amino acids to replace the M at position 1 of the pIX phage coat protein will be genetically encoded amino acids which can then be encoded by a nucleic acid sequence and thereby readily included in the vectors of the invention.

[0014] In embodiments of the invention, the alternative amino acid residue which replaces the methionine (M) is selected from leucine (L), glycine (G), isoleucine (I), phenylalanine (F), tryptophan (W), tyrosine (Y), asparagine (N), glutamine (Q), glutamic acid (E), aspartic acid (D), proline (P), arginine (R), lysine (K), histidine (H), cysteine (C), serine (S), threonine (T), alanine (A), or valine (V).

[0015] In other embodiments of the invention, the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, H, C, S, T, or A.

[0016] In other embodiments of the invention, the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, or H, or selected from L, G, I, C, S, T, or A.

[0017] In some preferred embodiments, the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, or H.

[0018] In other preferred embodiments of the invention, the alternative amino acid residue is selected from L, G, or I, more preferably is L or G, most preferably L.

[0019] Although any alternative amino acid residue can be used, the choice may also be guided by the desired outcome of the phage display process. For example, in many instances it will be desired to have as high phage production as possible combined with as high fusion protein functionality as possible. This would ensure the ability to cover the largest possible functional diversity in any fusion protein library, that should maximize the ability to retrieve and identify desired novel fusion proteins resulting from library selection. In this case, the results presented herein show that the exchange of the Methionine with Leucine (M1 L), Glycine (M1G) or Isoleucine (M 11) appears to translate into the optimal blend of these two disparate but connected features.

[0020] However, in situations where for example large fusion protein diversity is of lesser importance, e.g. where a smaller library is sufficient (in which case high phage production is not needed), other Mix identities can be appropriate to maximize fusion protein functionality. In this case, the results presented herein show that the M1 E or M1 D represent appropriate alternative amino acids which give rise to very high target reactivity and that M1F, M1W, M1Y, M1N or M1Q would be other options. M1 P, M1R, M 1 K or M 1 H would be further options.

[0021] Alternatively, in situations where high phage production is of higher importance (but fusion protein functionality is a lesser priority, e.g. where high affinity clones are not necessarily required), then other Mix identities can be appropriate to maximize phage production. In this case, the results presented herein show that M1C or M1S or M1T or M1A represent appropriate alternative amino acids in this scenario. M1V would be a further option.

[0022] In some embodiments, the alternative amino acid residue is not R, K, D, S, A, V, T, C, H, P, E, Q, N, Y, W, F, I, G or L. For example, in some embodiments, the alternative amino acid is not R. In some embodiments it is not K. In some embodiments it is not D. In some embodiments it is not S. In some embodiments it is not A. In some embodiments it is not V. In some embodiments it is not T. In some embodiments it is not C. In some embodiments it is not H. In some embodiments it is not P. In some embodiments it is not E. In some embodiments it is not Q. In some embodiments it is not N. In some embodiments it is not Y. In some embodiments it is not W. In some embodiments it is not F. In some embodiments it is not I. In some embodiments it is not G. In some embodiments it is not L.

[0023] In some embodiments, position 2 of the modified pIX phage coat protein of the invention is S. In some embodiments, position 3 of the modified pIX phage coat protein is V. In some embodiments, position 4 of the modified pIX phage coat protein is L. In some embodiments, one or more, two or more, preferably all, of positions 2, 3, and 4 of the modified pIX phage coat protein is S, V and L, respectively. In some embodiments, position 2 of the modified pIX phage coat protein of the invention is not S. In some embodiments, position 3 of the modified pIX phage coat protein is not V. In some embodiments, position 4 of the modified pIX phage coat protein is not L. In some embodiments, one or more, two or more, preferably all, of positions 2, 3, and 4 of the modified pIX phage coat protein is not S, V and L, respectively. In some embodiments, position 16 is not C. In some such embodiments, said modified positions are not M.

[0024] For the avoidance of doubt, the modified pIX phage coat proteins of the invention (or the pIX parts / components of the vector constructs or nucleic acid molecules of the invention), have (or encode) an amino acid residue at position 1 of the pIX protein, but it is not methionine (M). In other words, the methionine (M) at position 1 of the pIX protein is replaced or exchanged for an alternative amino acid. It is not sufficient to merely delete or remove the methionine (M) at position 1. Thus, for example modified pIX proteins or pIX fragments in which the methionine (M) at position 1 has been deleted or removed (e.g. deleted or removed without replacement with an alternative amino acid residue) are not encompassed by the invention. Thus, for example, pIX fragments containing or consisting of position 2 onwards, e.g. position 2 to 32 for a full- length pIX protein, or pIX fragments in which the M at position 1 has been removed, are not encompassed by the invention. Indeed, it has been shown that vector constructs with a deleted methionine (M) at position 1 of the pIX protein do not show the same advantages and improved properties over wild-type pIX as the vector constructs of the invention.

[0025] It can be noted that the improvements observed using the vectors and systems, etc., of the present invention as described herein, typically refer to improvements over equivalent vectors, systems, etc., where a wild-type or native pIX phage coat protein is used.

[0026] As used herein, the term open reading frame (ORF) takes its standard art recognised meaning. Thus, the term open reading frame (ORF) is used herein to refer to a span of a nucleic acid molecule, typically DNA, between a start and stop codon, or between a translation start and translation stop site. Such ORFs typically encode a polypeptide, in this case a polypeptide which comprises a modified pIX phage coat protein of the invention as described herein. Preferred ORFs encode a fusion protein of a POI and a modified pIX phage coat protein of the invention. Appropriate start codons would be well known to a person skilled in the art. A typical and exemplary start codon would be ATG encoding methionine. The start codon would be positioned in the vector (or nucleic acid molecule) of the invention at an appropriate distance upstream of the sequence encoding the modified pIX phage coat protein in order for translation of the modified pIX phage coat protein to be initiated under appropriate conditions. Where a POI is also encoded by the vector (or nucleic acid molecule) of the invention, the start codon is positioned at an appropriate distance upstream of, for example close to, or directly adjacent to, the sequence encoding the POI-modified pIX fusion in order for translation of the POI-modified pIX fusion protein to be initiated under appropriate conditions. Appropriate stop codons would be well known to a person skilled in the art. Typical and exemplary stop codons would be TAA, TGA or TAG. One or more stop codons can be used. In some embodiments, vectors or nucleic acid molecules of the invention contain a single ORF.

[0027] The term "pIX phage coat protein" or “pIX protein” or “pIX phage protein” or “pIX coat protein”, etc., as used herein refers to a pIX protein originating from or derived from a filamentous phage, for example wild-type or native pIX filamentous phage coat protein sequences, or a pIX protein with a sequence which corresponds to the sequence of such a pIX protein. Preferred filamentous phages from which the pIX protein is derived or the pIX protein corresponds to are M13, fd, or f1 phages. Any appropriate pIX protein can be used providing it has the ability to display a POI as a pIX fusion protein on the surface of a phage particle. Although wild-type (or native) pIX proteins or wildtype like pIX proteins, e.g. which comprise all the amino acids of wild-type (or native) pIX, but also incorporating one or more additional amino acids, e.g. in the form of conditional mutations, are used in some embodiments of the invention, e.g. in the various helper phages as described herein, the pIX protein encoded by the vectors of the invention, or otherwise used in a fusion protein with a POI in order for the POI to be displayed on the phage surface, corresponds to a pIX protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue. In other words it is a modified pIX phage coat protein of the invention as described elsewhere herein.

[0028] Preferably, the modified pIX phage coat proteins encoded by the vectors of the invention comprise or consist of a full-length pIX phage coat protein providing that the methionine (M) residue at position 1 of the full-length pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein. Such full-length pIX phage coat proteins can typically have 32 amino acids.

[0029] Preferably, the modified pIX phage coat proteins encoded by the vectors of the invention comprise or consist of the following amino acid sequence, which corresponds to the wild-type pIX protein from the VCSM13 helper phage (Genbank AY598820.1).

[0030] MSVLVYSFASFVLGWCLRSGITYFTRLMETSS

[0031] (SEQ ID NO:1) providing that the methionine (M) residue at position 1 of the pIX filamentous phage coat protein (here SEQ ID NO:1) is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0032] An exemplary nucleic acid sequence encoding this sequence for inclusion in the vectors of the invention is provided elsewhere herein as SEQ ID NO:2, again providing that the nucleic acid sequence encoding the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by a nucleic acid sequence encoding an alternative amino acid residue, e.g. as described elsewhere herein. Thus, a yet further embodiment of the invention provides a vector construct of the invention, wherein the encoded modified pIX filamentous phage coat protein corresponds to the pIX coat protein from M13, fd or f1 phage, or a variant thereof, providing that the methionine (M) residue at position 1 is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0033] In other embodiments, e.g. when a variant of the modified pIX protein, e.g. a variant M13, fd or f1 pIX phage coat protein, is used, the encoded pIX protein comprises or consists of an amino acid sequence with a sequence identity of at least 60%, 65%, 70%, 75% or 80% to that of SEQ ID NO: 1 , such as at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95 %, or 96 % identity, providing that the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue. In other words, in the modified pIX proteins of the invention, position 1 of the pIX protein sequence should not be methionine (M). Preferred alternative amino acid sequences for inclusion at position 1 are described elsewhere herein and can result in improvements in phage display, for example improved production of infectious phage particles (infectious phage titre) and / or improved functional protein display, e.g. improved POI / antibody / fusion protein display, e.g. improved antigen or target binding. Without wishing to be bound by theory, it is believed that replacement of the M at position 1 of the pIX protein with an alternative amino acid improves the translation efficiency of the ORF of the vector constructs, and in particular has a positive effect on (or improves) the amount of POI-modified pIX fusion protein that is produced.

[0034] Equally the nucleic acid molecule encoding a variant of the modified pIX protein of the invention can for example comprise or consist of a nucleotide sequence with a sequence identity of at least 60%, 65%, 70%, 75% or 80% to that of SEQ ID NO: 2, such as at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99 % identity, providing that the methionine (M) residue encoded at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0035] Thus, a yet further embodiment of the invention provides a vector construct of the invention, wherein the encoded modified pIX filamentous phage coat protein comprises SEQ ID NO:1 (MSVLVYSFASFVLGWCLRSGITYFTRLMETSS), or a sequence with at least 60%, 65%, 70%, 75% or 80% etc., identity to SEQ ID NO:1, providing that the methionine (M) residue at position 1 is replaced by an alternative amino acid residue, e.g. as described elsewhere herein. Exemplary and preferred % identity values are provided elsewhere herein.

[0036] Other preferred examples of modified pIX sequences, e.g. variant modified pIX sequences, encoded by the vectors or nucleic acid molecules of the invention are sequences containing up to 12, e.g. up to 12, 11 , 10, 9, 8, 7, 6, 5, 4, 3, 2 or 1 , altered amino acids in the pIX sequence, e.g. SEQ ID NO:1 , providing that the methionine (M) residue at position 1 of the pIX filamentous phage coat protein (e.g. SEQ ID NO:1) is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0037] Thus, a further embodiment of the invention provides a vector construct of the invention, wherein the encoded modified pIX filamentous phage coat protein comprises SEQ ID NO:1 (MSVLVYSFASFVLGWCLRSGITYFTRLMETSS), or a sequence containing up to 12, e.g. up to 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2 or 1 , altered amino acids in the pIX sequence, e.g. in SEQ ID NO:1, providing that the methionine (M) residue at position 1 is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0038] These modified pIX sequences or variants thereof, e.g. modified or variant pIX sequences of the invention, should retain or have the functional ability to display a POI as a pIX fusion protein on the surface of a phage particle. Functional C-terminal truncations or N-terminal fragments of SEQ ID NO:1 (or the variant sequences) or other pIX sequences, could also be used providing that the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein, and that the ability to display a POI as a pIX fusion protein is retained. In other embodiments, fragments are not used. Thus, alternatively, full-length pIX proteins can be used, for example pIX proteins where all 32 amino acids (or variants thereof, e.g. as described elsewhere herein) are present, with the proviso that the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0039] In the present invention, "sequence identity" is a measure of identity between proteins at the amino acid level and a measure of identity between nucleic acids at the nucleotide level. The protein sequence identity may be determined by comparing the amino acid sequence in a given position in each sequence when the sequences are aligned. Similarly, the nucleic acid sequence identity may be determined by comparing the nucleotide sequence in a given position in each sequence when the sequences are aligned. When variant molecules are referred to herein, for example by percent identity values, e.g. at least 60%, 65%, 70%, etc., then it should be understood that such variants are generally functional variants or derivatives, i.e. display retained or improved function.

[0040] Methods to determine the percentage identity of two amino acid sequences or of two nucleic acid sequences are well known and described in the art, and any of these may be used. For example, to determine the percent identity of two amino acid sequences or of two nucleic acid sequences, the sequences are aligned for optimal comparison purposes (e.g., gaps may be introduced in the sequence of a first amino acid or nucleic acid sequence for optimal alignment with a second amino acid or nucleic acid sequence). The amino acid residues or nucleotides at corresponding amino acid positions or nucleotide positions are then compared. When a position in the first sequence is occupied by the same amino acid residue or nucleotide as the corresponding position in the second sequence, then the molecules are identical at that position. The percent identity between the two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = # of identical positions / total # of positions x 100). In some embodiments the two sequences are the same length.

[0041] One may manually align the sequences and count the number of identical amino acids. Alternatively, alignment of two sequences for the determination of percent identity may be accomplished using a mathematical algorithm. Such an algorithm is incorporated into the NBLAST and XBLAST programs of (Altschul et al. 1990). BLAST nucleotide searches may be performed with the NBLAST program, score = 100, wordlength = 12, to obtain nucleotide sequences homologous to (or with a certain % identity to) a nucleic acid molecules of the invention. BLAST protein searches may be performed with the XBLAST program, score = 50, wordlength = 3 to obtain amino acid sequences homologous to (or with a certain % identity to) a protein molecule of the invention. To obtain gapped alignments for comparison purposes, Gapped BLAST may be utilised. Alternatively, PSI-Blast may be used to perform an iterated search which detects distant relationships between molecules. When utilising the NBLAST, XBLAST, and Gapped BLAST programs, the default parameters of the respective programs may be used. See http: / / www.ncbi.nlm.nih.gov. Alternatively, sequence identity may be calculated after the sequences have been aligned e.g. by the BLAST program in the EMBL database (www.ncbi.nlm.gov / cgi-bin / BLAST). Generally, the default settings with respect to e.g. “scoring matrix” and “gap penalty” may be used for alignment. In the context of the present invention, the BLASTN and PSI BLAST default settings may be advantageous. In calculating percent identity, only exact matches are counted.

[0042] In preferred vectors of the invention, the sequence encoding the modified pIX phage coat protein of the invention is linked, e.g. operably linked, to a sequence encoding a protein of interest (POI). Thus, in preferred embodiments the vector construct, or the open reading frame (ORF) of the vector construct, further comprises a sequence encoding a protein of interest (POI) linked, e.g. operably linked or fused, to the sequence encoding the modified pIX phage coat protein of the invention. Thus, preferred vector constructs and nucleic acid molecules of the invention encode a POI- modified pIX fusion protein.

[0043] The term “fusion protein”, “fused”, etc., is used herein to describe the functional joining of two or more protein components in the same polypeptide sequence or in the same open reading frame (ORF). Such fusion proteins can also be described as genetic fusions as they are encoded by the same nucleic acid sequence (sometimes called a “fusion gene” or “fusion nucleotide sequence”). Although two (or more) protein components (or encoding nucleic acid sequences) can be directly adjacent to each other in such a fusion protein, equally the components can be joined by appropriate peptide spacers or linkers. As is well known in the art, spacers or linkers can be important to allow each of the individual protein components to be expressed in a functional manner, e.g. allowing them to form the appropriate three-dimensional structure to perform or maintain their desired function.

[0044] Thus, in the fusion proteins encoded by the vectors or nucleic acid molecules of the invention, a peptide spacer (or linker) is generally included between the protein of interest (POI) and the modified pIX phage coat protein of the invention. In other embodiments, such linkers or spacers need not be included, or may only be included in between some of the components. Thus, in the vectors or nucleic acid molecules of the present invention, the sequences encoding the POI can be fused to the sequences encoding the modified pIX phage coat protein of the invention with or without a spacer or linker sequence between the components. All these possibilities (i.e. fusion proteins or encoding nucleic acids with or without spacer or linker sequences) are still regarded as direct fusions or direct genetic fusions. Although this discussion focuses on a linker or spacer between the modified pIX phage coat protein of the invention and the POI, linker sequences may be included elsewhere in the vectors of the invention as appropriate, e.g. between other components of the vectors as discussed herein, for example between the VH and VL domains of an antibody POI or other POIs which involve or comprise two or more separate polypeptide components.

[0045] Thus, the term "pIX fusion protein" refers to a pIX protein (pIX phage coat protein, pIX filamentous phage coat protein), fused to an exogenous peptide / polypeptide, e.g. a protein of interest (POI). Similarly, the term “modified pIX fusion protein” refers to a modified pIX protein (modified pIX phage coat protein, modified pIX filamentous phage coat protein) of the invention fused to an exogenous peptide / polypeptide, e.g. a protein of interest (POI).

[0046] Preferred vectors of the invention thus comprise a sequence (a nucleic acid sequence) encoding a modified pIX phage coat protein of the invention fused (genetically fused) to a sequence encoding a POI (sometimes referred to herein as POI-modified pIX or POI-modified pIX fusion protein). The POI and the modified pIX can be in any appropriate order or spacing in the vector providing that, once expressed and packaged into phage particles, a functional fusion protein between the POI and the modified pIX is formed wherein the modified pIX coat protein component of the fusion protein forms part of the phage coat and the POI is functionally expressed or displayed on the surface of the phage particle. The POI part of the fusion protein is thus positioned in frame with the modified pIX coat protein part of the fusion protein. This means that the POI and the pIX are expressed in the same polypeptide sequence (or as part of the same ORF), or, put another way, as a direct fusion. In the vectors of the invention it is generally preferred that the POI component of the fusion protein be positioned N-terminally (or at or near the N-terminus) of the modified pIX component of the fusion protein.

[0047] A yet further aspect of the invention provides a modified pIX filamentous phage coat protein of the invention. In other words, such aspects of the invention provide a modified pIX filamentous phage coat protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein. In preferred embodiments, fusion proteins comprising said modified pIX phage coat proteins of the invention are provided, e.g. a fusion protein comprising a POI and a modified pIX phage coat protein of the invention. In other words, such aspects of the invention provide a POI fused (preferably N-terminally) to a modified pIX filamentous phage coat protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein. Nucleic acid molecules encoding such modified pIX filamentous phage coat proteins and fusion proteins are also provided.

[0048] The present invention extends to nucleic acid sequences or nucleic acid molecules which can form part of the vectors of the invention or which comprise components of the vectors of the invention. Thus, another aspect of the invention provides a nucleic acid molecule or nucleic acid sequence comprising an open reading frame comprising a nucleic acid sequence encoding a modified pIX filamentous phage coat protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue, e.g. as described elsewhere herein.

[0049] In addition, the present invention provides a nucleic acid molecule or nucleic acid sequence encoding the modified pIX phage coat protein of the invention or a fusion protein of the invention comprising a POI fused to a modified pIX phage coat protein of the invention.

[0050] In preferred embodiments of the invention one or more ribosome (ribosomal) binding sites (RBS) are included in the vector constructs. Such components can also be referred to as a translational initiation region (TIR).

[0051] The RBS sequence is located in the vector at an appropriate position for the RBS sequence to function. The role of the RBS is to recruit a ribosome during the initiation of protein translation and thus is conveniently placed at an appropriate distance upstream from the start codon of the protein it is desired to translate, or upstream of the ORF for the protein it is desired to translate. Thus, in the vectors of the present invention, the RBS sequence is conveniently placed upstream of the sequence encoding the POI-modified pIX fusion protein. In embodiments where a signal peptide is also part of the ORF, e.g. as discussed elsewhere herein, the RBS sequence is conveniently also placed upstream of the sequence encoding the signal peptide (or upstream of the start codon of the ORF comprising the nucleic acid sequence encoding the signal peptide). In embodiments where no signal peptide is present in the ORF, then the RBS sequence is conveniently placed at an appropriate distance upstream of the sequence encoding the POI-modified pIX fusion protein. The appropriate distance would be known or readily determined by a person skilled in the art depending on the RBS chosen. Exemplary distances might be seven or eight nucleotides from the ATG (or other) start codon, but this can vary.

[0052] The RBS / TIR sequence modulates the translation intensity (level of protein expression) of the sequences located downstream and different types of RBS can produce different levels of protein expression, for example weak or strong expression. Weak or strong RBS / TIR sequences are well known in the art and can readily be selected by a skilled person depending on the level of protein expression desired. As expected, a strong RBS facilitates or induces more translation (strong translation) as compared to a weak RBS. Both weak and strong RBS sequences can be used in the vectors of the invention. In some embodiments, a weak RBS is used.

[0053] In particular, in preferred vectors of the invention an RBS is included upstream (or 5’ or N-terminal to) to the start codon of the sequence encoding (or upstream, etc., of the start codon of the ORF comprising the nucleic acid sequence encoding) the POI- modified pIX fusion protein. A preferred RBS for use in the present invention is a Shine Dalgarno (SD) sequence or a SD based sequence which can be included in the vector constructs. SD sequences are well known and described in the art and any of these may be used. For example the core SD sequence is GAGG (SEQ ID NO:3) and other consensus sequences are GAGGAG (SEQ ID NO:4) or AGAGGAG (SEQ ID NO:5) or AGGAGAA (SEQ ID NO:6), for example comprising the sequence AGGAG (SEQ ID NO:7). Thus, SD sequences comprising these core or consensus sequences can be used.

[0054] An exemplary structure for a construct of the invention with a POI-modified pIX fusion protein is shown in Figure 3B. As described elsewhere herein, the POI as shown in Figure 3B is an scFv antibody fragment, but this is just an example of a POI or library of POIs that could be used.

[0055] In preferred vectors of the invention a sequence encoding a spacer or linker (typically a peptide spacer or linker) is included between the sequence encoding the POI and the sequence encoding the modified pIX phage coat protein. Such sequences are typically synthetic or artificial (e.g. non-native or non-natural) linker or spacer sequences, e.g. sequences that do not encode functional proteins or protein domains. Composite linker sequences can also be used. Such linker or spacer sequences may include tag sequences such as c-Myc or a FLAG tag (e.g. DYKDDDDK; SEQ ID NO:8). Full or full-length linker or spacer sequences are generally used, for example such sequences are generally not truncated sequences. The inclusion of such a sequence can aid the folding of the connected proteins, in particular the N-terminal protein (here generally the POI), and thus the spacer or linker length can be adjusted as appropriate to enable the best or satisfactory functional folding of both components (i.e. the POI and the modified pIX). Appropriate lengths could readily be determined by a person skilled in the art. However exemplary lengths would be between five and 15 amino acids (Weiss et al., 2000, Protein Sci. , 9:647-654), e.g. 6 to 10 amino acids. A particular linker used in the present invention is AAAGSKDIR (SEQ ID NO:12). Alternatively, a linker such as a GS linker, for example a linker with a certain number of GS repeats, e.g. G4S repeats, could be used.

[0056] If present in the vectors or fusion proteins of the present invention, such spacers or linkers form a distinct part of the vector or fusion protein than the modified pIX phage coat protein. In other words such spacers or linkers are not part of the modified pIX phage coat protein of the invention; the modified pIX phage coat protein of the invention is a distinct or separate component, e.g. there is a junction between the modified pIX phage coat protein and the upstream part of the vector. Thus, any alternative amino acid residue used as a replacement at position 1 of the modified pIX phage coat protein of the invention is part of the pIX component (part) of the vector and is not part of the spacer or linker sequence (or any other part of the vector); the spacer or linker sequence (or the POI sequence) is a distinct or separate component. Such distinct parts of the vector constructs are often separated by restriction enzyme sites or site-specific recombination sites. Thus, in some embodiments, a restriction enzyme site or site-specific recombination site is incorporated between the linker or spacer component (or other parts of the vector) and the modified pIX component.

[0057] As described elsewhere herein, preferred vectors of the invention can encode a protein of interest (POI) or targeting unit fused to the modified pIX phage coat protein. Such embodiments allow the display of a POI, e.g. a targeting protein, on a modified pIX coat protein of the invention. As described elsewhere herein the use of the modified pIX phage coat proteins of the invention can result in improved display of the POIs.

[0058] Thus, the POI (and indeed any linker or spacer sequence placed between the POI and the modified pIX phage coat protein) is typically exogenous or heterologous. When referring to an exogenous or heterologous protein, what is meant is a protein or peptide not originally part of the relevant phage coat protein, e.g. the pIX protein, etc., which is fused (with or without any linker or spacer amino acids, which are also exogeneous or heterogeneous and thus not part or originally part of the relevant phage coat protein) to the modified pIX phage coat protein of the invention, e.g. fused to the N-terminal end of the modified pIX phage coat protein of the invention, e.g. fused to the N-terminal amino acid residue used as an alternative to the M residue at position 1 of the pIX phage coat protein.

[0059] Any protein of interest (POI) can be encoded in the vectors of the invention providing that it is suitable for display on a phage and in particular as a fusion with a pIX phage coat protein, for example a modified pIX phage coat protein of the invention.

[0060] Appropriate examples are well known and documented in the art. However, preferred examples would be targeting molecules / targeting units or binding partners / binding proteins which can bind to other entities (targets / target entities, e.g. target proteins). Some preferred examples of POIs would be antibodies or fragments thereof (e.g. Fab, scFv, nanobodies), MHC molecules (class I or class II), T cell receptors (TCRs), or non-lg derived binding proteins such as DARpins, Ankyrin family, fibronectin family, knottins, anticalins, etc., (Hosse et al., 2006, Protein Sci 15:14-27) and peptides.

[0061] Appropriate designs for the vectors of the invention in order to display the chosen type of POI on the surface of phage would be readily determined by a person skilled in the art. For example, if the chosen type of POI is in the form of a single polypeptide chain, e.g. scFv antibodies or single chain TCRs or single chain MHC molecules, e.g. single chain MHC class I or MHC class II, then nucleic acid molecules encoding these polypeptides can simply be positioned in the vectors such that a fusion protein with the modified pIX phage coat protein of the invention is produced. If the chosen type of POI is in the form of two or more polypeptide chains, e.g. Fab antibody fragments or TCRs or MHC molecules with two chains, then nucleic acid molecules encoding one of the polypeptides (one of the chains) can be positioned in the vectors such that a fusion protein with the modified pIX phage coat protein of the invention is produced, and the other polypeptide chain(s) can be produced separately or independently.

[0062] The vectors of the present invention can be used for classical phage display in order to select binding partners (e.g. antibodies) for a particular target entity, e.g. target protein or target antigen. In such applications a library of POIs can be expressed on phage particles as part of a fusion protein with the modified pIX phage coat protein of the invention and selected for binding to a target entity by standard and well-known techniques.

[0063] Another preferred component of the vector constructs of the invention is an appropriate promoter sequence in order to control the expression of the ORF comprising the modified pIX protein of the invention and fusion proteins containing said modified pIX protein. Appropriate promoter sequences would be well known to a person skilled in the art and any of these could be used. An exemplary promoter sequence might be a lac promoter which can for example be induced with IPTG. Other promoters may include tac, arabB, or psp.

[0064] Optionally a signal sequence or signal peptide, e.g. a pelB signal sequence or signal peptide, can be included in the ORF comprising the nucleic acid sequence encoding the modified pIX phage coat protein of the invention. Thus, such a signal sequence or signal peptide can be present or absent in the vectors or nucleic acid molecules of the invention. If present, then an appropriate location would readily be determined. Such signal sequences are generally located upstream of (N-terminal of), but as part of the same ORF as, the POI-modified pIX fusion protein of the invention. In some embodiments a signal sequence or signal peptide is not used or present. Signal sequences or signal peptides can also sometimes be referred to as leader sequences or leader peptides.

[0065] Other optional features which may be present in the vectors (or nucleic acid molecules) of the invention would be well-known to a person skilled in the art. For example, the vectors, e.g. the phage vectors or phagemid vectors (which can collectively be termed phage display vectors or constructs) may optionally additionally contain other appropriate components, for example origins of replication, inducible or non-inducible promoters / operators for initiating transcription, enhancers, termination sequences, antibiotic resistance genes and markers, sequences encoding chaperone proteins (e.g. periplasmic chaperone proteins such as FkpA), signal sequences, linkers, protease sites, general tags or reporter molecules, restriction enzyme or sitespecific recombination sites to enable cloning and other manipulations, e.g. for cloning appropriate POIs into the vectors of the present invention in an appropriate position to form a fusion protein with the modified pIX phage coat protein, primer binding sites to enable amplification of the constructs by e.g. PCR, or other desirable sequence elements, for example, DNA sequences to allow the discrimination between different libraries by e.g. PCR. Appropriate sources and positioning of such additional components within the phage display constructs so that they perform their desired function would be well within the normal practice of a skilled person in the art.

[0066] As described elsewhere herein, nucleic acid molecules encoding the modified pIX phage coat proteins of the invention, or the fusion proteins of the invention which comprise said modified pIX phage coat proteins of the invention form yet further aspects.

[0067] As described elsewhere herein the vectors of the invention are primarily used for phage display and can therefore be phagemid vectors or phage vectors. Thus, in a yet further embodiment of the invention, the vector construct is a phagemid or a phage vector.

[0068] Phage display is a technique that is well known and described in the art. In this regard, in 1985, G. P. Smith established a method to display polypeptides on the surface of filamentous phage, a virus that infects E. coli cells (Smith, G.P., 1985, Science 228, 1315-1317). Since then, so called phage display has evolved into a powerful technology for protein engineering and selection of peptides and proteins binding a specific target (Loset and Sandlie, 2012, Methods 58, 40-46). The filamentous phage M13 is built from five different structural proteins. Protein VIII (pVII I) is the major coat protein, and the particle is capped at one end by 5 copies of pill and pVI, and at the other end by 5 copies of pVII and pIX. The particle infects F pilus+ E.coli by way of pill, and its ssDNA is injected into the bacterial cell. Here, phage DNA is replicated and transcribed, and new phage particles are assembled before nonlytic secretion into the growth medium.

[0069] In phage display, a gene encoding a protein of interest (POI) is normally placed between a gene encoding a coat protein (often pill but here pIX) and its N-terminal signal sequence, to produce a POI-coat protein fusion, although in some embodiments of the present invention signal sequences are not present. The term “phage library” or “library of phage particles” or similar refers to a collection of unique phages that differ in the amino acid sequence of the POI, and can be prepared by standard molecular cloning techniques. A library may well contain >1O10members, and can be used for selection of specific binders.

[0070] Thus, the present invention further provides phage or phage particles comprising the vectors or nucleic acid molecules of the invention and expressing a modified pIX filamentous phage coat protein or a modified pIX-fusion protein of the invention on the surface. The phage particles may thus comprise a phage genome or phagemid, preferably a phagemid. Such phage or phage particles can be any filamentous phage. Preferred examples are Enterobacteria phage, for example M13, fd or f1 phages.

[0071] Another aspect of the present invention provides a library of phages / phage particles, e.g. filamentous phages, produced using and therefore comprising the vectors (or nucleic acid molecules) of the invention as described herein. Said phages comprise fusion proteins of POIs with modified pIX phage coat proteins as described herein. Thus, said filamentous phages display a POI or a library of POIs as fusions to the modified pIX phage coat protein of the invention. As with other phage display libraries, each individual phage particle expresses / displays the same POI, but the presence of multiple particles expressing different POIs allows the display of multiple (or a library or a plurality of) different POIs.

[0072] Thus, in such libraries, a collection of diverse protein fusions, e.g. diverse antibody fusions, with differing properties are displayed and selected on a desired target in the form of a phage display library, e.g. an antibody phage display library. In phage display used for target discovery, e.g. antibody discovery, a library is generally made up of a collection of either artificially or endogenously diversified proteins of interest, e.g. antibodies, fused to a phage capsid (here modified pIX), and these proteins of interest, e.g. antibodies, vary in their biophysical, biochemical and target binding properties. Such libraries are then employed to identify those variants that harbour the property of interest through a cyclic process termed panning where each clone in the library competes with each other to enrich for the favourable variants.

[0073] Thus, a yet further aspect of the invention provides a library of phage particles, wherein the phage particles comprise the vectors (or nucleic acid molecules) of the invention as described herein, and wherein multiple different proteins of interest are expressed on the surface of the phage particles fused to a modified pIX phage coat protein of the invention.

[0074] For aspects of the invention involving phage display, a general purpose phage display textbook such as "Phage Display in Biotechnology and Drug Discovery" by Sachdev S. Sidhu, 1995, or "Phage Display: A Laboratory Manual" by Barbas et al., 1994 can be referred to for relevant techniques and definitions. POI-modified pIX fusion proteins of the invention can be encoded either in a complete phage genome by insertion of the sequences encoding the POI-modified pIX fusion protein into the phage genome (phage vector display), or on a phagemid (phagemid display). A phagemid is a high copy number plasmid that can encode the POI-modified pIX fusion protein, and superinfection with a helper phage that provides the genetic material required for phage production, is required. Thus, in phagemid display, there are generally two sources of the coat protein that is utilized for POI display (here the pIX phage coat protein); the helper phage encoded pIX protein (e.g. a pIX protein that is not fused to a POI; a non-fused pIX protein), and the phagemid encoded POI-pIX phage coat protein fusion (here a POI-modified pIX fusion protein). The new virions that are produced will then have a mixture of phagemid derived POI-modified pIX fusion proteins and helper phage derived pIX coat proteins (non-fused pIX coat proteins). Such helper phage encoded pIX proteins / non-fused pIX proteins can be wild-type (or native), or wild-type like pIX proteins. Similarly, if a phage genome system is used, then non-fused, e.g. wild-type or wild-type like, pIX phage proteins also generally need to be present, although in some embodiments of the present invention, no non-fused, e.g. no wild-type or wild-type like, form of the pIX coat protein is present.

[0075] Thus, in the present invention, it is possible for the phage particles to be engineered to have one copy or to have multiple copies of the POI displayed on the modified pIX coat protein.

[0076] In a phage genome system this can for example be achieved by modifying the phage genome to contain a sequence encoding (or an ORF comprising a sequence encoding) a POI-modified pIX fusion protein of the invention. If this is the only version / form of the pIX phage coat protein in the phage genome then multiple copies of the POI will be displayed on the modified pIX coat protein, and high valency (HV) display should be achieved as there will be no other form of the pIX phage coat protein to compete for surface display. If, on the other hand, an alternative version / form of the pIX phage coat protein is provided in the system such that two versions of the pIX phage coat protein are present in the phage genome, e.g. by further modifying the phage genome to contain a sequence encoding (or an ORF comprising a sequence encoding) another pIX protein (non-fused pIX protein) as well as the modified-pIX phage coat protein of the invention, then the two forms of pIX will compete with each other for surface display and a mixture of POI-modified pIX fusion protein and non- fused pIX protein will be present on the surface, thereby achieving low valency (LV) display.

[0077] In a phagemid system this can for example be controlled by the helper phage which is used and in preferred embodiments of the invention phage particles with multiple copies of the POI displayed on the modified pIX coat protein are used. This can be achieved in any appropriate manner. However, in particular, to increase the display level and result in multiple copies of the POI on the surface of the phage (which can also increase the avidity of binding), it is possible to use a modified type of helper phage, for example a helper phage termed DeltaPhage, that allows high valency (HV) display on pIX. Such modified helper phages contrast the use of normal helper phages such as M13K07, VCSM13, R408 or similar that only allows for low valency (LV) display.

[0078] The helper phage called DeltaPhage reported by Nilssen et al (Nilssen et al., 2012, Nucleic acids research, 40, e120; WO 2011 / 036555), has at least one (e.g. two) amber mutations inserted close to the pIX start codon, i.e. close to the codon encoding the methionine (M) residue at position 1 of the pIX phage coat protein, thereby conditionally inactivating (conditionally suppressing) the helper phage encoded pIX. Specifically, these amber mutations were placed between position 2 and position 3 of the pIX phage coat protein, i.e. between the residues S and V of the wild-type pIX phage coat protein. However, other positions would be possible providing that said mutants would act to conditionally inactivate (conditionally suppress) the helper phage encoded pIX.

[0079] If this helper phage is then superinfected into a host cell (e.g. E. coli) transformed with a phagemid encoding a POI-pIX fusion, e.g. a POI-modified pIX fusion protein of the invention, then in a host cell which suppresses the amber mutation, such as an amber suppressor strain (e.g. a supE+ strain) of E coli, intermediate (low) valency display of the POI-pIX is seen, whereas in a host cell which does not suppress the amber mutation, such as an amber non-suppressor strain (e.g. a supE- / supE negative strain) of E. coli, high valency display of the POI-pIX is seen. This is because, in amber suppressor host cell strains (e.g. supE+ E. coli), pIX (non-fused pIX) from the helper phage is produced which results in intermediate (low) valency display as the pIX (nonfused pIX) from the helper phage competes with the POI-pIX fusion protein from the phagemid for display and a mixture of both are displayed. However, production of pIX (non-fused pIX) from the helper phage is suppressed or blocked in amber non- suppressor strains (e.g. supE- / supE negative strains), and only the phagemid encoded POI-pIX fusions should be present, here the POI-modified pIX fusion protein of the invention, thereby resulting in high valency display of said fusion protein on the modified pIX.

[0080] Thus, in preferred embodiments of the invention, the vector construct is a phagemid vector which encodes a POI-modified pIX fusion protein of the invention and such a vector construct is used in combination with a helper phage which has a conditional mutation such that expression / production of the helper phage encoded pIX (non-fused pIX) phage protein can be controlled, which in turn can enable control of the number of POI-modified pIX fusion proteins on the surface of the phage. In embodiments where the conditional mutation is not suppressed, e.g. when a non-suppressor strain of E coli, e.g. an amber non-suppressor strain, e.g. sup E- / sup E negative strain, is used, then the helper phage encoded pIX should not be expressed / produced (will be suppressed) and only the POI-modified pIX should be expressed / produced resulting in only POI-modified pIX fusion proteins of the invention on the surface (high valency, HV, display). In embodiments where the conditional mutation is suppressed, e.g. when a suppressor strain of E coli, e.g. an amber suppressor strain, e.g. sup E+ strain, is used, then the helper phage encoded pIX will be produced / expressed resulting in a mixture of helper phage pIX (non-fused pIX) and POI-modified pIX fusion proteins on the surface (low valency, LV, display).

[0081] Suitable conditional mutations would be well known to a person skilled in the art and helper phage vectors can readily be designed, and appropriate host cells chosen, so that the expression of the helper phage encoded pIX is under control of the conditional mutation. For example, in the helper phage called DeltaPhage, as described above, conditional mutations in the form of one or more suppressible stop codons (e.g. amber mutations / amber stop codons, or other suppressible stop codons such as ochre or opal mutations / stop codons) are used in conjunction with appropriate host cells to suppress or not suppress the conditional mutation.

[0082] Other means of achieving high valency (HV) display with a phagemid (plus helper phage) system are possible. For example, any helper phage in which the pIX phage coat protein is lacking (e.g. has been deleted) or which does not produce a functional pIX phage coat protein (e.g. due to mutation or truncation), can be combined with a phagemid of the invention, i.e. a phagemid comprising a sequence encoding (or an ORF comprising a sequence encoding) a POI-modified pIX fusion protein of the invention, in order to achieve HV display. Again, such systems are designed such that the POI-modified pIX fusion protein of the invention should be the only functional pIX coat protein in the system and hence should be the only pIX coat proteins displayed on the surface of the phage thereby resulting in HV display.

[0083] In some embodiments of the present invention, high valency display, and systems that allow high valency display, is preferred, as the modified-pIX vectors of the invention have been shown to be particularly effective and advantageous when combined with high valency display. Equally, however, the modified-pIX vectors of the invention are compatible with low valency display, and systems that allow low valency display. Methods and systems to achieve low valency display would be well known to a person skilled in the art. For example, the modified-pIX vectors of the invention, e.g. phagemid vectors, can be used with a helper phage such as DeltaPhage under suppressing conditions as discussed above. Alternatively, however, the modified-pIX vectors of the invention can be used with more conventional helper phages encoding pIX (non-fused pIX) coat proteins, for example where the expression of the pIX coat protein is not subject to specific control or suppression, such as M13K07 or VCSM13, to achieve low valency display with a mixture of helper phage (non-fusion) pIX and POI-modified pIX fusion proteins on the surface. Thus, in some embodiments low valency display, and systems that allow low valency display, can be used with the modified-pIX vectors of the invention.

[0084] HV display has not traditionally been used to identify high affinity binders, as it is believed that the avidity effect through the display of multiple copies of the POI might compromise high affinity selection. Instead LV display is generally used to allow for high affinity binders to be identified. However, it has been shown that the modified pIX vectors of the present invention advantageously can be used in a HV display system to identify high affinity binders. An HV display system has the further advantage of maximising the functional fraction of the phage particles, as more particles will have a POI-fusion protein meaning that more extensive functional diversity in the displayed POIs will be present, in turn meaning that finding a binder is more likely. Thus, the ability of the HV display systems of the invention to combine an improved functional fraction with improved functional properties of the candidates displayed, is highly advantageous, e.g. in terms of successfully identifying binders to a target of interest.

[0085] The term high valency (HV) display as used herein refers to a phage display system which is designed to maximise the number of copies of a particular phage coat fusion protein, here the number of modified-pIX fusion proteins, displayed on the surface of the phage when compared to the number of wild-type (or non-fusion) pIX proteins displayed on the surface. Such systems are thus designed such that all (theoretically all) of the copies of a particular phage coat protein, here the pIX phage coat protein, displayed on the surface of the phage should be POI fusion proteins, here POI- modified pIX fusion proteins. Thus, in HV display, the systems are designed such that 5 copies of the POI-modified pIX fusion protein can be displayed on each particle.

[0086] The term low valency (LV) display as used herein refers to a phage display system which is designed such that a mixture of coat protein fusion proteins, here modified- pIX fusion proteins, and wild-type (or non-fusion) pIX proteins are displayed on the surface of the phage. Such systems are thus designed such that not all (theoretically not all), i.e. less than 5, e.g. 4, 3, 2, or 1, copies (or non-maximum copies or low copies, e.g. less than 5, 4, 3, 2, or 1 , copies) of a particular phage coat protein, here the pIX phage coat protein, displayed on the surface of the phage are POI fusion proteins, here POI-modified pIX fusion proteins. Such systems are typically set up to achieve an average of 1 , or less than 1, copy of a POI fusion protein per phage particle (although such systems can be set up to achieve a higher average number if desired). Thus, in systems set up to achieve an average of 1 , or less than 1 , copy of a POI fusion protein per phage particle, many phage particles will not display a POI- modified pIX fusion protein at all.

[0087] The invention as described herein is designed for use in a prokaryotic system and not for example in a eukaryotic system. Thus, appropriate host cells are prokaryotic cells and in particular bacterial cells. Appropriate bacterial hosts for phage display which can be used to express the vectors and nucleic acid sequences of the invention and to package and produce phage particles would be well-known to a person skilled in the art and could be selected accordingly. Preferred bacterial host cells are Gram negative bacteria such as strains of E. coli. Exemplary E. coli strains would include XL-1 blue, TG1, ER2738, AVBIOOFmkll’, MC1061, SS320,TGP10F’, and K91 K. In some embodiments, non-suppressor strains, e.g. amber non-suppressor strains, are preferred, examples of which are SS320, TOP10F’, AVBIOOFmkll’, MC1061 , and K91 K. In other embodiments, suppressor strains are used, e.g. XL-1 Blue, TG1 or ER2738.

[0088] The term “phage”, often called bacteriophage, is used herein in its art recognised form as meaning a virus infecting, replicating and which is secreted from bacteria. A filamentous bacteriophage, or filamentous phage, is a phage with a single stranded DNA genome (ssDNA genome) which is packaged with phage coat proteins. The secreted filamentous phage particle has phenotypically a filamentous structure. Filamentous bacteriophage or filamentous phage are preferred for use in the present invention.

[0089] The term phage or filamentous phage or filamentous bacteriophage as used herein encompasses both phage genome derived virions and phagemid-derived virions. The term “phagemid” is a term of the art and refers to a type of cloning vector developed as a hybrid of the filamentous phage Ff and plasmids to produce a vector that can propagate as a plasmid, and also be packaged as single stranded DNA in viral particles. Similarly to a plasmid, a phagemid can be used to clone DNA fragments and be introduced into a bacterial host by a range of techniques (e.g. transformation, electroporation). However, infection of a bacterial host containing a phagemid with a 'helper' phage, for example VCSM13 or M13K07 or the above described DeltaPhage, provides the necessary viral components to enable single stranded DNA replication and packaging of the phagemid DNA into phage particles.

[0090] The term "helper phage" is a term of the art and refers to a virus which helps a separate and unrelated defective virus, e.g. a phagemid, which in itself is not a phage genome or a functional virus, but merely a plasmid containing one or several elements derived from a phage genome (here at least a modified pIX protein of the invention), to reproduce by infecting the same host cell that is already occupied by the defective virus (e.g. phagemid) and providing the proteins which the defective virus (e.g. phagemid) is missing and needs to complete its life cycle and form virions, e.g. containing the phagemid.

[0091] Preferred helper phage for use in the present invention are described elsewhere herein and include M13K07 (Stratagene), Hyperphage (Progen Biotechnik GmbH), R408 (Agilent Technologies) and VCSM13 (Stratagene). In preferred embodiments, the helper phage may be a helper phage with a conditional (or suppressible) mutation as described herein e.g. the DeltaPhage helperphage as described herein and in the art, or Phaberge, or Ex-phage.

[0092] Another aspect provided by the present invention is a phage display system comprising a vector (or nucleic acid molecule) of the invention. Preferred phage display systems comprise a vector (or nucleic acid molecule) of the invention, e.g. a phagemid vector of the invention, and a helper phage, e.g. as described herein, e.g. a helper phage capable of expressing a pIX phage coat protein (e.g. a non-fused pIX phage coat protein). Other preferred phage display systems of the invention comprise a vector (or nucleic acid molecule) of the invention, e.g. a phagemid vector of the invention, and a bacterial host cell, e.g. an E. coli host cell / strain. Appropriate host cells / strains are also described elsewhere herein and can be included as a component in all the phage display systems, kits, methods and uses described here. Other preferred phage display systems comprise a vector (or nucleic acid molecule) of the invention, a helper phage, e.g. as described herein, e.g. a helper phage capable of expressing a pIX phage coat protein (e.g. a non-fused pIX phage coat protein) and a bacterial host cell, e.g. as described herein, e.g. an E. coli host cell / strain.

[0093] Thus, a yet further embodiment of the invention provides a phage display system of the invention as described elsewhere herein, further comprising a helper phage and / or bacterial host cell strain, e.g. an E. coli host strain.

[0094] In some embodiments, the pIX phage coat protein encoded by the helper phage can complement or compete with the modified pIX phage coat protein encoded by the vector construct or nucleic acid molecule of the invention. In such embodiments, the pIX phage coat protein encoded by the helper phage can in effect provide additional copies of a pIX phage coat protein, e.g. additional copies of a non-fused pIX phage coat protein, e.g. additional copies of a functional pIX phage coat protein, which can be used to form the phage coat. In some such embodiments, the pIX filamentous phage coat protein encoded by the helper phage is produced or expressed under the control of one or more conditional mutations, for example one or more suppressor mutations, e.g. as described elsewhere herein. For example, in preferred embodiments, the suppressor mutation is a suppressible stop codon, preferably selected from the group consisting of amber, ochre and opal stop codons, more preferably an amber stop codon. A preferred helper phage for use in such systems is DeltaPhage, details of which are described elsewhere herein and in the art. Appropriate and preferred E coli host strains for use in such embodiments are suppressor strains, preferably an amber suppressor strain, more preferably XL-1 Blue, TG1 or ER2738.

[0095] In some embodiments, the pIX phage coat protein encoded by the helper phage cannot complement or compete with the modified pIX phage coat protein encoded by the vector construct or nucleic acid molecule of the invention. Such inability to complement or compete can result in any appropriate way. For example, such inability may arise because the pIX phage coat protein encoded by the helper phage is not functional, for example due to mutation or truncation, or because the pIX phage coat protein encoded by the helper phage is absent, for example due to deletion.

[0096] Alternatively, such inability may arise because the pIX phage coat protein encoded by the helper phage is not produced or expressed, e.g. the pIX phage coat protein is produced or expressed under the control of one or more conditional mutations, for example one or more suppressor mutations, e.g. as described elsewhere herein, and the production or expression is suppressed. For example, in preferred embodiments, the suppressor mutation is a suppressible stop codon, preferably selected from the group consisting of amber, ochre and opal stop codons, more preferably an amber stop codon, and the production or expression of the pIX phage coat protein is suppressed by using an appropriate bacterial host strain. A preferred helper phage for use in such systems is DeltaPhage, details of which are described elsewhere herein. Appropriate and preferred E coli host strains for use in such embodiments are nonsuppressor strains which do not allow production or expression of the pIX phage coat protein encoded by the helper phage, preferably an amber non-suppressor strain (or an ochre non-suppressor strain, or an opal non-suppressor strain), more preferably SS320 or TOP-1 OF’.

[0097] The vectors (or nucleic acid molecules) of the invention also find utility in phage display methods, i.e. can be used in phage display methods.

[0098] Thus, a yet further aspect of the invention provides a method for producing phage particles comprising the use of a vector construct or nucleic acid molecule of the invention or the use of a phage display system of the invention as described herein. Such phage particles are typically produced by methods involving the steps of introducing the vector constructs of the invention, together with appropriate helper phages if necessary, into an appropriate bacterial host cell, examples of which are described elsewhere herein.

[0099] A yet further aspect of the invention thus provides a method of phage display comprising the steps of: a. providing a bacterial host cell / strain, e.g. an E coli host strain, comprising a vector construct of the invention wherein the open reading frame further comprises a sequence encoding a protein of interest fused to the sequence encoding the modified pIX phage coat protein of the invention, wherein expression of said vector construct results in production of a protein of interest-modified pIX fusion protein; b. providing a helper phage; and c. infecting said bacterial host cell / strain, e.g. said E coli host strain, with said helper phage under conditions such that said host strain produces phage particles that display said protein of interest-modified pIX fusion protein.

[0100] In some embodiments, such methods can be used for high valency phage display. Thus, in some embodiments, said method is a method for high valency phage display, wherein the pIX filamentous phage coat protein encoded by the helper phage cannot complement the POI-modified pIX filamentous phage coat protein encoded by the vector construct of the invention.

[0101] In particular, the present invention provides a method for high valency phage display comprising the steps of: a) providing a non-suppressor bacterial host cell / strain, e.g. a non-suppressor E coli host strain, comprising a vector construct of the invention wherein the open reading frame further comprises a sequence encoding a protein of interest fused to the sequence encoding the modified pIX phage coat protein of the invention, wherein expression of said vector construct results in production of a protein of interest-modified pIX fusion protein; b) providing a helper phage, wherein expression of the pIX phage coat protein of the helper phage is under the control of one or more suppressor mutations; c) infecting said non-suppressor bacterial host cell / strain, e.g. said E coli host strain, with said helper phage under conditions such that the pIX phage coat protein encoded by the helper phage is not expressed or produced, such that said non-suppressor host cell / strain produces phage particles that display multiple copies of said protein of interest-modified pIX fusion protein.

[0102] As described elsewhere herein, in some embodiments such high valency display is preferred.

[0103] In other embodiments, the methods of the invention can be used for low valency phage display. Thus, in such embodiments, said method is a method for low valency phage display, wherein the pIX filamentous phage coat protein encoded by the helper phage can complement the POI-modified pIX filamentous phage coat protein encoded by the vector construct of the invention.

[0104] In particular, the present invention provides a method for low valency phage display comprising the steps of: a) providing a suppressor bacterial host cell / strain, e.g. a suppressor E coli host strain, comprising a vector construct of the invention wherein the open reading frame further comprises a sequence encoding a protein of interest fused to the sequence encoding the modified pIX phage coat protein of the invention, wherein expression of said vector construct results in production of a protein of interest-modified pIX fusion protein; b) providing a helper phage, wherein expression of the pIX phage coat protein of the helper phage is under the control of one or more suppressor mutations; c) infecting said suppressor bacterial host cell / strain, e.g. said E coli host strain, with said helper phage under conditions such that the pIX phage coat protein encoded by the helper phage is expressed or produced, such that said suppressor host cell / strain produces phage particles that display single or low copies of said protein of interest-modified pIX fusion protein.

[0105] In any method of phage display in accordance with the invention and as described herein, e.g. high valency or low valency phage display, preferably a library of vector constructs of the invention are used which encode multiple proteins of interest. In other preferred embodiments said phage display methods are used for the selection of a protein which binds to a desired target molecule.

[0106] When one or more proteins (POIs) have been selected using the methods of the invention, these proteins, or a component, fragment, variant, or derivative thereof may be manufactured or produced, and if desired formulated with at least one pharmaceutically acceptable carrier or excipient. Such manufactured molecules, or components, fragments, variants, or derivatives thereof, are also encompassed by the present invention. Alternatively, these molecules may take the form of nucleic acids encoding said proteins, which nucleic acids may in turn be incorporated into an appropriate expression vector and / or be contained in a suitable host cell. Thus, nucleic acid molecules encoding said proteins, or expression vectors containing said nucleic acid molecules form further aspects of the invention. Thus, a yet further aspect of the invention provides a method of producing or manufacturing a protein (POI) comprising the steps of selecting the protein according to the methods of the invention as described herein, manufacturing or producing said protein, or a component, fragment, variant, or derivative thereof, and optionally formulating said manufactured protein with at least one pharmaceutically acceptable carrier or excipient. Put another way said methods of the invention as described herein, e.g. methods for selecting a protein, may further comprise the step of manufacturing or producing said protein, or a component, fragment, variant, or derivative thereof, and optionally formulating said manufactured or produced antibody with at least one pharmaceutically acceptable carrier or excipient. Said variants or derivatives of protein may have at least 60, 70, 80, 90, 95 or 99% sequence identity to the original polypeptide from which they are derived.

[0107] Another aspect described herein, for example for use in the methods of the present invention, is a kit comprising a vector (or nucleic acid molecule) of the invention or a kit comprising a phage display system of the invention as described above, for example comprising a phagemid of the invention and a helper phage, preferably a helper phage as described herein, e.g. a helper phage wherein the pIX phage coat protein is expressed under the control of a conditional (or suppressible) mutation as described herein, or a kit comprising a vector (or nucleic acid molecule) of the invention, e.g. a phagemid vector of the invention, and a bacterial host cell, e.g. an E. coli host strain as described herein, e.g. a non-suppressor E. coli host strain. The kit could also include the necessary instructions for use. A kit comprising a phagemid of the invention, a helper phage, and a bacterial host cell as described herein is also provided. Preferred vectors of the invention, helper phage and bacterial host cells for use in such kits are as described elsewhere herein.

[0108] Preferred vectors (or nucleic acid molecules) of the invention for inclusion in such kits could comprise a modified pIX phage coat protein vector of the invention as described herein, further comprising one or more cloning sites (e.g. a multiple cloning site) suitable for cloning in a POI which would then be fused to the modified pIX phage coat protein.

[0109] A preferred kit can thus comprise or consist of a collection of reagents for generating phage particles with a fusion protein of a POI to a modified pIX coat protein of the invention. As well as the vector of the invention, a kit could include one or more components selected from: other phagemids, helper phages, bacterial strains and instructions. Preferred options for such additional components are as described elsewhere herein.

[0110] A yet further aspect of the invention provides the use of a vector construct, a nucleic acid molecule, a phage display system or a kit of the invention to produce phage particles, or for use in phage display. Put alternatively, the present invention provides a method for producing phage particles (or a method of phage display), said method comprising the use of a vector construct, a nucleic acid molecule, a phage display system or a kit of the invention. Such methods for producing phage particles typically involve the steps of introducing the vector constructs or nucleic acid molecules of the invention, together with appropriate helper phages if necessary, into an appropriate host cell, e.g. a bacterial host cell, examples of which are described elsewhere herein.

[0111] The phage particles of the invention as defined herein may also be used as molecular tools for in vitro applications and assays. The particles may be used in any assay in which display of a POI on a pIX phage protein is desired.

[0112] As preferred phage particles of the invention also display a POI which can be a specific binding partner or targeting unit, e.g. an antibody etc., as described elsewhere herein, these can function as members of specific binding pairs or targeting reagents, and such phage particles can be used in any assay where the particular binding pair member or targeting unit is required.

[0113] Thus, yet further aspects of the invention provide a reagent that comprises phage particles of the invention as defined herein and the use of such phage particles as molecular tools, for example in in vitro assays.

[0114] As used throughout the entire application, the terms "a" and "an" are used in the sense that they mean "at least one", "at least a first", "one or more" or "a plurality" of the referenced components or steps, except in instances wherein an upper limit is thereafter specifically stated.

[0115] In addition, where the terms “comprise”, “comprises”, “has” or “having”, or other equivalent terms are used herein, then in some more specific embodiments these terms include the term “consists of” or “consists essentially of’, or other equivalent terms. Methods comprising certain steps also include, where appropriate, methods consisting of these steps. In the methods described herein the method steps can be carried out in any appropriate order.

[0116] The term "increase" or “improve” or “enhance” (or equivalent terms) as described herein includes any measurable increase or improvement when compared with an appropriate control. Appropriate controls would readily be identified by a person skilled in the art and might include a level of a particular parameter as determined when a wild-type pIX phage coat protein is used in comparison to a modified pIX phage coat protein of the invention. Preferably the increase, etc., will be significant, for example statistically significant, for example with a probability value of <0.05, when compared to an appropriate control level or value. Methods of determining the statistical significance of differences are well known and documented in the art.

[0117] Some of the sequences referred to herein are summarised in the Table below, along with relevant identifiers.

[0118] All sequences in this Table are recited herein 5' to 3' or from the N-terminus to the C- terminus in line with convention in this technical field.

[0119] The invention will be further described with reference to the following non-limiting Examples with reference to the following drawings in which:

[0120] Figure 1. Polyclonal phage ELISA and single-clone screening for OMV-reactivity. (A) Normalized phage samples from RO and R3 outputs were analyzed for binding to OMV by ELISA. Phages displaying an irrelevant specificity (scFv anti-NIP) were included as control. (B) Random single colonies after R3 were rescued to high valence (HV) display for all libraries. Samples were analyzed for OMV reactivity by ELISA and scored positive with a S / B (signal / background) ratio >3. The percentage of OMV positive clones within each library group is indicated. Supernatant from empty E. coli SS320 was included as a control. (C) The phage libraries after R3 were reformatted by batch-cloning for soluble scFv E. coli expression and random single clones analyzed for OMV binding by ELISA. (Hoydahl et al., 2016).

[0121] Figure 2. SDS-PAGE / Western blot analysis. Normalized amounts of phages (left) of the fully human antibody phagemid library (Hoydahl et al., 2016), and (right) a defined anti-phOx scFv phagemid control clone, were separated by 4-12% SDS PAGE followed by anti-pIX Western blot analysis probed with polyclonal rabbit anti-pIX serum. A M13K07 helper phage was included as a control (C). Both phagemid samples were packaged as either low valence (LV - rescued with M13K07) or high valence (HV - rescued with DeltaPhage) display. The pIXwt and scFv-pIX fusions are indicated.

[0122] Figure 3. (A) Schematic illustration of the pV, pVII, pIX and pVIII encoding genomic region of M13 filamentous phage. The pIX ORF has the start codon internal in the pVII ORF and is expressed as a complete protein without any post-translational processing. (B) Schematic illustration of the scFv-pIX phagemid expression cassette. The heterologous scFv fusion is placed / V-terminally to the complete pIX capsid coupled through an artificial linker / spacer as published (Loset et al., 2011). The complete phagemid sequence can be accessed through the GenBank accession code HQ528250. Transcription is controlled by the lac promoter (LacPO) through the T7 terminator and the subsequent mRNA to protein translation of the continuous open reading frame (ORF) initiated at the Shine-Dalgarno (SD) sequence starting on the Met encoding start codon* (C) The amino acid sequence of the start on the scFv (*) and the junction (**) between the linker and native pIX is shown (upper region), where the native pIX Met (M) start residue is indicated in bold underlined (**). This residue was subject to site-specific mutagenesis (***) as indicated (lower panel).

[0123] Figure 4. Phage production following phagemid rescue determined by either infectious titer (cfuampR / rnl), or total virion content determined by OD using the formula (((A269nm 16

[0124] - A320nm) x 6.083 x10 ) / genome size = virions / ml). Figure 5. The Mix anti-phOx scFv phage were separately produced at LV in E.coli XL1-blue (A) or HV in E.coli SS320 (B), titrated and assessed for target binding to phOx-BSA in phage capture ELISA using phage about similar total amounts of virions per sample. Levels of target binding (left x-axis) with corresponding titers (right x-axis) were plotted in the same figure. (C) The anti-phOx scFv phage displayed on pIXwt were produced at low valence (LV) in E.coli XL1-blue or high valence (HV) in E.coli SS320, titrated and assessed for target binding to phOx-BSA in phage capture ELISA using phage concentrations at low titer (LT); 108cfuampR / rnl and high titer (HT) 101° cfuampR / rnl. (D) Mix anti-phOx scFv displaying phages were separated by 4-12% SDS PAGE followed by anti-pIX Western blot analysis probed with polyclonal rabbit anti-pIX serum (upper part). All phagemid were produced at HV in E.coli SS320, except for the wt phage, which also was produced at LV in E.coli XL1-blue. For comparing of amounts of phages used in the experiment, western blot analysis was performed with mouse anti-pl 11 (lower part).

[0125] Figure 6. Phage production following phagemid rescue determined by either infectious titer (cfuampR / rnl), or total virion content determined by OD using the formula (((A269nm 16

[0126] - A320nm) x 6.083 x10 ) / genome size = virions / ml). Each individual phage Mix version was individually produced and titrated. Thereafter, the phage samples were grouped according to their biochemical similarity and the averaged means ± SD of the combined titers within each group are shown as indicated.

[0127] Figure 7. The Mix Anti-NIP scFv phage were separately rescued with DeltaPhage at LV (A) and HV (B) using in E. coli XL1-Blue and SS320, respectively, titrated and assessed for target binding to NIP-BSA in phage capture ELISA using serial dilution of each individual Mix variant. Thereafter, the variants were grouped according to their biochemical similarity and the averaged means ± SD of the combined data of each individual binding curve within each group are shown as indicated. Notably, some of the samples produced very little phage and could only be tested at low titers.

[0128] Figure 8. Phage target binding and phage titers were given scores from 1 to 10 based on how the individual values relates to each other (A). Since all groups did not reach signal saturation in target binding, specific target binding was defined as the titer yielding 2x the baseline value and the hierarchy scored accordingly. The M1 L_I_G group representing the best score on the combination between target binding and phage titer (Q4) was separated into the individual amino acids (B). Figure 9. Comparison of M1L and M1G. The M1L, M1G and wt anti-phOx and anti- NIP scFv phages were separately produced at LV and HV in E. coli XL1-blue or SS320, respectively. Phages were titrated and assessed for target binding to phOx- BSA (A, C) or NIP-BSA (B, D) in phage capture ELISA at serial dilutions.

[0129] Figure 10. Low valent versus high valent display. The M1L and wt anti-phOx and anti-NIP scFv phages were separately produced at LV and HV in E. coli XL1-blue or SS320, respectively. Phages were titrated and assessed for target binding to phOx- BSA (A) or NIP-BSA (B) in phage capture ELISA at serial dilutions. Performance of the LV and HV phages were compared with the standard low valence protocol using M13K07 rescue and E. coli XL1-blue (M1L_standard-LV and wt_standard-LV).

[0130] Figure 11. Functional binding versus target concentration. Anti-phOx and anti-NIP scFv displayed on pIXwt and M1L were produced at HV in E. coli SS320. Phages were titrated and assessed for target binding to reducing amount of phOx-BSA (A) or NIP- BSA (B) in phage capture ELISA using serial phage dilutions.

[0131] Figure 12. Assessment II of preferential target-specific enrichment between pIXwt and plX-M1L in spiked panning.

[0132] The NIP-specific scFv was prepared in E. coli SS320 (HV) using DeltaPhage helper phage and spiked into a target-irrelevant scFv at 1 :107followed by 3 rounds of immobilized NIP-BSA panning. Forty randomly chosen single colonies were then packaged from each mock library before (RO) and after each round of selection (R1 - 3) and tested for target reactivity using an antigen-specific phage capture ELISA. Clones were regarded as positive if they exhibited at least 3-fold higher response than the background signal. The results are given as number of positive clones / total number of clones tested as indicated.

[0133] Figure 13. The fully human scFv antibody phage library displayed on pIXwt previously reported (Hoydahl et al., 2016) was reformatted to plX-M1 L and both libraries prepared at standard-LV and HV display from E. coli SS320 using the M13K07 and DeltaPhage helper phages, respectively. The apparent level of functionally folded scFv on the phages was then assessed by binding to the conformational specific superantigen protein L (pL) in serial dilutions of titrated phages using phage capture ELISA. A non-pL binding scFv control phage was included as control. Figure 14. The two indicated versions of the fully human scFv antibody phage library were used to individually select for pHLA-specific binders in three consecutive rounds (R1 - 3) of parallel panning using identical protocol towards two unrelated tumor associated antigen (TAA) specific pHLA targets. Following panning, equal amounts of polyclonal phages prepared from the R3 output, as well as the unselected libraries (RO), were tested for target specific binding in phage capture ELISA. All samples were tested on both targets to serve as both specific screen and mutual negative controls on apparent specificity. The results for each sample are shown as the ratio in signal on the specific versus the unspecific target as an indirect measure of target specific enrichment. An irrelevant scFv control phage was included as negative control (NC).

[0134] Figure 15. Random single clones from the R3 output selected towards pHLA TAA target 1 were expanded and phages produced using DeltaPhage rescue independent of the originating standard LV or HV form used in selection to maximize sensitivity in screening. The phages were individually tested for binding to the matched (TAA target 1) and mis-matched (TAA target 2 and 3) pHLA targets in phage capture ELISA. The clones are separated into their display capsid (A and B) and display version (standard- LV or HV) used in selection. The number of target specific clones in each version is indicated.

[0135] EXAMPLES

[0136] Improved antibody discovery through modified pIX display

[0137] Example 1. Identification of a modified pIX version yielding improved antibody display

[0138] MATERIALS AND METHODS

[0139] Single-clone phage expression

[0140] Single clones were packaged into 96-deep well plates using 400 pl culture medium for screening experiments or in 50 ml cultures for larger-scale expression. Brifely, clones were inoculated into YT-AG and cultured ON / 37 °C. For 96-deep well expression 10 pl were transferred to new plates containing fresh medium and grown for 3 h / 37 °C / 600rpm before superinfection with 109cfu DeltaPhage per well. For the larger-scale expression, cultures were reinoculated into fresh medium to ODeoonm of 0.05 and grown with rigorous shaking at 37 °C until ODeoonm reached 0.2 before superinfection with DeltaPhage or M13K07 at MO110. Plates and culture flasks were further incubated at 37 °C / 30 min with gentle agitation, and further 30 min with vigorous shaking before cells were pelleted and resuspended in 2x YT-AK. Phage were packaged ON / 30 °C. 100 pL cleared supernatants from the deep well packaging were used for screening in ELISA, whereas phage from the larger-scale expression was PEG-precipitated before being spot titrated1before being used in ELISA and WB analysis.

[0141] Single clone phage ELISA with anti-phOx and NIP phages

[0142] ELISA plates were coated with serial dilutions of phOx-BSA or NIP-BSA in PBS starting at 5 pg / mL, incubated ON / 4 °C and blocked with 4% skim milk powder in PBST for 1 h / RT. Phage were added at serial dilutions and incubated for 1 h / RT. Bound phage particles were detected with anti-M13-HRP (Amersham Biosciences, 1 :5,000). Phage samples and antibodies were diluted in PBST. Plates were developed with TMB solution and read at 450 nm using a microplate reader. Between each step, the plates were washed 3x with PBST.

[0143] SDS-PAGE and Western blot

[0144] Normalized phage samples of 2x109cfuAmpRphage samples were heated at 95 °C for 5 min with BOLT™ LDS sample buffer before separation on 4-12% Tris Plus gels in Bolt MES SDS running buffer (reagents from Invitrogen) at 220 V for 22 min along with Broad-range ladder. Proteins were blotted onto lmmobilon™-P membrane (Sigma) in Tris-Glycine buffer (25 mM Tris, 192 mM Glycine, 20% methanol, pH 8.3) using a semi-dry blotting apparatus and membranes were blocked with PBSM. Mouse anti-pl 11 (MoBiTec, 1 :5000) and anti-mouse IgG-HRP (1 :10.000) was used for pill detection. For pIX detection a polyclonal anti-pIX rabbit serum was generated by immunization using a peptide (N-CITYFTRLMETSS-C; SEQ ID NO:9) of the C-terminal pIX portion (AbMART). The anti-pIX serum was used at 1 :2000 in combination with anti-rabbit IgG-HRP (1 :5000). Western blots were detected by reading chemiluminescent signals.

[0145] Spiked panning

[0146] Phages displaying the NIP-specific scFv either on pIXwt or plX-M1 L were rescued with DeltaPhage from either of the two amber non-suppressor E. coli strains TOP10F’ or SS320, titrated and blended in a roughly 50 / 50 blend, followed by one round of selection on immobilized NIP-BSA. Briefly, ELISA wells (NUNC) were coated with 5 pg / mL phOx-BSA or NIP-BSA in PBS ON / 4°C. Phages were incubated with antigen for 1.5 h at RT with agitation. Wells were washed with 10xPBST and 5xPBS, and elution of bound phages was performed by 10 min incubation with 0.5 ml 0.5% trypsin. Both the input phage and eluates were used to infect E.coli, and 20-24 random single clones from each group were sequenced (Eurofins Genomics).

[0147] Reformatting of pIX library

[0148] A naive human scFv-pIX phage library2as reformatted to the scFv-plX_M1 L scaffold. The scFv cassette was PCR amplified directly from the scFv-pIX library phages using 1010cfuAmpRas template, 0.25 pM each of forward 5' - ATTAAAGAGGAGAAATTAACCATGGCC-3' (SEQ ID NO: 10) and reverse 5' - TTTGGATCCAGCGGCCGC-3' (SEQ ID NO:11) biotinylated primers (Eurofins Genomics) containing Ncol and Notl RE-sites and 0.05 ll / rnl Phusion High- Fidelity DNA polymerase. The correct band was extracted from an agarose gel, followed by digestion and capture of biotinylated ends using MyOne Streptavidin T1 magnetic beads (Invitrogen). scFv cassettes were purified and ligated into phosphatase-treated vectors in the presence of polynucleotide kinase at 16 °C overnight, followed by purification using Pellet Paint Coprecipitant (Novagen). 600 ng library DNA was transformed into E. coli SS320 (Lucigen) using aliquots of 350 pl and an ECM 600 electroporator (BTX) essentially as previously described3The transformation mixtures were plated on Bio-Assay dishes (Nunc) and incubated ON at 30°C. Amounts of 8.7x109primary transformants were obtained which were scraped from the plate and rescued (see phage rescue and PEG / NaCI precipitation section). The size of the library was limited to the size of the repertoire of the scFv-pIX library which was determined have a diversity of 3x108.2

[0149] Phage rescue and PEG / NaCI precipitation

[0150] Phagemid rescue was done by inoculating scraped material to ODeoo nm 0.05 in 2x YT supplemented with 30 p g / ml tetracycline, 100 p g / ml ampicillin and 0.1 M glucose (2x YT-TAG) and incubated with rigorous shaking at 37 °C until OD reached 0.1-0.2. The cultures were superinfected with (DeltaPhage for plX_M1 L Library, R1 , R2 and R3 or M13K07 for R3 only) at MOI20 and incubated with gentle shaking at 37 °C for 60 min, followed by rigorous shaking for 30 min, before centrifugation and medium replacement to 2x YT supplemented with 100 p g / ml ampicillin and 50 p g / ml kanamycin (2x YT-AK) and incubation further 7 h at 30 °C. Phage particles were purified and concentrated by 2x PEG / NaCI precipitation and resuspended in PBS and cfu was determined by spot titration.1

[0151] Protein L ELISA

[0152] ELISA plates were coated with 5 pg / mL protein L (pL) in PBS and incubated ON / 4 °C and blocked with 2% skim milk powder in PBST for 1 h / RT. Phage samples were added in serial dilutions and incubated 2h / RT. Bound phage particles were detected with an anti-M13 antibody (generated by immunization of chickens with M13 bacteriophages by Norwegian antibodies) conjugated to HRP. Phage samples and antibodies were diluted in PBST. Plates were developed with TMB solution and read at 450 nm using a microplate reader. Between each step, the plates were washed 3x with PBST.

[0153] Phage selection

[0154] Selection was performed using both solid- and solution-based panning. Pre-blocked phage samples were incubated for 1 hour with 100 nM biotinylated HLA-A2 loaded with a panel of irrelevant TAAs captured onto MyOne Streptavidin T1 beads. Unbound phage was transferred to new tubes and incubated with 100 nM biotinylated HLA- A2:TAA target 1, either pre-captured onto beads (solid based panning for R1) or in solution (solution-based panning for R2 and R3) for 1 hour, followed by capture on beads. Antigen concentration was decreased 10-fold in each round, and washing stringency increased from 8xPBST + 2xPBS in R1 to 13xPBST + 2xPBS in R2 and 18xPBST + 2xPBS in R3. Before R2 and R3, the phage samples were heat challenged for 15 min at 65°C, before used in panning. Tubes were briefly vortexed between each wash. 4% (w / v) non-fat skim milk powder (PBSM) or 2% (w / v) bovine serum albumin (essentially fatty acid free) was used as blocking reagents in alternating selection rounds. Elution was performed by 10 min incubation with 0.5 ml 0.5% trypsin followed by E. coli infection using half of the eluate. Infected colonies were scraped and rescued (see phage rescue and PEG / NaCI precipitation section). A small sample of the infected culture was removed for determination of output.

[0155] Screening of selection output

[0156] Single clones were packaged into 96-deep well plates as described.4Briefly, clones were inoculated into YT-AG and cultured ON / 37 °C / 600 rpm. 10 pl were transferred to new plates containing fresh medium and grown for 3 h before superinfection with 109cfu DeltaPhage per well. Plates were incubated at 37 °C / 30 min with gentle agitation, and further 30 min with vigorous shaking before cells were pelleted and resuspended in 50 l 2x YT-AK and phage were packaged ON / 30 °C. 100 pL cleared supernatants were used for screening in ELISA.

[0157] Single clone phage ELISA from phage selection

[0158] ELISA plates were coated with 5 pg / mL NeutrAvidin in PBS and incubated ON / 4 °C and blocked with 5% skim milk powder in PBST for 1 h / RT. Biotinylated pH LA variants were captured for 1 h / RT, followed by addition of phage. Bound phage particles were detected with anti-M13-HRP (Amersham Biosciences, 1:5,000). pHLA, phage samples and antibodies were all diluted in PBST. Plates were developed with TMB solution and read at 450 nm using a microplate reader. Between each step, the plates were washed 3x with PBST.

[0159] References

[0160] 1 Loset, G. A., Kristinsson, S. G. & Sandlie, I. Reliable titration of filamentous bacteriophages independent of pill fusion moiety and genome size by using trypsin to restore wild-type pill phenotype. BioTechniques 44, 551-554 (2008).

[0161] 2 Hoydahl, L. S. et al. Multivalent pIX phage display selects for distinct and improved antibody properties. Sci. Rep. 6, 39066, doi:10.1038 / srep39066 (2016).

[0162] 3 Tonikian, R., Zhang, Y., Boone, C. & Sidhu, S. S. Identifying specificity profiles for peptide recognition modules from phage-displayed peptide libraries. Nat Protoc 2, 1368-1386 (2007).

[0163] 4 Frick, R. et al. A high-affinity human TCR-like antibody detects celiac disease gluten peptide-MHC complexes and inhibits T cell activation. Science Immunology 6 (62), eabg4925, doi:10.1126 / sciimmunol.abg4925 (2021).

[0164] RESULTS

[0165] We have previously shown that antibody discovery using pIX may improve efficacy in both identifying desired target-specific antibody candidates, as well as yielding clones of superior biophysical properties as compared to those discovered by standard pill display (Figure 1)1. This beneficial property was achieved through the combined use of pIX as antibody display scaffold and a modified helper phage termed DeltaPhage2, that allows for multivalent high valency (HV) display of the pIX fusion protein.

[0166] In that previous study, there was an apparently staggering difference in the target specific-signal sampling bulk output in target binding assays following phage library selection and the actual downstream success rate in discovery, when comparing the use of pIX display with pill display as antibody display scaffolds (Figure 1A versus 1C). Traditionally, one expects a higher target signal from phage selection output to be an indication of more successful enrichment of specific clones, but this turned out not to be a good correlate in our case. Moreover, in HV phage display it is expected that functional affinity effects (also termed avidity) compromises high affinity selection, which contrasts low valency display (LV) that efficiently allow for higher affinity binders to be identified3. However, in our previous study this turned out not to be the case across two full-scale antibody library selections, as HV pIX display both outperformed LV and HV pill display in successful antibody identification (Figure 1 C), as well as consistently identified the strongest target binders1. There is a strong correlation between phage genome size and virion size4. Since, pIX is encoded by a much smaller gene than pill, the virions in pIX display have a smaller virion coat than pill-expressing phages4. This would lead to lower signals in target binding assays such as that shown in Figure 1A. In addition, we have previously found a higher tendency in forming large polyphage virions in the pill system further amplifying this effect as target-specific detection is based on virion recognition1 4.

[0167] To better understand if also differences in productive scFv antibody display between the use of pIX and pill as display scaffolds could influence how to explain these results, we performed both binding experiments of the diverse antibody repertoire- matched pIX and pill libraries to pL (a superantigen that binds functional antibodies5), as well as capsid-specific western blots to assess actual display levels. Indeed, the two most successful routes of display, LV pill and HV pIX, both had apparent similar antibody levels in these assays. Moreover, as expected there was a pronounced difference in display levels on LV and HV display.

[0168] Further studies focused on pIX display using western blotting followed by anti- pIX detection indeed showed a clear difference in apparent scFv antibody display between the LV and HV form (Figure 2).

[0169] However, we also made some additional observations in these assays. Firstly, for unknown reasons, the pIX devoid of any scFv fusion consistently yielded two distinct bands of about 12.5 and 18 kDa, which is higher than its estimated MW of about 3.6 kDa. Deviations in SDS PAGE migratory behaviour are however well known also for other M13 capsids such as pill6. Secondly, despite the fact that the DeltaPhage helper phage system has been devised to completely block pIXwt expression in amber non-suppressor E. coli strains such as TOP10F’ and SS3202, we consistently observed detection of apparent scFv fusion-devoid pIX in the HV samples (Figure 2, HV samples). Thus, in particular we were interested in the latter observation as it did not immediately point to unspecific degradation which usually gives multiple intermediates of lower MW.

[0170] It is well known that pIXwt has a complex and only partially characterised translational initiation apparently coupled to the upstream pVII capsid due to an inefficient pIX-specific Shine-Dalgarno (SD) site (Figure 3A)78. Thus, we re-inspected the original scFv-pIX fusion design to consider if there could be an alternative explanation for the apparent defined blend of both scFv-pIX and free pIX in the HV display samples (Figure 2). Our previous design contained only one SD that is located upstream of the scFv ORF, and which therefore should ensure an efficient single translational initiation of the complete heterologous fusion protein ORF. However, we use the complete native pIX ORF and therefore this design also contains two naturally occurring start codons encoding Methionine (M), namely the / V-terminal start of the scFv, as well as the native Methionine of pIX. We considered the possibility that a hitherto undescribed ribosomal wobbling, or a cryptic pIX-intrinsic SD-like sequence could lead to independent simultaneous translation of both products (Figure 3B). Such non-typical translation, such as skipping and translational wobbling, has not, to our knowledge, been described in the context of pIX in filamentous phages9 10.

[0171] To test the hypothesis of independent translation initiated by the native pIX Methionine (hereafter referred to as M1) in our system, we therefore made a limited Mix targeted design of alternatives to this M1 (Figure 3C). We made a number of modifications, for example removed the side-chain (M1A and M1G), introduced negatively charged residues compatible with the overall negative charge of the virion coat (M1D and M1E)11, as well as introduced a small hydrophobic side-chain (M1 L). In this initial test, we avoided positively charged R-chains, as that might confer virion instability in a virion coat proximal context12. The various pIX variants were tested in the context of phagemid display of a well characterized previously described anti-phOx scFv4. Virions were prepared using phagemid rescue using DeltaPhage both in the amber suppressor E. coli strain XL1-Blue and the non-suppressor E. coli strain SS320, rendering LV and HV display, respectively. First, we assessed if there was an apparent difference in virion production between the different Mix versions and unmodified pIX (termed wt) (Figure 4).

[0172] Here, we both determined the virion content based on infectious particles (cfuampR), as well as the total amount of virions based on genome count (A268)2. In normal phage production, about 10 - 50% of the virions are infective13, whereas in the case of polyphage where several genome unit lengths virions are formed there is a corresponding discrepancy between infectious and total virion titer1. In the case of LV display where helper phage pIXwt expression is allowed and thus complements the heterologous fusion protein, we observed uniform and about identical virion production measured by both methods, as expected (Figure 4, left panel). However, in the case of HV display this changed such that all the pIX Mix variants displayed varying degree of primarily reduced infectious titer (Figure 4, right panel). The most pronounced deleterious effect was seen with the M1 D and M1 E variants, which all exhibited an about 100-fold reduction in infectious titer, and without a concomitant reduction in the total genome count. On the other hand, the M1A, M1G and M1 L variants only showed a mild reduction in infectious titer as compared with the pIXwt variant and there was a corresponding reduction in the total virion titer.

[0173] We also tested the different phage variants for target binding ability in ELISA (Figure 5). A normalized amounts of virions set to the detection limit of the unmodified pIX display response (corresponding to 8x 107 / ml based on A268, Figure 5C), was used both for the LV and HV samples to better visualize the Mix response variation, and the individual total undiluted infectious titer was plotted in the graphs together with the target binding response. In correspondence with the result in Figure 4, the infectious phage input was roughly identical in all the LV samples (Figure 5A) whereas this differed substantially in the HV samples (Figure 5B). Moreover, in the LV samples there was a clear difference in antigen binding where all Mix variants in varying degree bound better than the scFv displayed on unmodified pIX (Figure 5A). Further, two samples clearly stood out as superior in target reactivity and to a similar level, namely the M1 D and M1 L variant. When changing the scFv display mode from LV to HV, it is expected that the apparent target reactivity increases due to both increased scFv display, as well as varying degree of functional affinity effects due to multivalent scFv display2. Indeed, we did observe improved target reactivity of all the Mix variants in HV form as compared with both unmodified pIX and also compared with the LV counterparts. This effect was pronounced for both the M1 D and M1 E variants, but importantly, these samples also exhibited a huge discrepancy between the virion and infectious titers. Notably, the M1 L variant again also stood out with respect to improved target reactivity, and this was without a severe reduction in infectious titer.

[0174] To get insight in the actual scFv-pIX display levels, in particular on the HV display versions, we also performed a pIX-specific western blot analysis comparing about equal amounts of total virions between each sample (Figure 5D). This analysis showed several interesting trends. Firstly, the apparent amount of scFv was similar between all samples. Secondly, there was a very clear difference in the amount of free pIX between the different Mix versions, where the unmodified pIX again showed a substantial amount of fusion devoid free pIX, and this was also the case for in particular the M1A variant. The M1G and M1 L variants showed somewhat intermediate phenotype and with clearly less free pIX than the unmodified pIX. In stark contrast, no free pIX were seen with the M1 D and M1 E variants. As mentioned above, it is here important to have in mind that these HV samples bear strong indication of polyphage formation due to varying discrepancy between total virion count and infectious titer. Independent of virion length (which varies with polyphage distribution), the stoichiometry of the tip capsids pVII / pIX and plll / pVI will be similar to the infectious titer14, thus we in parallel performed an anti-pl 11 western blot to account for actual virion count. This analysis indeed confirmed the infectious titers in that virion counts were particularly low in the two samples showing the strongest target reactivity (M1 D and M1 E). For this reason, we therefore cannot entirely exclude that also these samples have traces of free pIX that were below detection limit in the assay.

[0175] In summary, the combined results of these analyses of the anti-phOx scFv clearly single out the M 1 L variant as the most favorable variant that both preserves an overall good virion production both at LV and HV, as well as exhibiting a clear improvement in target reactivity that must be assigned to a pronounced reduction in scFv devoid pIX protein component. The results also strongly suggest that it is the native / V-terminal pIX Methionine that allows for this pIX off-product to be produced, and that the effect at large can be abolished by changing this particular residue.

[0176] Antibody phage display has a primary objective to be used in antibody engineering and discovery where collection of diverse antibody fusions with differing properties are displayed and selected on target in the form of an antibody library15. These antibodies will inevitably have varying intrinsic efficiency in display and the human anti-phOx scFv represents a rather well-behaving unit in this context due as it is derived from multiple rounds of optimization using phage display engineering16. We therefore expanded the analysis of Mix effects to also include an anti-NIP scFv originating from a hybridoma that is known to perform less optimal with phage display2 4. Since the initial analysis clearly showed that the impact on reduction in pIX off-target product, virion production and target reactivity varied depending on which amino acid M1 was exchanged to, we here tested all 20 genetically encoded amino acid variants. The individual Mix variants were again prepared in both the LV (XL1- Blue) and HV (SS320) version using DeltaPhage and the corresponding titers assessed with respect to infectious and total virion count (Figure 6).

[0177] At large, the result was similar to that seen with the anti-phOx scFv in that rather uniform and high titers were obtained in the LV version (Figure 6, left panel), whereas considerable differences emerged in the HV version when considering infectious virion count (Figure 6, right panel). To ease interpretation, we grouped the data of the individual Mix variants into bins representing similar trends which to a certain extent also coincided with the biochemical properties of the amino acids (positive versus negative charge, hydrophobic versus hydrophilic, etc). There was a clear difference in infectious phage production between some of the mutant groups as compared to the unmodified pIX - all revealed primarily in the HV form - and the most negative impact was seen with the M1E and M1D versions that exhibited a 4-order of magnitude drop in infectious titer as compared with unmodified (wt) pIX (Figure 6, right panel). We also assessed antigen specific reactivity against NIP-BSA in phage capture ELISA with the same phage preparations (Figure 7).

[0178] Here, we used serial dilutions of all samples instead of normalized single-point phage concentrations as with the anti-phOx scFv (Figure 5). Again, we grouped the data of the individual Mix variants into bins here representing similar profiles in antigen reactivity. First, as expected presumably due to the avidity effects associated with HV display, it was a clear difference in overall reactivity between the LV and HV groups (Figure 7A versus B). Notably, this effect was primarily restricted to the Mix variants. Second, the unmodified (wt) pIX version was clearly in the least target reactive group independent of LV or HV display. Third, the reactivity bin profile was somewhat more complex between the various Mix variants depending on LV or HV display. However, the unmodified pIX and M1V variants were consistently in the least reactive bin. As indicated, the M1 L and M1 I variants were also consistently in the bin with the highest target reactivity across both LV and HV form. As pointed to previously, effective use of phage display as an engineering and discovery tool requires both the highest possible functional display in combination with the highest possible infective virion production. This will ensure that the largest heterologous fusion protein library pool can be screened for the desired variant properties.

[0179] Thus, to integrate these two main properties into a combined analysis based on the HV display data in Figures 6 and 7, we separately scored phage production and target binding from 1 - 10 (where 10 is most favorable) and plotted these against each other (Figure s). Here, it became clear that the group containing the MI L, M1I, and M1G singled out as the most favorable group (Q4) with only minor individual variation (Figure 8A). A further stratification of this group into their individual amino acids showed that the M1L variant represented the best compromise between improved target binding and preserved phage production (Figure 8B). However, also the M1G variant also showed rather good performance and appeared potentially even marginally better in target binding. Their similar phenotypes were also in good correspondence with the initial analysis on the unrelated anti-phOx scFv showing a strong and similar reduction in the free pIX production (Figure 5D). Thus, we focused the further assessment on comparing matched samples of unmodified pIX with the M1 L and M1G variants in target binding ability using phage capture ELISA comparing both these two scFvs (Figure 9).

[0180] This focused side-by-side comparison further underscored the very similar effects of exchanging the M1 position to either M1 L or M1G, possibly with a marginally improved benefit with the M1 L version. Both in the case of the anti-phOx and the anti- NIP scFv the M1 modification was clearly superior to the unmodified counterpart. The effect was larger in the case of the anti-NIP scFv (Figure 9B and D) as compared with the anti-phOx scFv (Figure 9A and C), and for both scFvs the effect was larger for the HV display form as compared with the LV display form.

[0181] To confirm and further expand the previous analysis, we again repeated the LV and HV rescue using DeltaPhage with both the anti-phOx and anti-NIP scFv constructs, but now also including phagemid rescue with the standard M13K07 helper phage in E. coli XL1-Blue17. M13K07 (and other equivalent helper phages such as VCSM13 and R408) is known to give significantly lower antibody display than with use of DeltaPhage, which also translates to reduction in apparent target sensitivity2. To test if this was the case also here, we thus repeated the corresponding phage captures ELISAs (Figure 10).

[0182] As seen previously, the M1 L variant consistently gave strongest target reactivity with both scFvs and both in LV and HV display. The difference was again greatest with the HV version of the anti-NIP scFv. Moreover, the M13K07-rescued phages were clearly subordinate in target reactivity for both scFvs and compared with both the LV and HV versions. Interestingly, the differences between the unmodified pIX and M1 L also appeared equalized when using M13K07 pointing to the use of DeltaPhage as a key helper phage reagent to disclose the underlying disparate profiles.

[0183] In our previous study, it was a non-obvious difference to which degree classical pill and our original pIX display system were influenced by functional affinity effects diminishing the ability to discriminate between high and low affinity binders when comparing LV and HV versions as discovery tool1. Since the current Mix modification appeared to improve antibody display levels, and in particular in the HV version, it could point to the possibility that the modified pIX display method would also be prone to functional affinity limitations. To test this hypothesis, we thus repeated the phage capture ELISAs where we varied target density of immobilized antigen, as functional affinity is dependent on high target density and high valency3(Figure 11).

[0184] If functional affinity is operational, it is well recognized that the effect is gradually reduced concomitantly with a reduction in target density when probed with a multivalent antigen binding unit such as HV display (Crothers and Metzger, 1972, Immunochemistry 9(3): 341-357).. However, we did not observe any difference in binding when reducing antigen density neither for the anti-phOx nor for the anti-NIP scFv independent of using unmodified pIX or the M 1 L variant as display scaffold. Thus, the M1 L modification did not appear to alter the already beneficial property of the original design pIX in this regard1.

[0185] In summary, the results clearly point to the finding that the “leaky” free pIX phenotype observed with the original pIX display system1, is likely coupled to distinct translational initiation of two separate polypeptides encoded by the single scFv-pIX fusion expression cassette through employment of both the scFv and the pIX initiation codons (Figure 3). Moreover, the effect can be abolished or diminished by changing the native / V-terminal pIX Methionine to an alternative amino acid. All the alternative amino acids show good results. However, for certain applications, the alternative amino acids are preferentially Leucine, Isoleucine or Glycine, as these three residues offer the best compromise in preserving original high infectious virion production concomitantly with improving antibody display efficiency (improved functional display).

[0186] Example 2. Improved antibody discovery through use of M1 L modified pIX display

[0187] In Example 1 , we identified that the native / V-terminal Methionine of the pIX capsid in filamentous phage display could be beneficially exchanged with alternative amino acids, and in particular Leucine, to obtain improved heterologous fusion protein display. In those experiments, we used predictable single clone evaluation of performance. However, one of the primary applications of phage display is as a combinatorial engineering and discovery tool15’18. In phage display used for antibody discovery, a library is made up by a collection of either artificially or endogenously diversified antibodies fused to a phage capsid, and these antibodies vary in their biophysical, biochemical and target binding properties. Such libraries are then employed to identify those variants that harbour the property of interest through a cyclic process termed panning where each clone in the library competes with each other to enrichment for the favourable variants19. In the case of the original pIX phagemid display system, this has been extensively tested in panning and found to work very well1’4’20-22. In a first attempt to test if the apparently improved phenotype of the M1 L modification also translated to putatively further improved performance in panning, we performed a single cycle target enrichment assay where the anti-NIP scFv was displayed either on unmodified pIX or on the M1 L variant (Table 1). The two scFv variants were prepared in the HV version by DeltaPhage rescue from either the amber non-suppressor E. coli strains SS320 or TOP10F’. The phages were mixed at a tentatively 1 :1 blend and panned against immobilized target (NIP-BSA) followed by sequencing of random single clones before and after panning to disclose if one version was preferentially enriched at the expense of the alternative clone.

[0188] _ .. _ .. . Cycling strain / Seq.freq.

[0189] Spike Specificity

[0190] Helper phage |nput Outputwt MIL wt MIL

[0191] , TOPIOF / , , , ,

[0192] 50 / 50 NIP , , 10 / 24 14 / 24 0 / 24 24 / 24

[0193] Deltaphage

[0194] 55320 /

[0195] 50 / 50 NIP , , 15 / 20 5 / 20 1 / 18 17 / 18

[0196] Deltaphage

[0197] Table 1. Assessment I of preferential target-specific enrichment between unmodified pIX and plX-M1L Phages displaying the NIP-specific scFv either on pIXwt or plX-M1 L were rescued with DeltaPhage (rendering HV display) from either of the two amber non-suppressor E. coli strains 10F’ or SS320, titrated and blended in a roughly 50 / 50 blend, followed by one round of selection on immobilized NIP-BSA. Both the input blend and eluates were used to infect E. coli and 20 - 24 random single clones from each group were sequenced. Results are shown as the pIX identity of all clones yielding a resolved sequencing result.

[0198] The sequence analysis revealed only some minor variation in the initial spiking ratio whereas after panning a very strong preference for the M 1 L variant was seen, which indeed pointed to improved panning performance. We then extended this spiking approach to represent a more realistic library context by spiking the anti-NIP scFv into a large background of an irrelevant scFv at a 1 : 107level mimicking a medium sized diverse antibody library. We made matched blends where antibodies were displayed on either unmodified pIX or on M1 L and in HV display form followed by 3 rounds of consecutive cyclic panning to address how efficiently the specific scFv was retrieved (Figure 12).

[0199] Between 10 - 24 random single clones from each round of selection (R1 - 3) were sequenced to reveal the clonal identity and the result showed a striking preference for the M 1 L version in the HV display form as only here efficient enrichment of the anti-NIP scFv was obtained. Thus, both spiking experiments clearly showed the improved and highly beneficial effect coupled to the M1 L modification of pIX in ability to retrieve a desired antibody by panning and also when at low abundance in the initial repertoire. Motivated by these encouraging results, we then reformatted the previously reported diverse pIX displayed fully human scFv antibody library also to the M1 L variant1. This library has an estimated diversity of about 3x 108unique antibody clones and hence serves as a suitable source for discovery of potentially new antibody specificities for further drug development. Care was taken to preserve the original antibody repertoire and based on transformation frequency sequence analysis the new library was estimated to have about the same diversity as prior to reformatting. To address if putative differences in functional scFv display occurred between the original and reformatted library, we rescued both libraries from E. coli SS320 in both the standard LV (using M13K07) and HV (using DeltaPhage) display form. We then tested normalized amounts (based on infectious titer) of each library in serial dilutions towards binding to the conformational specific superantigen protein L (pL) in phage capture ELISA (Figure 13).

[0200] Indeed, the results showed that both libraries were highly reactive in the HV display form and the M1 L variant was clearly superior. Moreover, in line with the previous single clone analysis (Figure 10), both libraries showed about similar and significantly less reactivity in the LV form.

[0201] We then went on to do a full-scale antibody discovery campaign against two clinically validated unrelated pHLA targets equipped with tumor associated antigen (TAA) peptides overexpressed in human cancer. Each library was used in HV form through R1 and R2, whereas we split both display forms into either standard LV or HV form in R3. The panning was conducted essentially according to a previously described protocol21, followed by a polyclonal pH LA target-specific phage capture ELISA comparing the unselected primary libraries (RO) and the final R3 output samples (Figure 14). The two different pHLA complexes served as mutual negative controls for putative pHLA cross-reactivity or unspecific HLA binding.

[0202] In line with the previous results using defined model scFv clones, we also here observed an enrichment that was primarily restricted to the modified plX-M1 L library across both targets. Notably, there was a clear difference in apparent target specific signal between the two targets that could indicate either stronger enrichment of more specific clones and / or higher affinity of clones towards pHLA TAA target 1 as compared with TAA target 2. To clarify this aspect, we therefore picked 96 random single clones from each individual panning experiment focusing on TAA target 1 , which were expanded (using DeltaPhage rescue to maximize target binding sensitivity) and tested for specific target binding in pHLA phage capture ELISA (Figure 15). Here, the results clearly showed both an improved hit rate as well as apparent better target binding and specificity from the M 1 L variant both panned in the LV and HV form as compared with the unmodified pIX version.

[0203] In summary, our results give a clear explanation for the observation of both full- length scFv-pIX as well as fusion-devoid pIX in the HV display conditions where no free pIX should be expressed. The inclusion of the native / V-terminal Methionine of pIX in the original design appears to give rise to two alternative ORFs encoded from the same phagemid expression cassette and the amino acid identity in the M1 position is the single determinant regulating the feature. By altering the identity of the position 1 amino acid in pIX, the effect may be diminished or abolished. Multiple alternative identities appear to be allowed with varying effects, which can be stratified into the two key parameters important for the use of phage display as an engineering and discovery tool, namely phage production and fusion protein functionality. From a theoretical point of view, one would like as high phage production as possible combined with as high fusion protein functionality as possible. This would ensure the ability to cover the largest possible functional diversity in any fusion protein library, that should maximize ability to retrieve and identify desired novel fusion proteins resulting from library selection. In this regard, the exchange of the Methionine with Leucine (M1L) appears to translate into the optimal blend of these two disparate but connected features. In a spiked model selection of limited diversity this effect was very clear in that the M1L modification clearly showed the best efficiency in specific antibody retrieval, and this was in particular observed with the HV display form. When the M1L modification was further tested side-by-side with the unmodified counterpart in two fully-fledged diverse human antibody discovery campaigns using pHLA a target bait, we also here observed a pronounced improvement in apparent selection efficacy with the M1L variant, which indeed translated to increased ability to identify an expanded number of target-specific clones which also had superior target reactivity. Here, the effect could also be seen in the classical LV version underpinning the generic beneficial effect on the pIX display system independent of helper phage system employed.

[0204] Though we currently do not have a complete understanding of the precise mechanism underlying the dual ORF phenomenon, which is likely explained by either ribosomal skipping, wobbling or independent translational initiation, this will translate to the same conclusion on design for practical application to obtain improved functionality, namely to alter the M1 position preferably into either M1L, M1I or M1G. Further, our results point to a finding that the optimal compromise on phage production and fusion protein functionality may be best obtained using the M1L if generically the largest and most flexible diversity space is to be covered. In situations where e.g. large fusion protein diversity is of lesser importance, e.g. where a smaller library is sufficient (in which case high phage production is not needed), other Mix identities may be considered to maximize fusion protein functionality. In this case, our results show that the M1E or M1 D may represent a choice which gives very high target reactivity (M1 F, M1W or M1Y, or M1N or M1Q would be other options; M1P, M1R, M1 K or M1H would be further options). Alternatively, in situations where high phage production is of higher importance (but fusion protein functionality is a lesser priority, e.g. where high affinity clones are not necessarily required), then other Mix identities may be considered to maximize phage production. Here the results show that M1C or M1S or M1T or M1A may represent appropriate choices in this scenario.

[0205] We have also shown that the improvements seen with the modified pIX system described herein are not solely limited to antibody phage display libraries. For example, similar results and improvements have been observed with both T cell receptor phage display libraries and peptide MHC (pMHC) phage display libraries.

[0206] References

[0207] 1 Hoydahl, L. S. et al. Multivalent pIX phage display selects for distinct and improved antibody properties. Sci. Rep. 6, 39066, doi:10.1038 / srep39066 (2016).

[0208] 2 Nilssen, N. R. et al. DeltaPhage — a novel helper phage for high-valence pIX phagemid display. Nucleic acids research 40, e120, doi:10.1093 / nar / gks341 (2012).

[0209] 3 O'Connell, D., Becerril, B., Roy-Burman, A., Daws, M. & Marks, J. D. Phage versus Phagemid Libraries for Generation of Human Monoclonal Antibodies. Journal of molecular biology 321 , 49-56. (2002).

[0210] 4 Loset, G. A., Roos, N., Bogen, B. & Sandlie, I. Expanding the Versatility of Phage Display II: Improved Affinity Selection of Folded Domains on Protein VII and IX of the Filamentous Phage. PLoS ONE S, e17433 (2011).

[0211] 5 Graille, M. et al. Complex between Peptostreptococcus magnus protein L and a human antibody reveals structural convergence in the interaction modes of Fab binding proteins. Structure 9, 679-687. (2001).

[0212] 6 Hust, M. et al. Single chain Fab (scFab) fragment. BMC biotechnology?, 14 (2007).

[0213] 7 Endemann, H. & Model, P. Lcoation of Filamentous Phage Minor Coat Proteins in Phage and in Infected Cells. Journal of molecular biology 250, 496-506 (1995). 8 Blumer, K. J., Ivey, M. R. & Steege, D. A. Translational control of phage f1 gene expression by differential activities of the gene V, VII, IX and VIII initiation sites. Journal of molecular biology 197, 439-451 (1987).

[0214] 9 Cruz-Teran, C. A., Tiruthani, K., Mischler, A. & Rao, B. M. Inefficient Ribosomal Skipping Enables Simultaneous Secretion and Display of Proteins in Saccharomyces cerevisiae. ACS Synthetic Biology 6, 2096-2107, doi:10.1021 / acssynbio.7b00144 (2017).

[0215] 10 Goldman, E., Korus, M. & Mandecki, W. Efficiencies of translation in three reading frames of unusual non-ORF sequences isolated from phage display. FASEB J 14, 603-611 (2000).

[0216] 11 Passaretti, P., Sun, Y., Dafforn, T. R. & Oppenheimer, P. G. Determination and characterisation of the surface charge properties of the bacteriophage M13 to assist bio-nanoengineering. RSC Adv 10, 25385-25392, doi:10.1039 / d0ra04086j (2020).

[0217] 12 Lamboy, J. A. et al. Phage wrapping with cationic polymers eliminates nonspecific binding between M13 phage and high pl target proteins. Journal of the American Chemical Society 131 , 16454-16460, doi:10.1021 / ja9050873 (2009).

[0218] 13 Thomas, W. D. & Smith, G. P. The case for trypsin release of affinity-selected phages. BioTechniques 49, 651-654, doi:10.2144 / 000113489 (2010).

[0219] 14 Loset, G. A. & Sandlie, I. Next generation phage display by use of pVII and pIX as display scaffolds. Methods 58, 40-46 (2012).

[0220] 15 Ponsel, D., Neugebauer, J., Ladetzki-Baehs, K. & Tissot, K. High affinity, developability and functional size: the holy grail of combinatorial antibody library generation. Molecules 16, 3675-3700 (2011).

[0221] 16 Marks, J. D. et al. By-passing immunization: building high affinity human antibodies by chain shuffling. Biotechnology (N Y) 10, 779-783 (1992).

[0222] 17 Vieira, J. & Messing, J. Production of single-stranded plasmid DNA. Methods in enzymology 153, 3-11 (1987).

[0223] 18 Rakonjac, J., Bennett, N. J., Spagnuolo, J., Gagic, D. & Russel, M. Filamentous Bacteriophage: Biology, Phage Display and Nanotechnology Applications. Curr Issues Mol Biol 13, 51-76 (2011).

[0224] 19 Ledsgaard, L. et al. Advances in antibody phage display technology. Drug discovery today 27, 2151-2169, doi:10.1016 / j.drudis.2022.05.002 (2022).

[0225] 20 Huszthy, P. C. et al. B cell receptor ligation induces display of V-region peptides on MHC class II molecules to T cells. PNAS 116, 25850-25859, doi: 10.1073 / pnas.1902836116 (2019). Frick, R. et al. Affinity maturation of TCR-like antibodies using phage display guided by structural modeling. Protein Engineering, Design and Selection 35, gzac005, doi:10.1093 / protein / gzac005 (2022).

[0226] Frick, R. et al. A high-affinity human TCR-like antibody detects celiac disease gluten peptide-MHC complexes and inhibits T cell activation. Science Immunology 6, eabg4925, doi:10.1126 / sciimmunol.abg4925 (2021).

Claims

CLAIMS:

1. A vector construct comprising an open reading frame comprising a nucleic acid sequence encoding a modified pIX filamentous phage coat protein in which the methionine (M) residue at position 1 of the pIX filamentous phage coat protein is replaced by an alternative amino acid residue.

2. The vector construct of claim 1 , wherein the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, H, C, S, T, A, or V.

3. The vector construct of claim 1 or claim 2, wherein the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, H, C, S, T, or A.

4. The vector construct of any one of claims 1 to 3, wherein the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, or H, or selected from L, G, I, C, S, T, or A.

5. The vector construct of any one of claims 1 to 4, wherein the alternative amino acid residue is selected from L, G, I, F, W, Y, N, Q, E, D, P, R, K, or H.

6. The vector construct of any one of claims 1 to 5, wherein the alternative amino acid residue is selected from L, G, or I, more preferably is L or G, most preferably L.

7. The vector construct of any one of claims 1 to 6, wherein the modified pIX filamentous phage coat protein corresponds to the pIX coat protein from M13, fd or f1 phage or a variant thereof, providing that the methionine (M) residue at position1 is replaced by an alternative amino acid residue.

8. The vector construct of any one of claims 1 to 7, wherein the modified pIX filamentous phage coat protein comprises SEQ ID NO:1 (MSVLVYSFASFVLGWCLRSGITYFTRLMETSS), or a sequence with at least 70% identity to SEQ ID NO:1, providing that the methionine (M) residue at position 1 is replaced by an alternative amino acid residue.

9. The vector construct of any one of claims 1 to 8, wherein the open reading frame further comprises a sequence encoding a protein of interest fused to the sequence encoding the modified pIX filamentous phage coat protein.

10. The vector construct of claim 9, wherein the protein of interest is an antibody, a T cell receptor, or an MHC molecule.

11. The vector construct of any one of claims 1 to 10, wherein the vector is a phagemid or a phage vector.

12. A nucleic acid molecule encoding the modified pIX filamentous phage coat protein as defined in any one of claims 1 to 8, or the fusion protein as defined in claim 9 or claim 10.

13. Phage particles comprising the vectors or nucleic acid molecules as defined in any one of claims 1 to 12 and expressing a modified pIX filamentous phage coat protein or a modified pIX-fusion protein on the surface.

14. A library of phage particles, wherein the phage particles are as defined in claim 13, and wherein multiple different proteins of interest are expressed on the surface of the phage particles.

15. A phage display system comprising a vector construct or nucleic acid molecule as defined in any one of claims 1 to 12.

16. The phage display system of claim 15, further comprising a helper phage and / or an E. coli host strain.

17. The phage display system of claim 16, wherein the helper phage encodes a pIX filamentous phage coat protein, and wherein the pIX filamentous phage coat protein encoded by the helper phage can complement the modified pIX filamentous phage coat protein encoded by the vector construct or nucleic acid molecule as defined in any one of claims 1 to 12.

18. The phage display system of claim 16, wherein the helper phage encodes a pIX filamentous phage coat protein, and wherein the pIX filamentous phage coat protein encoded by the helper phage cannot complement the modified pIX filamentous phage coat protein encoded by the vector construct or nucleic acid molecule as defined in any one of claims 1 to 12.

19. The phage display system of any one of claims 16 to 18, wherein the helper phage encodes a pIX filamentous phage coat protein, and wherein the pIX filamentous phage coat protein encoded by the helper phage is expressed under the control of one or more conditional mutations, for example one or more suppressor mutations.

20. The phage display system of claim 19, wherein the suppressor mutation is a suppressible stop codon, preferably selected from the group consisting of amber, ochre and opal stop codons.

21. The phage display system of any one of claims 16 to 20, wherein said helper phage is DeltaPhage.

22. The phage display system of any one of claims 18 to 21, further comprising an appropriate non-suppressor E. coli host strain which does not allow expression of the pIX filamentous phage coat protein encoded by the helper phage.

23. The phage display system of claim 22, wherein said non-suppressor E coli host strain is an amber non-suppressor strain, preferably SS320 or TOP-1 OF’.

24. The phage display system of any one of claims 15 to 17, wherein said E coli host strain is a suppressor strain, preferably an amber suppressor strain, more preferably XL-1 Blue, TG1 or ER2738.

25. A method for producing phage particles comprising the use of a vector construct or nucleic acid molecule as defined in any one of claims 1 to 12 or the phage display system of any one of claims 15 to 24.

26. A method of phage display comprising: a. providing an E coli host strain comprising a vector construct of claim 9 or claim 10, wherein expression of said vector construct results in production of a protein of interest-modified pIX fusion protein; b. providing a helper phage; c. infecting said E coli host strain with said helper phage under conditions such that said host strain produces phage particles that display said protein of interest- modified pIX fusion protein.

27. The method of claim 26, wherein said method is a method for high valency phage display, wherein the pIX filamentous phage coat protein encoded by the helper phage cannot complement the modified pIX filamentous phage coat protein encoded by the vector construct of claim 9 or claim 10.

28. The method of phage display of claim 26 or claim 27, wherein said method is a method for high valency phage display comprising: a. providing a non-suppressor E coli host strain comprising a vector construct of claim 9 or claim 10, wherein expression of said vector construct results in production of a protein of interest-modified pIX fusion protein; b. providing a helper phage, wherein expression of the pIX filamentous phage coat protein of the helper phage is under the control of one or more suppressor mutations; c. infecting said non-suppressor E coli host strain with said helper phage under conditions such that the pIX filamentous phage coat protein encoded by the helper phage is not expressed, such that said non-suppressor host strain produces phage particles that display multiple copies of said protein of interest-modified pIX fusion protein.

29. The method of claim 26, wherein said method is a method for low valency phage display, wherein the pIX filamentous phage coat protein encoded by the helper phage can complement the modified pIX filamentous phage coat protein encoded by the vector construct of claim 9 or claim 10.

30. The method of phage display of claim 26 or claim 29, wherein said method is a method for low valency phage display comprising: a. providing a suppressor E coli host strain comprising a vector construct of claim 9 or claim 10, wherein expression of said vector construct results in production of a protein of interest-modified pIX fusion protein;b. providing a helper phage, wherein expression of the pIX filamentous phage coat protein of the helper phage is under the control of one or more suppressor mutations; c. infecting said suppressor E coli host strain with said helper phage under conditions such that the pIX filamentous phage coat protein encoded by the helper phage is expressed, such that said suppressor host strain produces phage particles that display single or low copies of said protein of interest- modified pIX fusion protein.

31. The method of any one of claims 25 to 30, wherein a library of said vector constructs are used which encode multiple proteins of interest.

32. The method of any one of claims 25 to 31 for the selection of a protein which binds to a desired target molecule.

33. The method of claim 32, further comprising the step of manufacturing or producing said protein, or a component, fragment, variant, or derivative thereof, and optionally formulating said manufactured or produced protein with at least one pharmaceutically acceptable carrier or excipient.

34. A kit comprising a vector construct of any one of claims 1 to 11 , a nucleic acid molecule of claim 12, phage particles of claim 13 or claim 14, or a phage display system of any one of claims 15 to 24.

Images