WO2015078840A1

WO2015078840A1 - Full and partial protein secretion and cell surface display using type iii secretion system

Info

Publication number: WO2015078840A1
Application number: PCT/EP2014/075478
Authority: WO
Inventors: Thomas MARLOVITS; Lisa KOENIGSMAIER; Julia Radics; Daniel STREBINGER
Original assignee: IMBA Institut fur Molekulare Biotechonologie GmbH; Boehringer Ingelheim International GmbH
Current assignee: IMBA Institut fur Molekulare Biotechonologie GmbH; Boehringer Ingelheim International GmbH
Priority date: 2013-11-26
Filing date: 2014-11-25
Publication date: 2015-06-04
Anticipated expiration: 2016-05-26

Abstract

The present invention provides a cell which fully or partially secretes a protein of interest via a type III secretion system (T3SS). The cells comprising a type III secretion system and a protein of interest with having an N-terminal secretion tag. Partial secretion is achieved by using an anchoring sequence which plugs the entry point of the T3SS to prevent the complete secretion of the protein of interest.

Description

Full and Partial Protein Secretion and Cell Surface Display

Using Type III Secretion System

Field of Invention

[001] The present invention generally relates to the field of recombinant biotechnology and cell culture technology, and more specifically to cells constructed to export desired biomolecules from the cells to extracellular environment such as cell supernatant or another cell. The present invention also relates to polypeptides exported for medical purposes such as for treating a disease. It concerns a method of generating novel host cells for biopharmaceutical manufacturing. In another aspect, the present invention is related to the field of displaying polypeptides on the surface of a cell membrane, a method for producing a library of cells displaying proteins or polypeptides on the cell surface and a method of screening them.

Background of the Invention

[002] The plasma membranes of cells present a barrier to the passage of intracellular proteins. Protein secretion is a useful tool for applications in biotechnology when proteins need to be exported for their function or for purification. Highly charged molecules in particular experience difficulty in passing across membranes, and many therapeutic proteins recombinantly expressed in host cells cannot be obtained unless the cells are disrupted and they are often engineered to be exported via a secretion path. Thus, there is a constant need for reliable means of exporting therapeutic proteins from cells.

[003] It is known that some bacteria secrete proteins into the host to subvert the antibacterial response and promote infection. This is often mediated by large, bacterial membrane protein-based complexes such as a type-3 secretion system (T3SS). Found in many gram negative bacteria, type III secretion system (T3SS) is a molecular machine that exports proteins through both membranes to the extracellular environment. T3SS is central to the virulence of many bacteria, including animal pathogens in the genera Salmonella, Yersinia, Shigella, and Escherichia and plant pathogens in the genera Pseudomonas, Erwinia, Xanthomonas, Ralstonia and Pantoea. These pathogens use T3SS to inject a series of proteins, so called bacterial "effector proteins," into the cytosol of host cells.

[004] Work has been previously done to identify the properties of natural effector proteins that make them conducive to secretion. Mutation of a single effector protein gene commonly was found to have little or no effect because of apparent redundancies among the effectors, and loss of the secretion system usually abolishes pathogenicity in mutants (Cornells and Van Gijsegem, Annu. Rev. Microbiol 54:735-774 (2000)). However, the range of foreign proteins that can be exported and factors limiting their secretion properties remain largely unknown. T3SSs from several bacteria have been utilized to export recombinant proteins. For example, T3SS from Vibrio parahaemolyticus was cloned into the nonpathogenic Escherichia coli K-12 strain to explore its ability to translocate to foreign proteins into host cell (Akeda et al., "Functional cloning of Vibrio parahaemolyticus type III secretion system 1 in Escherichia coli K-12 strain as a molecular syringe," Biochem Biophys Res Commun 427(2):242-7 (2012)). Akeda et al. have also used the secretion and translocation domains of type III secreted proteins to export the 438 amino acid long alkaline phosphatase (Akeda et al., "Chaperone release and unfolding of substrates in type Hi secretion," Nature 437(7060):91 1-5 (2005)).

[005] Other researchers observed that the length of polypeptide that can be exported through the T3SS is limited. It has been found that proteins having fewer than 550 amino acids can be reliably secreted. However, secretion declines significantly after about 700 amino acids and proteins larger than 864 amino acids were not transported at all (Widmaier et al., "Quantification of the physiochemical constraints on the export of spider silk proteins by Salmonella type III secretion," Microb Cell Fact 9:78 (2010)). No successful studies have been published which uses T3SS system to export proteins which are larger than 864 amino acids. [006] Because of its ability to export protein to the extracellular environment, T3SS represents a powerful tool in basic and clinical biotechnology strategies. It would be especially desirable if therapeutic proteins, which are often large in size, can be reliably exported via T3SS by non-pathogenic bacteria. There is a need for a method and a cell for secreting large polypeptides, especially polypeptides which are larger than 864 amino acids, via T3SS.

_***

[007] It must be noted that as used herein, the singular forms "a", "an" and "the" include plural references unless the context clearly indicates otherwise. Thus, for example, a reference to "a reagent" includes one or more of such different reagents and a reference to "the method" includes equivalent steps and methods that could be modified or substituted known to those of ordinary skill in the art.

[008] Unless otherwise indicated, the term "at least" preceding a series of elements is to be understood to refer to every element in the series. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the present invention.

[009] The term "and/or" wherever used herein includes the meaning of "and", "or" and "all or any other combination of the elements connected by said term". For example, A, B and/or C means A, B, C, A+B, A+C, B+C and A+B+C.

[0010] The term "about" or "approximately" as used herein means within 20%, preferably within 10%, and more preferably within 5% of a given value or range. It includes also the concrete number, e.g., about 20 includes 20.

[0011] The term "less than", "more than" or "larger than" includes the concrete number. For example, less than 20 means≤20 and more than 20 means≥20. [0012] Throughout this specification and the claims or items, unless the context requires otherwise, the word "comprise" and variations such as "comprises" and "comprising" will be understood to imply the inclusion of a stated integer (or step) or group of integers (or steps). It does not exclude any other integer (or step) or group of integers (or steps). When used herein, the term "comprising" can be substituted with "containing", "composed of", "including", "having" or "carrying." When used herein, "consisting of" excludes any integer or step not specified in the claim/item. When used herein, "consisting essentially of" does not exclude integers or steps that do not materially affect the basic and novel characteristics of the claim/item. In each instance herein any of the terms "comprising", "consisting essentially of" and "consisting of may be replaced with either of the other two terms.

[0013] It should be understood that this invention is not limited to the particular methodology, protocols, material, reagents, and substances, etc., described herein. The terminologies used herein are for the purpose of describing particular embodiments only and are not intended to limit the scope of the present invention, which is defined solely by the claims/items.

[0014] All publications and patents cited throughout the text of this specification (including all patents, patent applications, scientific publications, manufacturer's specifications, instructions, etc.), whether supra or infra, are hereby incorporated by reference in their entirety. Nothing herein is to be construed as an admission that the invention is not entitled to antedate such disclosure by virtue of prior invention. To the extent the material incorporated by reference contradicts or is inconsistent with this specification, the specification will supersede any such material.

Summary

[0015] The present invention provides for the first time a cell which is able to secrete large protein via the type III secretion system. The present invention is based in part on the surprising finding that the size of the protein does not limit its ability to engage with the secretion system as observed before. In addition, the inventors discovered that entry into the secretion system requires the substrate to be at least in partially unfolded state, and by fusing the substrate C-terminally to a conformationally stable moiety which is resistant to unfolding enables partial secretion as the moiety plugs the entry point of the needle complex and prevents substrate to depart fully from the needle complex.

[0016] In one aspect, the present invention provides one or more non-pathogenic cells for secreting therapeutic proteins via a type III secretion system (T3SS), wherein the cell comprises (i) a nucleotide sequence encoding a type III secretion system and (ii) a nucleotide sequence encoding a protein comprising a) an N-terminal secretion tag recognized by T3SS and b) the therapeutic protein. Preferably the protein is more than 864 amino acids long.

[0017] Also provided herein is one or more cells comprising (i) a type III secretion system and (ii) a protein which is more than more than 864 amino acids long and comprises an N-terminal secretion tag recognized by T3SS and a therapeutic protein.

[0018] In one preferred embodiments, the cell secreting a therapeutic protein via a type III secretion system (T3SS) comprises:

(i) a nucleotide sequence encoding a type III secretion system,

(ii) a nucleotide sequence encoding a protein comprising an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain, and said therapeutic protein, and

(iii) optionally a nucleotide sequence encoding a chaperone capable of binding to said chaperone binding domain.

The T3SS system comprised in the cell of the present invention is encoded by genes from Salmonella Pathogenicity Island 1 (SPI-1 ) locus and more preferably that from Salmonella typhimurium. In a preferred embodiment, the nucleotide sequence (ii) further comprises an anchoring sequence at its C-terminus. [0019] Such cells can be used in a method for the production and secretion of the therapeutic protein as well as for the treatment of disease where the secretion of the therapeutic protein is desired. The present invention also provides a pharmaceutical or diagnostic composition comprising the cell which secretes therapeutic proteins via a type III secretion system (T3SS). When using the cell for the production and secretion of the therapeutic protein, the cells are cultured to express therapeutic protein and the therapeutic proteins are subsequently harvested from the cell culture.

[0020] The complete polypeptide or protein (including the secretion tag and any other sequences) may have a length of more than 864 amino acid residues via T3SS, such as more than 865, 866, 867, 868, 869, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1 100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000 residues or more. The present invention cell encodes in particular proteins or polypeptides that are heterologous to the cell. Particularly preferred proteins are antibodies or fragments thereof.

[0021 ] As used throughout this specification, the terms "peptide", "polypeptide", and "protein" are used synonymously and refer to any proteinaceous compound comprising an ammo acid sequence of two or more ammo acid residues.

[0022] Therapeutic proteins secreted or produced (excluding the secretion tag, linkers, chaperone binding domain or other non-therapeutic sequences if present) according to the present invention may have a length of more than 629 amino acid residues via T3SS, such as more than 630, 631 , 632, 633, 634, 635, 636, 637, 638, 639, 640, 650, 660, 670, 680, 690, 700, 710, 720, 730, 740, 750, 760, 770, 780, 790, 800, 810, 820, 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000 residues or more. [0023] In a further aspect, the present invention discloses a cell comprising a type III secretion system (T3SS) for partial secretion. The term "partial secretion" refers to the secretion of a partial stretch of a polypeptide through T3SS. In contrast, a "full secretion" means that the substrate enters and completely departs from the T3SS to enter into extracellular environments. Substrates for partial secretion comprise an N-terminal secretion tag, a protein of interest, one or more effector domains of a T3SS effector protein to span the needle complex, and finally, a final portion (anchoring sequence) for plugging the entry point of the needle complex. Such cells are particularly useful for producing a library of cells to display polypeptides at cell surface and for screening a polypeptide of interest.

[0024] In a preferred embodiment, the present invention provides a cell comprising: a type III secretion system (T3SS), and

a protein comprising, from N- to C-terminus, an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain, a protein of interest, one or more effector domains of a T3SS effector protein having a length in amino acids that has at least the length of at least three SptP effector domains, and a protein having a globular structure or zinc finger domain.

[0025] The invention is also based partially on the discovery that by fusing effector domains having a given length C terminally to the protein allows the protein to pass through the needle complex. At the same time, to retain the protein on the cell surface, a protein having a globular structure or zinc finger domain is present C-terminally to the effector domains to prevent the protein of interest from fully departing the needle complex (see Example 2).

[0026] It has been discovered that the entire length of the injectisome may alter during secretion; it was observed that the needle appears longer when substrates are transported inside. The length of at least three SptP effector domains was found to be effective to span across the needle complex. The sequence of one SptP effector domain for example is listed in SEQ ID NO: 2.

[0027] In preferred embodiments, the spanning sequence when unfolded in the T3SS is at least 700 A, such as at least 710, 720, 730, 740, 750, 760, 770, 780, 790, 800, 810, 820, 830, 840, 850, 860, 870, 880, 890, 900 A, such as 1000, 1 100, 1200, 1300, 1400, 1500 A. It is within the skill of the artisan to measurement such length; methods provided in the example can for example be used.

[0028] The recombinant cells described in the present invention can be used to identify a cell secreting a certain protein of interest via T3SS. Thus, in this aspect, the present invention provides a method for identifying a cell secreting a protein of interest via a T3SS, comprising (a) providing one or more cell comprising a type III secretion system (T3SS) and a protein comprising, from N to C-terminus, an N-terminal secretion tag recognized by the T3SS, a protein of interest, one or more effector domains of a T3SS effector protein, and a protein having a globular structure or zinc finger domain; and (b) identifying said protein of interest with a detection moiety which recognizes said protein of interest. In a preferred embodiment, the protein of interest is an antibody or an antigen-binding fragment of an antibody.

[0029] The exact nature of this invention, as well as its advantages, will become apparent to a skilled person from the following description and examples. The present invention is not limited to the disclosed preferred embodiments or examples. A skilled person can readily adapt the teaching of the present invention to create other embodiments and applications.

Drawings Brief Description

Figure 1 : (a) Designed substrates used for secretion test. One, three or five tandemly repeated effector domains from the natural substrate SptP (SptP1 , SptP3, SptP5) are fused downstream of a secretion signal (N-term signal). Each substrate was constructed without (Set-1 ) and with (Set-2) GFP followed by a 3x-FLAG-tag at the C-terminus. (b) Test results showing substrate secretion into supernatant and expression in the cell. Wild-type SptP, SptP1 , SptP3 and SptP5 were all secreted into the cell culture medium. GFP-containing substrates (Set-2) were expressed but not secreted, (c) Ubiquitin fused substrates used in secretion test. Wild-type ubiquitin prevented substrate secretion whereas conformationally- destabilized mutant did not. (d) Overview of the secretion result from Fig. 1 (a)-(c) (see Example 1 and 2).

Figure 2: Reconstructed cryo electron tomogram of Salmonella typhimurium cells (SB905) containing T3SSs. Cells with T3SS and expressing wild type SptP are shown in the upper panel. Cells expressing SptP3-GFP is shown in the lower panel.

Figure 3: (a) Overview of the experimental set-up to monitor substrate-loading, (b) Inhibitory effect of recombinant substrates on the secretion of the natural substrates SptP and SipA. Increasing levels of SptP3-GFP correlated with a decreased secretion of SptP and SipA. (c) Inhibitory effect of recombinant substrates on the secretion of the natural substrate SipB (translocase). Increasing levels of SptP3-GFP correlated with a decreased secretion of SipB.

Figure 4: (a) SptP3-GFP containing needle complexes were immunopurified using anti FLAG which binds to the 3x FLAG-tag C-terminally fused to SptP3-GFP. (b) Electron microscopy of isolated needle complexes. An extra density is visible at the terminal end of the needle filament (white arrow) for substrate-trapped needle complexes.

Figure 5: (a) Model for immunogold labeling on GFP for substrate-trapped needle complexes. Anti GFP antibody was conjugated to colloidal gold for immunogold labeling of GFP on substrate trapped needle complexes (b) Electron microscopy of trapped SptP3-GFP labeled with an anti-GFP antibody conjugated to colloidal gold. The gold labels are exclusively located at the basal side of the needle complex.

Figure 6: (a) The longitudinal central section through the three-dimensional structure of substrate-free (w.t.) and substrate-trapped (+pSptP3-GFP) needle complexes reconstituted by single particle analysis on images obtained by cryo electron microscopy using 74964 and 82403 particles. Three-dimensional volumes (C3) have been filtered to 10 A of resolution (FSC 0.5). The presence and the position of the trapped-substrate within injectisomes are visualized by difference imaging, (b) Density scanning along single pixel-rows at different vertical positions (y=125 and y=250) clearly indicated a difference (see Δ) between substrate-trapped and substrate-free needle complexes at the position corresponding to the central, tunnel-like part, (c) Surface views of half-sectioned and 15 degrees tilted three- dimensional volume of substrate-free and difference volume of substrate-trapped needle complexes (Difference corresponding to presence of substrate (Asub); difference observed at the larger, inner ring-1 , (AIR1 )). (d) Illustration of the secretion path reconstituted from the structure examination. Hollow spaces ('portal', 'channel', 'atrium' and 'tunnel') define the secretion path within the needle complex. The narrowest part for substrate passage is the 10 A wide 'channel'. The direction of substrate movement is indicated with arrows.

Figure 7: Comparison of needle filament lengths of substrate-free (w.t.) and substrate- trapped (+pSptP3-GFP) needle complexes.

Figure 8: Sequences of Spt N terminal secretion tag, effector domain, chaperone binding domain, SptP3-GFP and SptP5GFP used in the examples.

Figure 9: Overview of the N signal sequences (SecSig) and chaperone binding domain (CBD), and effector domains of various Salmonella effector proteins.

Figure 10: (a) Schematic representation of the secretion assay in Example 7.

(b) Results of secretion of various proteins of interest.

(c) Blots for the precipitated AvrBs3 variant Fokl and VP64. Both constructs are efficiently expressed, but only AvrBs3-VP64 was secreted.

Figure 11 : (a) pCASP plasmid used in the secretion test in Example 7.

(b) Restriction sites on the plasmid.

Figure 12: Constructs expressed for T3SS secretion in Example 7. Items of the Invention

The present invention can also be characterized by the following items:

1 . A non-pathogenic cell secreting a therapeutic protein via a type III secretion system

(T3SS), preferably that encoded by Salmonella Pathogenicity Island 1 (SPI-1 ), said cell comprising:

(i) a nucleotide sequence encoding a type III secretion system,

(ii) a nucleotide sequence encoding a protein comprising an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain and said therapeutic protein, and

2. The non-pathogenic cell of item 1 , wherein expression of said nucleotide sequence(s) is driven by a constitutive or inducible promoter.

3. The non-pathogenic cell of item 1 or 2, wherein said T3SS is from Salmonella

typhimurium, Salmonella enterica, Pseudomonas aeruginosa, Yersinia

pseudotuberculosis, Yersinia pestis, Shigella felxneri, Citrobacter rodentium, Escherichia coli EHEC, Escherichia colia EPEC, Pseudoman syringae, Ralstonia solanacearum, Xanthominas campestris, or Erwinia amylovora.

4. The non-pathogenic cell of any one of the preceding items, wherein said N-terminal secretion tag and chaperone binding domain is selected from SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP, AvrA.

5. The non-pathogenic cell of any one of the preceding items, wherein said chaperone is selected from the Salmonella protein SicA, InvB, SicP, SigE, InvB, SicP, SigE, SicA.

6. The non-pathogenic cell of any one of the preceding items, wherein the following N- terminal secretion tag and chaperone binding domain/chaperones are applied SipA/lnvB, SipC/SicA, SopA/lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE.

7. The non-pathogenic cell of any one of the preceding items, wherein said nucleotide

sequence (i) is integrated into the genome of said host cell. The non-pathogenic cell of any one of the preceding items,

- wherein nucleotide sequence (ii) and (iii) are contained in a plasmid,

- wherein nucleotide sequence (ii) and (iii) are contained in an artificial bacterial chromosome,

- wherein nucleotide sequence (ii) is contained in a plasmid and nucleotide sequence (iii) is contained in an artificial bacterial chromosome, or

- wherein nucleotide sequence (iii) is contained in a plasmid and nucleotide sequence (ii) is contained in an artificial bacterial chromosome.

The non-pathogenic cell of any one of the preceding items, wherein nucleotide sequence (ii) and (iii) are co-expressed from the same plasmid. The non-pathogenic cell of any one of item 7 to 9, wherein the plasmid has a copy number of 20-30 or 30-100 or more per host cell. The non-pathogenic cell of any one of the preceding items, wherein the therapeutic protein does not have basic amino acid stretches oligomerization interfaces and/or a stable tertiary structure. The non-pathogenic cell of any one of the preceding items, wherein said host cell is E. coli, preferably E. coli Nissle. The non-pathogenic cell of any one of the preceding items, wherein said host cell is capable of injecting said therapeutic protein into eukaryotic cells via the T3SS of said host cell. The non-pathogenic cell of any one of the preceding items, wherein the chaperone binding domain and the therapeutic protein are separated by a linker. The non-pathogenic cell of any one of the preceding items, wherein the therapeutic protein can be cleaved off by a protease recognizing a protease cleavage site preceding the therapeutic protein. The non-pathogenic cell of any one of the preceding items, wherein said therapeutic protein is in the form of a concatemer. The non-pathogenic cell of item 16, wherein the concatemer contains 2, 3, 4, 5 or more copies of the therapeutic protein. The non-pathogenic cell of item 17, wherein each copy of the therapeutic protein can be cleaved off by a protease recognizing a protease cleavage site preceding the therapeutic protein. The non-pathogenic cell of any one of the preceding items, wherein the therapeutic protein has a length of more than 864 amino acids. The non-pathogenic cell of any one of the preceding items, wherein the therapeutic protein comprises an antibody, polyclonal antibody, monoclonal antibody, recombinant antibody, antibody fragment, Fab', F(ab')₂, Fv, scFv, di-scFvs, bi-scFvs, tandem scFvs, bispecific tandem scFvs, sdAb, V_H, V_L, humanized antibody, chimeric antibody, IgA antibody, IgD antibody, IgE antibody, IgG antibody, IgM antibody, intrabody, minibody or monobody. A pharmaceutical or diagnostic composition comprising a non-pathogenic cell of any one of the preceding items. A non-pathogenic cell of any one of items 1-20 for use in a method of treating a disease. Use of a non-pathogenic cell of any one of items 1-20 for the production and secretion of a therapeutic protein. A method for the production of a protein of interest, comprising culturing a nonpathogenic cell of any one of items 1 -20 and harvesting said therapeutic protein from the culture. A cell comprising a type III secretion system (T3SS) and a protein comprising from N- to C-terminus an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain, a protein of interest, a spanning sequence having a length in amino acids of at least three SptP effector domains, and a protein having a globular structure or zinc finger domain. The cell of item 25, wherein the spanning sequence comprises one or more effector domains a T3SS effector protein, and optionally wherein the domain is selected from SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP and AvrA. The cell of item 25 or 26, wherein the secretion tag and chaperone binding domain is selected from SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP and AvrA. The cell of any one of items 25-27, wherein T3SS is from Salmonella typhimurium, Salmonella enterica, Pseudomonas aeruginosa, Yersinia pseudotuberculosis, Yersinia pestis, Shigella felxneri, Citrobacter rodentium, Escherichia coli EHEC, Escherichia colia EPEC, Pseudoman syringae, Ralstonia solanacearum, Xanthominas campestris, or Erwinia amylovora. The cell of item 28, wherein the T3SS is encoded by genes from the Salmonella typhimurium pathogenicity island 1 (SPI-1 ) locus. The cell of any one of the preceding items, wherein said cell comprises a chaperone selected from SicA, InvB, SicP, SigE, InvB, SicP, SigE, SicA. The cell of any one of the preceding items, wherein the following N-terminal secretion tag and chaperone binding domain/chaperones are applied SipA lnvB, SipC/SicA, SopA lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE.

The cell of any one of the preceding items, wherein said cell is capable of injecting said protein into eukaryotic cells via its T3SS. The cell of any one of the preceding items, wherein the chaperone binding domain and the protein of interest are separated by a linker. The cell of any one of the preceding items, wherein the protein of interest can be cleaved off by a protease recognizing a protease cleavage site preceding and/or following the protein of interest. The cell of any one of the preceding items, comprising a plasmid comprising nucleotides encoding the N-terminal secretion tag, the optional chaperone binding domain, the therapeutic protein, and/or the chaperone. 36. A plurality of cells as defined in any one of items 25-35.

37. Use of the cell as defined in any one of items 25-35 or the plurality of cells of item 36 for screening the protein of interest.

38. A method for identifying a cell secreting a protein of interest via a T3SS comprising

(a) providing a cell as defined in any one of items 25-35 or a plurality of cells of item 36; and

(b) identifying said protein of interest with a detection moiety which recognizes said protein of interest.

39. A T3SS comprising a partially secreted protein in its secretion pathway, wherein the protein comprises a globular structure or zinc finger domain at the C terminus.

Detailed Description of the Invention

[0030] In a first aspect, the present invention provides in a non-pathogenic cell which secretes therapeutic proteins via a type III secretion system (T3SS), wherein the cell comprises (i) a nucleotide sequence encoding a type III secretion system and (ii) a nucleotide sequence encoding a protein comprising an N-terminal secretion tag recognized by T3SS and the therapeutic protein.

[0031 ] The present invention also provides a non-pathogenic cell which secretes therapeutic proteins via a type III secretion system (T3SS), wherein the cell comprises (i) a type III secretion system and (ii) a protein comprising an N-terminal secretion tag recognized by T3SS and the therapeutic protein.

[0032] It is also characterized in preferred embodiments that the protein has a length of more than 864 amino acids.

[0033] A "cell" as used in the present invention should be understood broadly. As used herein, a "cell" refers to any type of lipid bilayer-enclosed structures such as a prokaryotic cell or eukaryotic cell, which are generally regarded as the smallest structural and functional unit of an organism which is considered alive. A cell, however, can also refer to other lipid bilayer-endosed structures which are not considered alive, such as a cell organelle like liposome, a subbacterial component like bacterial ghosts, or achromosomal cells like minicells.

[0034] Preferably, cells used in the present invention are prokaryotes or eukaryotes. Examples include, but are not limited to, vertebrate cells, mammalian cells, human cells, animal cells, invertebrate cells, nematodal cells, insect cells, stem cells, yeast cell or fungal cells. It is preferred that the cell is a prokaryotic cell such as bacterial cells from Gram- negative bacteria, including cells from Enterobacteriaceae, and E. coli, Pseudomonadaceae, e.g., P. putida as well as Gram-positive bacteria such as cells from Lactobacteriaceae or Bacillaceae. Preferably, the cell is E. coli, Bacillaceae, Salmonella, Serratia, or Pseudomonas species.

[0035] Most preferably, the cells are E. coli, such as Escherichia coli strain Nissle 1917, which is also known as Escherichia coli 083:K24:H31 . It is non-pathogenic and has been used as probiotic agents in medicine for the treatment of various gastroenterological diseases, including inflammatory bowel disease.

[0036] As mentioned earlier, the T3SS can be present in a lipid bilayer-endosed structures which are not considered alive such as minicells. T3SS engineered in minicells is known in the art and has for example been described in Carleton et al., "Engineering the type III secretion system in non-replicating bacterial minicells for antigen delivery," Nature Commun. 4, 1590 (2013).

[0037] The term "non-pathogenic" as defined herein means that the cells do not cause significant disease in healthy animals such as human. The cell used in the present invention is especially not a pathogenic strain of E. coli such as enteropathogenic or enterohemorrhagic E. coli. Examples of enteropathogenic E. coli strains include strains from the serogroup 0127 such as 0127:1-16. Examples of enterohemorrhagic E. coli strains include strains from the serogroup 0157 such as 0157:1-17. Enteropathogenic or enterohemorrhagic E. coli is known to cause acute gastroenteritis in humans. Enteropathogenic E. coli is a frequent cause of infantile diarrhea and enterohemorrhagic E. coli causes a wide spectrum of illnesses ranging from mild diarrhea to hemorrhagic colitis and hemolytic uremic syndrome.

[0038] The cell of the present invention is preferably non-pathogenic. Furthermore, in one preferred embodiment a non-pathogenic cell is not a pathogenic bacterium which is attenuated because attenuated pathogenic bacteria whose pathogenicity characteristics have not been fully characterized may have unknown harmful effects if administrated to mammalian culture cells or the living body.

T3SS system

[0039] The cell of the present invention secrets a protein via a type III secretion system (also herein also referred to as "T3SS"). Type-3 secretion systems is known to involve 3.5 MDa syringe-like, membrane embedded injectisomes containing needle complex to connect intracellular compartments of infectious bacteria and hosts. T3SS are found in a variety of gram-negative pathogens such as Yersinia pestis, Yersinia pseudotuberculosis, Yersinia enterocolitica, Pseudomonas, Shigella flexneri, Shigella dysenteriae, Xanthomonas and some Salmonella sp. including Salmonella enterica, Salmonella typhimurium, Salmonella typhi, Salmonella enteritica as well as other pathogens including Vibrio cholerae, Hafnia alvei, Bordetella sp. and Chlamydia species. "Effector proteins" are injected by the bacteria via T3SS into the cytosol of host cells, which in turn modulate eukaryotic regulatory or signaling pathways during bacterial infection in the host cell. The system resembles a large supramolecular cylindrical structure embedded in both bacterial membranes. A needle filament protruding from the bacterial surface allows the transfer of proteins directly from inside the bacteria to the cytoplasm. The needle complex

[0040] The hallmark of T3SS is the needle complex (referred to also as NC) (also called injectisome when the ATPase is excluded). The needle complex is embedded within the inner and outer bacterial membrane, spans the periplasmic space, and extends into the extracellular environment with a needle-like filament. The cylindrically shaped needle complex (referred to as "injectisome") is composed of structural proteins forming a multi-ring base associated to the bacterial envelope and a needle-like extension that protrudes several nanometers from the bacterial surface. The needle is anchored to the base through another substructure, the inner rod, which together with the needle filament forms a channel that serves as conduit for the traveling of the effector proteins (Marlovits et al., Science 306, 1040-1042 (2004)).

[0041 ] The needle provides a smooth passage through the highly selective and almost impermeable membranes. A single bacterium can have several hundred needle complexes on its membrane. The needle complex is believed to be universal to all T3SSs. The needle complex shares similarities with bacterial flagella. The base of the needle complex is structurally very similar to the flagellar base; the needle itself is analogous to the flagellar hook, which is a structure connecting the base to the flagellar filament.

(i) A nucleotide sequence encoding a type III secretion system

[0042] As mentions earlier, the cell of the present invention may in some aspects comprise (i) a nucleotide sequence encoding T3SS. Such nucleotide sequences encode structural proteins making up the type III secretion system. "Structural proteins" as defined here includes proteins which form the base, the inner rod and the needle of T3SS.

[0043] Structures of T3SS have been intensively studied in the art for nearly 20 years, for example, as disclosed in Hueck, "Type III protein secretion systems in pathogens of animals and plants," Microbiol Mol Biol Rev 62:379-433 (1998), Keyser et al., "Virulence blockers as alternatives to antibiotics: type III secretion inhibitors against Gram-negative bacteria," J Intern Med 264:17-29 (2008), Coburn et al., "Type III Secretion Systems and Disease," Clin Microbiol Rev 20(4):535-549 (2007), Connelius, "The type III secretion injectisome," Nat Rev Microbiol 4(11 ):811-25 (2006), Kosarewicz et al., "The blueprint of the type-3 injectisome," Philos Trans R Soc Lond B Biol Sci 367(1592):1140-1 154 (2012), Mota et al., "The bacterial injection kit: Type III secretion systems," Ann Med 37: 234-249 (2005), and Yip et al., "New structural insights into the bacterial type III secretion system," Trends Biochem Sci 31 , 223-230 (2006). It is within the skill of ordinary person in the art to determine the nucleotide sequences which encodes T3SS. Location of the genes encoding T3SS in pathogens has also been identified. Salmonella, for instance, has a chromosomal region in which most T3SS genes are gathered, the so-called Salmonella pathogenicity island (SPI). Shigella, on the other hand, has a large virulence plasmid on which all T3SS genes reside. It is possible to clone these genes into another cell for recombinant expression of T3SS. It is known that type III secretion systems are highly conserved in a variety of gram- negative pathogenic bacteria.

[0044] This nucleotide sequence (i) can encodes any type III secretion system known in the art, such as from Yersinia, Salmonella, Bordetella, Pseudomonas, Chlamydia, Burkholderia, Escherichia, Shigella, Erwinia, Ralstonia, Xanthomonas and Rhizobium species. Examples include, but are not limited to, T3SS from Salmonella typhimurium, Salmonella enterica, Pseudomonas aeruginosa, Yersinia pseudotuberculosis, Yersinia pestis, Shigella flexneri, Citrobacter rodentium, Escherichia coli EHEC, Escherichia coli EPEC, Pseudomona syringae, Ralstonia solanacearum, Xanthomonas campestris, and Erwinia amylovora. Structural proteins forming the T3SS do not have to all origin from the same pathogen. Because type III secretion systems are highly conserved in a variety of gram- negative pathogenic bacteria, it is possible that the structural proteins are derived from different pathogens.

[0045] The nucleotide sequence (i) is preferably integrated into the genome of said cell. It may be present in in the chromosome and/or on a plasmid or vector in the cell. Preferably, the nucleotide sequence (i) is integrated into the chromosome of said cell. In other embodiments, the nucleotide sequence can be integrated in a plasmid or vector.

[0046] The term "nucleotide sequence" as used herein refers to either DNA or RNA. "Nucleic acid sequence" or "polynucleotide sequence" refers to a single or double-stranded polymer of deoxyribonucleotide or ribonucleotide bases read from the 5' to the 3' end. It includes both self -replicating plasmids, infectious polymers of DNA or RNA, and nonfunctional DNA or RNA.

[0047] For the purpose of the present invention, nucleotide encoding the T3SS comprises genes which encode structural proteins for the needle monomer, the inner rod of the needle, the ring proteins, two translocators, the needle-tip protein, the ruler protein, the ATPase, the export apparatus proteins and the sorting platform. A skilled person in the art can readily determine the nucleotides encoding such genes as they are known in the art and conserved between various T3SS systems (see for example, Kosarewicz et al., "The blueprint of the type-3 injectisome," Philos Trans R Soc Lond B Biol Sci. 367(1592):1140-54 (2012)). For example, the following proteins encode a T3SS in salmonella (listed with SwissProt accession number):

Salmonella size [kDa] function Acc. Nr.

InvA 76.1 Export Apparatus P0A1 I3

InvC 47,6 ATPase P0A1 B9

InvG 61 .8 Outer Ring and Neck P35672

InvJ 36.4 Needle Length Control P40613

OrgA 22.8 Sorting Platform P0CL44

OrgB 26.5 Sorting Platform P0CL45

PrgH 44.5 Inner Rings P41783

Prgl 8,9 Needle Filament P41784

PrgJ 1 0.9 Inner Rod P41785

PrgK 28.2 Inner Rings P41786

SpaO 33.8 Sorting Platform P40699

SpaP 25.2 Export Apparatus P40700

SpaQ 9.4 Export Apparatus P0A1 L7

SpaR 28.5 Export Apparatus P40701

SpaS 40.1 Export Apparatus P40702

Table 1 : structural proteins of a T3SS system in Salmonella Salmonella Pathogenicity Island 1 (SPI-1 ).

[0048] Nucleotide sequence (i) encoding T3SS may be obtained from T3SS genetic locus in a pathogen which expressed T3SS. In one preferred embodiment, the nucleotide sequence (i) encoding the T3SS is derived from Salmonella Pathogenicity Island 2 (SPI-2) or more preferably from Salmonella Pathogenicity Island 1 (SPI-1 ) in Salmonella. T3SS encoded by SPI-1 is decribed herein in more detail. It is an approximately 40 kilobase (kb) gene segment, which is found in all organisms of the genus Salmonella and acquired by a lateral gene transfer event early in Salmonella evolution. Review of SPI-1 can be found in Herbert et al., "Pathogenicity islands in bacterial pathogenesis," Clin Microbiol Rev 17(1 ): 14-56 (2004) and Thomoson et al., "Comparative genome analysis of Salmonella enteritidis pt4 and Salmonella gallinarum 287/91 provides insights into evolutionary and host adaptation pathways," Genome Res 18(10):1624-37 (2008). SPI-1 encodes all structural genes required for a T3SS in three operons: PhoP repressed genes (prg operon) (Miller et al., "The phop virulence regulon and live oral salmonella vaccines" Vaccine 1 1 (2): 122-5 (1993)), surface presentation of antigens proteins (spa operon) and Invasion proteins (inv operon) (Groiman et al., "Cognate gene clusters govern invasion of host epithelial cells by Salmonella typhimurium and Shigella flexnerf EMBO J 12(10):3779-87 (1993)).

[0049] Additionally, a couple of T3SS effector proteins and their cognate chaperones are encoded in the Salmonella invasion proteins (sip operon) (Kaniga et al., "Identification of two targets of the type iii protein secretion system encoded by the inv and spa loci of Salmonella typhimurium that have homology to the Shigella ipad and ipaa proteins." J Bacteriol 177(24):7078-85 (1995)). The last operon in SPI-1 is the hyperinvasion locus (hil), which encodes the regulatory transcription factors for the T3SS (Bajaj et al., "hila is a novel ompr/toxr family member that activates the expression of salmonella typhimurium invasion genes," Mol Microbiol 18(4):715-27 (1995)). Environmental conditions such as growth phase, pH, oxygen tension, and osmolarity regulate expression of hilA, the positive regulator for the whole SPI-1 T3SS regulon. It upregulates the inv, prg and sic operon leading to the overexpression of the structural components of the needle complex and indirectly to the increased expression of the natural effector proteins. For the purpose of the present invention, it is also possible to use alternative transcription regulators/promoters for heterologous expression of T3SS structural proteins in a cell.

(ii) a nucleotide sequence encoding a protein comprising an N-terminal secretion tag recognized by T3SS, and the therapeutic protein.

[0050] T3SS machinery recognizes signals present in the effector proteins which are typically referred to as type III signal sequences. In general, the signal sequences tend to be located near the N-terminal end of the effector protein and form the first 15-30 amino acids of the effector protein. [0051 ] To secrete the therapeutic protein via T3SS, the cell according to the present invention comprises (ii) a nucleotide sequence encoding a protein comprising an N-terminal secretion tag and the therapeutic protein. The N-terminal secretion tag is required for the T3SS to recognize and secrete the therapeutic protein. The N-terminal secretion tag can be derived, for example, from the N-terminal end of any type III secretion effectors (the type III signal sequence), because the effectors are readily recognized by T3SS in natural context. These secretion tags have been studied and are known in the art. They can also be identified by examining sequence similarity shared between effectors (Lower et al., "Prediction of Type III Secretion Signals in Genomes of Gram-Negative Bacteria," PLoS One 4(6): e5917 (2009)). Use of secretion tags is known and has been described, for example, in WO2000/059537, WO2008/110653 and WO2012/012605. It is within the skill person of the art to determine, for example, by way of secretion assay, whether any given N-terminal secretion tag is recognized by T3SS. For example, structural proteins of salmonella is listed in Table 2 where. SwissProt Accession Number and functions are also listed. name Chap AA size [kDa] acc. nr. function

Sip A InvB 685 73.9 P0CL52 invasion, inflammation

SipB SicA 593 62.5 Q 56 1 9 translocation, apoptosis of macrophages

Si C SicA 409 43.0 P0CL47 translocation, invasion

SipD - 343 37.1 Q56026 translocation

SopA InvB 782 86.8 Q8ZNR3 inflammation, escape from SCV

SopB PipC 561 61.9 030916 invasion, inflammation

SopD - 317 36.1 P40722 membrane fission

SopE InvB 240 26.6 052623 invasion

SptP SicP 503 60.0 P74873 restoring of cytoskeleton

AvrA - 301 33.6 030621 reduction of inflammation

Table 2: Effector Proteins and cognate chaperon proteins in Salmonella.

Fig. 9 provides an overview of the signal sequences, chaperone binding domain and effector domains of the Salmonella effector proteins. [0052] In addition to the signal sequence, some effector proteins depend on their interaction with specific T3SS chaperones for their secretion, known as class I chaperones. Thus, the nucleotide sequence (ii) of the present invention can as an optional feature additionally comprise a chaperone binding domain (CBD) after the N-terminal secretion tag for engaging the chaperons. These chaperone-binding domains have been studied in the art and are readily identifiable. In natural effector proteins, the chaperone binding domains of class I chaperones are generally located immediately after the N-terminal signal sequence and are usually approximately 50-100 amino acid long. For the purpose of the present invention, it is preferred that the nucleotide sequence (ii) includes a chaperone binding domain. However, in other embodiments, the sequence (ii) does not include the chaperone binding domain.

[0053] Preferably, the nucleotide sequence (ii) is obtained from the first 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 1 10, 120, 130, 140 or 150 amino acids of a T3SS effector protein or homologues thereof.

[0054] In preferred embodiments, the N-terminal secretion tag is derived from SipA, SipB, SipC, SipD, InvJ, SpaO, AvrA, SopE2 and SptP proteins of Salmonella, the YopE, YopH, YopM and YpkA proteins of Yersinia, the I pa proteins of Shigella, or the ExoS proteins of Pseudomonas aeruginosa.

[0055] More preferably, the N-terminal secretion tag is obtained from SipA, SipB, SipC, SipD, SopA, SopB, SopD, AvrA, SopE2 and most preferably from SptP. These effector proteins are described in detail in the later sections of the present specification. In the examples provided in the present application, the inventors used the first 36 amino acid residues at the N-terminal end of SptP (as shown in SEQ ID NO: 1 in Fig 8) to direct secretion via Salmonella SPI-1 T3SS, although shorter sequence can be used, such as at least the first 30, 31 , 32, 33, 34 and 35 amino acid residues of SEQ ID NO: 1 . In a preferred embodiment, the N-terminal secretion tag comprises an amino acid sequence at least 90%, such as at least 91 , 92, 93, 94, 95, 96, 97, 98, 99 and 100% to the amino acid sequence as shown in SEQ ID NO: 1.

(iii) a nucleotide sequence encoding a chaperone capable of binding to said chaperone binding domain

[0056] The cell of the present invention further comprises, as an optional feature, a nucleotide sequence (iii) which encodes a chaperone which is capable of binding to the chaperone binding domain encoded by the nucleotide sequence (ii). Chaperones which bind to the chaperone binding domain of effector proteins have been studied. There is a considerable body of literature describing chaperone structure and how they interact with their partner proteins (e.g., see Feldman et al., "The multitalented type III chaperones: all you can do with 15 kDa," FEMS Microbiol Lett 219, 151-158 (2003); Ghosh, "Process of protein transport by the type III secretion system," Microbiol Mol Biol Rev 68, 771-795 (2004); Parsot et al., "The various and varying roles of specific chaperones in type III secretion systems," Curr Opin Microbiol 6, 7-14 (2003); "Wattiau et al., "SycE, a chaperone-like protein of Yersinia enterocolitica involved in the secretion of YopE," Mol Microbiol 8, 123-131 (1993); Akeda et al., "Identification and characterization of a type III secretion-associated chaperone in the type III secretion system 1 of Vibrio parahaemolyticus," FEMS Microbiology Letters 296, 18-25 (2009); Page et al., "Chaperones of the type III secretion pathway: jacks of all trades," Molecular Microbiology 46, 1-1 1 (2002)). The secretion of T3SS effectors generally becomes more efficient when the chaperones are present. It is thus preferred that the cell expresses chaperones which are capable of binding to the chaperone binding domain to increase the secretion efficiency of the protein.

[0057] The chaperone binding domain encoded by nucleotide sequence (ii) is preferably selected from the chaperone binding domain in SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP, AvrA. The chaperone encoded by nucleotide sequence (iii) is preferably selected from one of the following the Salmonella chaperone protein SicA, InvB, SicP, SigE, InvB, SicP, SigE, SicA. [0058] More preferably, one of the following N-terminal secretion tag and chaperone binding domain/chaperones is encoded by the nucleotides in the cell: SipA/lnvB, SipC/SicA, SopA lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE. This means, for example in case of SipA lnvB, the N-terminal secretion tag and chaperone binding domain is that from SipA, and the chaperone protein is InvB. In the example the chaperone binding domain shown in Fig. 8 (SEQ ID NO: 3) was used. In a preferred embodiment, the N-terminal secretion tag comprises an amino acid sequence at least 90%, such as at least 91 , 92, 93, 94, 95, 96, 97, 98, 99 and 100% to the amino acid sequence as shown in SEQ ID NO: 3.

Promoter

[0059] The expression of the nucleotide sequences described herein (any one of nucleotide sequences (i) (ii) and (iii)) can be driven by a promoter. The term "promoter" as used herein refers to a region that facilitates the transcription of a particular gene. A promoter is preferably operatively linked to the adjacent nucleotide sequence which is to be expressed. A promoter typically increases the amount of recombinant product expressed from a nucleotide sequence as compared to the amount of the expressed recombinant product when no promoter exists. A promoter from one organism can be utilized to enhance recombinant product expression from a sequence that originates from another organism. In addition, one promoter element can increase the amount of products expressed for multiple sequences attached in tandem. Hence, one promoter element can enhance the expression of one or more recombinant products. Multiple promoter elements are well known to persons of ordinary skill in the art. The expression of the nucleotide sequences described herein can be driven by a "constitutive" or "inducible" promoter. The term "expression" as used herein refers to the transcription and stable accumulation of mRNA from a given nucleotide sequence. The promoter could be a "inducible promoter" or "constitutive promoter." "Inducible promoter" refers to a promoter which can be induced by the presence or absence of certain factors, and "constitutive promoter" refers to an unregulated promoter that allows for continuous transcription of its associated gene. In a preferred embodiment, both the nucleotide sequences (i) and (ii) is driven by an inducible promoter. Upon induction of the promoter, the T3SS and the therapeutic protein with the secretion tag are expressed.

[0060] In preferred embodiments the nucleotide sequence (ii) and (iii) are contained in a plasmid or a bacterial artificial chromosome (BAC). As used herein, the term "bacterial artificial chromosome BAC" refers to a cloning vector derived from a bacterial chromosome, into which large DNA sequences from bacterial or nonbacterial sources can be inserted. The term "plasmid" refers to an autonomous circular DNA molecule capable of replication in a cell. Most plasmids exist in only one copy per bacterial cell. Some plasmids, however, exist in higher copy numbers. For example, the plasmid ColE1 typically exists in 10 to 20 plasmid copies per chromosome in E. coli. If the nucleotide sequences of the present invention are contained in a plasmid, the plasmid preferably has a copy number of 20-30, 30-100 or more per host cell. With a high copy number of plasmids, it is possible to increase the amount therapeutic proteins expressed by the cell. Large numbers of suitable plasmids or vectors are known to those of skill in the art and many are commercially available. Examples of suitable vectors are provided in Sambrook et al, eds., Molecular Cloning: A Laboratory Manual (2^nd Ed.), Vols. 1-3, Cold Spring Harbor Laboratory (1989), and Ausubel et al, eds., Current Protocols in Molecular Biology, John Wiley & Sons, Inc., New York (1997).

[0061 ] Embodiments in this connection include that (1 ) nucleotide sequence (ii) and (iii) are contained in a plasmid, (2) nucleotide sequence (ii) and (iii) are contained in an artificial bacterial chromosome, (3) nucleotide sequence (ii) is contained in a plasmid and nucleotide sequence (iii) is contained in an artificial bacterial chromosome, or (4) nucleotide sequence (iii) is contained in a plasmid and nucleotide sequence (ii) is contained in an artificial bacterial chromosome.

[0062] In another preferred embodiment, nucleotide sequence (ii) and (iii) are co- expressed from the same plasmid. "Co-expression" as used herein refers to the expression of two or more nucleic acid sequences at the same time. Therapeutic protein

[0063] As used herein, a "therapeutic protein" means any polypeptide, protein, protein variant, fusion protein and/or fragment thereof which may be administered to a mammal as a medicament. It is envisioned but not required that therapeutic protein according to the present invention is heterologous to the cell. Examples of proteins that can be produced by the cell of the present invention are, without limitation, enzymes, regulatory proteins, receptors, peptides, e.g. peptide hormones, growth factors, cytokines, structural proteins, lymphokines, adhesion molecules, receptors, membrane or transport proteins, and any other polypeptides that can serve as agonists or antagonists and/or have therapeutic or diagnostic use. Moreover, the proteins of interest may be antigens as used for vaccination, vaccines, antigen-binding proteins, immune stimulatory proteins. It may also be an antigen-binding fragment of an antibody, which can include any suitable antigen-binding antibody fragment known in the art. For example, an antibody fragment may include but not limited to Fv (a molecule comprising the VL and VH), single-chain Fv (scFV) (a molecule comprising the VL and VH connected with by peptide linker), Fab, Fab', F(ab')2, single domain antibody (sdAb) (molecules comprising a single variable domain and 3 CD ), and multivalent presentations thereof. The antibody or fragments thereof may be murine, human, humanized or chimeric antibody or fragments thereof. Examples of therapeutic proteins include an antibody, polyclonal antibody, monoclonal antibody, recombinant antibody, antibody fragment, Fab', F(ab')₂, Fv, scFv, di-scFvs, bi-scFvs, tandem scFvs, bispecific tandem scFvs, sdAb, V_H, V_L, humanized antibody, chimeric antibody, IgA antibody, IgD antibody, IgE antibody, IgG antibody, IgM antibody, intrabody, minibody or monobody. A "therapeutic protein" in the sense of the present invention is preferably not an effector protein of T3SS or mutants thereof.

[0064] Such therapeutic proteins include, but are not limited to, insulin, insulin-like growth factor, hGH, tPA, cytokines, e.g. interleukines such as IL-1 , IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-1 1 , IL-12, IL-13, IL-14, IL-15, IL-16, IL-17, IL-18, interferon (IFN) alpha, IFN beta, IFN gamma, IFN omega or IFN tau, tumor necrosisfactor (TNF) TNF alpha and TNF beta, TRAIL; G-CSF, GM-CSF, M-CSF, MCP-1 and VEGF.

[0065] The therapeutic protein have a length of more than 629 amino acid residues via T3SS, such as more than 630, 631 , 632, 633, 634, 635, 636, 637, 638, 639, 640, 650, 660, 670, 680, 690, 700, 710, 720, 730, 740, 750, 760, 770, 780, 790, 800, 810, 820, 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1 100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000 amino acid residues or more.

[0066] The complete polypeptide (including the secretion tag and any other non- therapeutic sequences forming a single polypeptide chain) may have a length of more than 864 amino acid residues via T3SS, such as more than 865, 866, 867, 868, 869, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000 amino acid residues or more.

[0067] Particularly preferred therapeutic proteins are antibodies or fragments thereof which are has a length of more than 629 amino acids.

[0068] In preferred embodiments, the therapeutic protein is an antibody, nanobody or monobody. Examples of antibodies that can be used in the invention include chimeric antibodies, non-human antibodies, human antibodies, humanized antibodies, and domain antibodies (dAbs). A "nanobody" is a small functional antibody fragment composed of the VHH domain of Camelidae heavy chain antibodies. They are homodimers of two heavy chains which are missing the first constant domain (CH1 ) due to a splice site mutation. The variable domain of these antibodies(denoted as VHH) binds to epitopes with a comparable affinity to conventional antibodies. A "monobody" refers to an artificial or synthetic single domain antibody or antibody mimic. Also known as an ADNECTIN™ (e.g., see U.S. Patent No. 7,1 15,396), monobodies are genetically engineered proteins that can bind to antigens. They are based on the structure of human fibronectin, more specifically on its tenth extracellular type III domain.

[0069] Preferably, the expressed therapeutic proteins do not have basic amino acid stretches, oligomerization interfaces or a stable tertiary structure. If therapeutic protein contains basic amino acid stretches (e.g., His, Arg or Lys) of more than 6 amino acid residues in length, the basic stretch might form a positively-charged domain which interacts with negatively charged cell components like nucleic acids or the surface of cell membrane or nuclear membrane. An example of basic amino acid stretch is a nuclear localisation signal.

[0070] It is also preferred that the therapeutic protein does not contain oligomerization interfaces such as a dimerization interface like leucine zipper because oligomerization may abolish secretion upon oligomerization. Exemplary oligomerization domains include, but are not limited to, coiled-coil domains, alpha- helical coiled-coil domains, collagen domains, collagen-like domains. Further examples include helix-loop-helix domains and

[0071 ] Furthermore, it is desirable that the therapeutic protein does not have a stable tertiary structure. As used herein, the "term "stable tertiary structure" refers to a tertiary structure that cannot be unfolded by the ATPase of T3SS. For example, the green fluorescent protein (GFP) and ubiquitin is known to have a very compact and stable tertiary structure which cannot be unfolded by the ATPase. A skilled person is able to determine whether a given therapeutic protein has a stable tertiary structure for the purpose of the present invention.

[0072] The therapeutic protein encoded by the nucleotide sequence (ii) can be in the form of a concatemer. A "concatemer" as used herein refers to a long continuous nucleic molecule that contains multiple copies of the same nucleic acid sequence, such as 2, 3, 4, 5 or more, linked in series. [0073] In a preferred embodiment, the chaperone binding domain and the therapeutic protein are separated by a linker. The term "linker" refers to an innocuous length of nucleotide sequences or protein that joins two nucleotide sequences or proteins.

Protease cleavage site

[0074] In a preferred embodiment, the therapeutic protein can be cleaved off from the secretion tag and the optional chaperone binding domain by a protease recognizing a protease cleavage site preceding the therapeutic protein. It may be desirable that the secretion tag and chaperon binding domain are cleaved after the protein is secreted via T3SS: A protease cleavage site is a specific amino acid sequence recognized by the protease for proteolytic cleavage. Many protease cleavage sites are known in the art (see, e.g., Matayoshi et al., Science 247: 954 (1990); Dunn et al., Meth Enzymol 241 : 254 (1994); Seidah et al., Meth Enzymol 244: 175 (1994); Thomberry, Meth Enzymol 244: 615 (1994); Weber et al., Meth Enzymol 244: 595 (1994); Smith et al., Meth Enzymol 244: 412 (1994); and Bouvier et al., Meth Enzymol 248: 614 (1995)). The protease cleavage site can also be contained within the linker.

[0075] If multiple therapeutic proteins are present in the form of concatemers, each copy of the therapeutic protein can be cleaved off by a protease recognizing a protease cleavage site preceding each copy of the therapeutic protein.

Pharmaceutical and diagnostic composition

[0076] The present cells can be used to generate of one or several therapeutic proteins for diagnostic purposes, research purposes or manufacturing of therapeutic proteins either on the market or in clinical development. Provided herein is a pharmaceutical or diagnostic composition comprising the non-pathogenic cell of the present invention.

[0077] For therapeutic use, the nonpathogenic cell or composition may be administered in a therapeutically effective amount in any conventional dosage form in any conventional manner to treat a disease. Routes of administration include, but are not limited to, intravenously, intramuscularly, subcutaneously, intrasynovially, by infusion, sublingually, transdermally, orally, topically or by inhalation, tablet, capsule, caplet, liquid, solution, suspension, emulsion, lozenges, syrup, reconstitutable powder, granule, suppository and transdermal patch. Methods for preparing such dosage forms are known (see, for example, Ansel et al., Pharmaceutical Dosage Forms and Drug Delivery Systems 5th ed., Lea and Febiger (1990)). A therapeutically effective amount can be determined by a skilled artisan based on factors such as weight, metabolism, and severity of the affliction etc. Preferably, the active compound is dosed at about 1 mg to about 500 mg per kilogram of body weight on a daily basis. More preferably the active compound is dosed at about 1 mg to about 100 mg per kilogram of body weight on a daily basis.

[0078] The composition may be administered alone or in combination with adjuvants to enhance the stability of the therapeutic proteins, to facilitate administration, to provide increased dissolution or dispersion, to increase the activity, to provide adjunct therapy and the like. Advantageously, combination with adjuvants may utilize lower dosages of the active ingredient, thus reducing possible toxicity and adverse side effects.

Production of proteins

[0079] This invention allows the cell to be used as production factories of recombinant proteins that can be secreted into the extracellular medium without the need for cell lysis. The present invention thus provides nonpathogenic cell used for the production and secretion of a therapeutic protein. The method for the production of the protein comprises culturing the non-pathogenic cell of the present invention and harvesting said therapeutic protein from the culture. Another possible application of the cells is for the delivery of therapeutic proteins into the target eukaryotic cells.

[0080] The cells should be cultured under conditions suitable for the expression of the aforementioned nucleotide sequences. Technique of cell culture is well-established in the art. The therapeutic protein is then secreted via T3SS into the culture medium which can then be harvested. The therapeutic proteins in the medium can be optionally treated by protease to cleave secretion tag off. It may be desirable to purify the protein to obtain substantially homogenous preparations of the protein. In general, methods which are routinely applied in the expression of recombinant proteins in a host cell can be employed.

[0081 ] The current invention also provides in another aspect use of the present cells to deliver therapeutic protein into eukaryotic cells. The cells can be brought into contact with eukaryotic cells in vivo or in vitro under suitable conditions for the therapeutic protein to be injected into host cell cytoplasm via T3SS.

Partial secretion

[0082] In another aspect, the present invention relates to the display of proteins on the cell surface. Bacterial surface display had been achieved using chimeric genes derived from bacterial outer membrane proteins, lipoproteins, fimbria proteins, and flagellar proteins. However, the present invention provides yet a novel means of surface display achieved by a partial secretion of a given protein via T3SS. It is generally considered that the expression of heterologous proteins on the surface of cells provides a powerful tool for diverse activities as obtaining specific antibodies, determining enzyme specificity, exploring protein-protein interactions, and introducing new functions into proteins. One advantage of the present invention is the ability to display large proteins.

[0083] For partial secretion, the present invention provides one or more cell comprising a type III secretion system and a protein comprising from N to C-terminus, (1 ) an N-terminal secretion tag recognized by said T3SS, (2) optionally a chaperone binding domain, (3) a protein of interest, (4) optionally a sequence for spanning the secretion path of T3SS (a "spanning sequence") having a length in amino acids that has at least the length of at least three SptP effector domains, and (5) an "anchoring sequence" which enables partial secretion of the protein. For surface display it is envisaged that the spanning sequence can be used so to enable the protein of interest to pass through T3SS and enter the extracellular environment. Thus, the present invention provides a T3SS comprising a partially secreted protein comprising an anchoring sequence at the C-terminus of the protein.

[0084] Preferably, the cell comprises a type III secretion system (T3SS) and a protein comprising, from N to C-terminus, (1 ) an N-terminal secretion tag recognized by said T3SS, (2) optionally a chaperone binding domain, (3) a protein of interest, (4) as the "spanning sequence" one or more effector domains of a T3SS effector protein having a length in amino acids that has at least the length of at least three SptP effector domains, and (5) an anchoring sequence encoding for example a protein having a globular structure or zinc finger domain. Such cells can be advantageously used to construct polypeptide display libraries comprising a plurality of expressed proteins of interest. Examples of cell display libraries include phage displayed peptide libraries, bacterial surface displayed polypeptides and monoclonal antibody libraries. The protein of interest could be for example an antibody, polyclonal antibody, monoclonal antibody, recombinant antibody, antibody fragment, Fab', F(ab')₂, Fv, scFv, di-scFvs, bi-scFvs, tandem scFvs, bispecific tandem scFvs, sdAb, V_H, V_L, humanized antibody, chimeric antibody, IgA antibody, IgD antibody, IgE antibody, IgG antibody, IgM antibody, intrabody, minibody or monobody. Examples of proteins that can be produced by the cell of the present invention are any polypeptides, with or without biological functions, including enzymes, regulatory proteins, receptors, peptides, e.g. peptide hormones, growth factors, cytokines, structural proteins, lymphokines, adhesion molecules, receptors, membrane or transport proteins, and any other polypeptides that can serve as agonists or antagonists and/or have therapeutic or diagnostic use. Moreover, the proteins of interest may be antigens as used for vaccination, vaccines, antigen-binding proteins, immune stimulatory proteins. It may also be an antigen-binding fragment of an antibody, which can include any suitable antigen-binding antibody fragment known in the art. For example, an antibody fragment may include but not limited to Fv (a molecule comprising the VL and VH), single- chain Fv (scFV) (a molecule comprising the VL and VH connected with by peptide linker), Fab, Fab', F(ab')2, single domain antibody (sdAb) (molecules comprising a single variable domain and 3 CDR), and multivalent presentations thereof. The antibody or fragments thereof may be murine, human, humanized or chimeric antibody or fragments thereof. Examples of therapeutic proteins include an antibody, polyclonal antibody, monoclonal antibody, recombinant antibody, antibody fragment, Fab', F(ab')₂, Fv, scFv, di-scFvs, bi-scFvs, tandem scFvs, bispecific tandem scFvs, sdAb, V_H, V_L, humanized antibody, chimeric antibody, IgA antibody, IgD antibody, IgE antibody, IgG antibody, IgM antibody, intrabody, minibody or monobody. Further proteins of interest include, but are not limited to, insulin, insulin-like growth factor, hGH, tPA, cytokines, e.g. interleukines such as IL-1 , IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-11 , IL-12, IL-13, IL-14, IL-15, IL-16, IL-17, IL-18, interferon (IFN) alpha, IFN beta, IFN gamma, IFN omega or IFN tau, tumor necrosisfactor (TNF) TNF alpha and TNF beta, TRAIL; G-CSF, GM-CSF, M-CSF, MCP-1 and VEGF as well as homologues thereof.

[0085] The display system mainly involves (1 ) a N-terminal secretion tag recognized by T3SS, (4) one or more effector domains of a T3SS effector protein having a certain length for secretion, and (5) an anchoring sequence which anchors the recombinant proteins onto the surface of transfected cells.

[0086] (1 ) N-terminal secretion tag recognized by said T3SS and the optional (2) chaperone binding domain as well as other characterization of the cells have been described earlier in the present specification and are applicable for constructing cells for surface display. However, the (3) protein of interest to be displayed is not limited to therapeutic proteins. Since cell display libraries are particularly useful for generating protein-specific affinity reagents for therapeutics and drug discovery, any type of proteins or polypeptides can be displayed by the cells according to the present invention.

[0087] The surprising discovery of the inventors has furthermore made it possible for the first time to partially secrete large proteins via T3SS. The present invention enables a method for displaying large polypeptides in a cell surface display system. [0088] As mentioned earlier, the T3SS comprised in the cell is preferably from Salmonella typhimurium, Salmonella enterica, Pseudomonas aeruginosa, Yersinia pseudotuberculosis, Yersinia pestis, Shigella flexneri, Citrobacter rodentium, Escherichia coli EHEC, Escherichia coli EPEC, Pseudomona syringae, Ralstonia solanacearum, Xanthomonas campestris, and Erwinia amylovora. The N-terminal secretion tag is preferably obtained from SipA, SipB, SipC, SipD, SopA, SopB, SopD, AvrA, SopE2 and most preferably from SptP. The chaperone binding domain is preferably from one of SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP, AvrA. The cell may optionally comprise a chaperone selected from SicA, InvB, SicP, SigE, InvB, SicP, SigE, SicA. More preferably, one of the following N-terminal secretion tag and chaperone binding domain/chaperones is encoded by the nucleotides in the cell: SipA/lnvB, SipC/SicA, SopA lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE. If present, the chaperone binding domain and the protein of interest are preferably separated by a linker. Also, the protein of interest can be cleaved off from the secretion tag and the optional chaperone binding domain by a protease recognizing a protease cleavage site preceding the protein of interest. Overview of the signal sequences, chaperone binding domain and effector domains of various Salmonella effector proteins is provided in Fig. 9.

[0089] In a particularly preferred embodiment, the cell comprises a type III secretion system (T3SS) and a protein comprising from N- to C-terminus an N-terminal secretion tag recognized by said T3SS, a chaperone binding domain, a protein of interest, three Spt effector domains from Salmonella, and a protein having a globular structure or zinc finger domain. Furthermore the secretion tag and chaperone binding domain can optionally be selected from SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP and AvrA from Salmonella.

[0090] In another embodiment, the cell comprises a type III secretion system (T3SS) and a protein comprising from N- to C-terminus an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain, a protein of interest, three Spt effector domains from Salmonella and a GFP or ubiquitin. Furthermore, the cell can optionally comprise a chaperone selected from SicA, InvB, SicP, SigE, InvB, SicP, SigE, SicA.

[0091 ] In yet another embodiment, the cell comprises a type III secretion system (T3SS) and a protein comprising from N- to C-terminus an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain, a protein of interest, one or more effector domains of a T3SS effector protein having a length in amino acids that has at least the length of at least three SptP effector domains and a protein having a globular structure or zinc finger domain, wherein the following N-terminal secretion tag and chaperone binding domain/chaperones are applied SipA lnvB, SipC/SicA, SopA lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE.

[0092] In all of the embodiments above, T3SS is preferably a salmonella T3SS derived from the Salmonella Pathogenicity Island 1 (SPI-1 ) locus. Also, the protein of interest can be cleaved off from the secretion tag, the optional chaperone binding domain, and/or the effector domains by a protease recognizing a protease cleavage site preceding the protein of interest and/or following. Successful cleavage will allow the protein of interest to be delivered into a host cell such as eukaryotic cells via its T3SS.

(4) Spanning Sequence

[0093] The present cell surface display system is characterized in a spanning sequence comprising one or more effector domains have at least a given length to span the T3SS secretion path and expose the preceding protein of interest to the extracellular environment. The term "spanning sequence" is a stretch of amino acid sequence of the partially secreted protein that occupies the secretion path when partial secretion is attained. The spanning sequence can be composed of any sequence which does not have a stable structure which cannot be unfolded by ATPase of the T3SS system. A skilled person can readily determine when needed by way of assay whether a given spanning sequence is unfoldable. Preferably, the spanning sequence comprises one or more "effector domains," which refer to one or more of "complete, partial or homologous sequences" of the portion in a given effector protein because such sequences are transportable by T3SS. By homologous it is meant that the sequences are at least 5%, such as at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 95, 97, 98 and 99% homologous in comparison. In preferred embodiments the spanning sequence comprises at least 1 , 2, 3, 4 or 5 effector domains from any one of SipA, SipB, SipC, SipD, SopA, SopB, SopD, AvrA, SopE2 and SptP from Salmonella. The "effector domain" for the purpose of the present invention does not include the N terminal secretion sequence.

[0094] In preferred embodiment, the spanning sequence comprises 1 , 2, 3, 4, 5, 6, or 7 effector domains from SipA, SipB, SipC, SipD, SopA, SopB, SopD, AvrA, SopE2 and/or SptP. Sequences of these proteins are provided in Table 2 with reference to SwissProt accession numbers. Preferably, the spanning sequence comprises effector domains from one effector protein as repeats. However, it is also possible to combine effector domains from different effector proteins. Furthermore, it is not necessary that the "effector domain" retains its biological activity as an effector protein.

[0095] The inventors have compared and analyzed the structural differences between empty T3SS and T3SS with partially secreted substrate (also referred to as substrate- trapped needle complexes) and it was surprisingly found that this secretion path of the needle complex is longer than empty T3SS. In the examples provided in the present application, it has been shown that the path is about 800 A and at least three SptP effector domains is sufficient to span the secretion path of the T3SS needle complex (see Example 5- 7). In a preferred embodiment, the spanning sequence comprises one or more effector domains where each effector domain comprises an amino acid sequence which is at least 10, 20, 30, 40, 50, 60, 70, 80, 90 91 , 92, 93, 94, 95, 96, 97, 98, 99 and 100% homologous to SEQ ID NO: 2.

[0096] A variety of sequence based alignment methodologies, which are well known to those skilled in the art, are useful in determining homology among sequences. These include, but not limited to, the local identity/homology algorithm of Smith, F. and Waterman, M. S. (1981 ) Adv. Appl. Math. 2: 482-89, homology alignment algorithm of Peason, W. R. and Lipman, D. J. (1988) Proc. Natl. Acad. Sci. USA 85: 2444-48, Basic Local Alignment Search Tool (BLAST) described by Altschul, S. F. et al. (1990) J. Mol. Biol. 215: 403-10, or the Best Fit program described by Devereau, J. et al. (1984) Nucleic Acids. Res. 12: 387-95, and the FastA and TFASTA alignment programs, preferably using default settings or by inspection. In one preferred embodiment, homology is calculated by Fast alignment algorithms based upon the following parameters: mismatch penalty of 1.0; gap size penalty of 0.33, joining penalty of 30 (see "Current Methods in Comparison and Analysis" in Macromolecule Sequencing and Synthesis: Selected Methods and Applications, p. 127-149, Alan R. Liss, Inc., 1998). Another example of a useful algorithm is PILEUP. PILEUP creates multiple sequence alignment from a group of related sequences using progressive, pairwise alignments. It can also plot a tree showing the clustering relationships used to create the alignment. PILEUP uses a simplification of the progressive alignment method of Feng, D. F. and Doolittle, R. F. (1987) J. Mol. Evol. 25, 351-60, which is similar to the method described by Higgins, D. G. and Sharp, P. M. (1989) CABIOS 5: 151-3. Useful parameters include a default gap weight of 3.00, a default gap length weight of 0.10, and weighted end gaps. Another example of a useful algorithm is the family of BLAST alignment tools initial described by Altschul et al. (see also Karlin, S. et al. (1993) Proc. Natl. Acad. Sci. USA 90: 5873-87). A particularly useful BLAST program is WU-BLAST-2 program described in Altschul, S. F. et al. (1996) Methods Enzymol. 266: 460-80. WU-BLAST uses several search parameters, most of which are set to default values. The adjustable parameters are set with the following values: overlap span=1 , overlap fraction=0.125, word threshold (T)=1 1. The HSP S and HSP S2 parameters are dynamic values and are established by the program itself depending upon the composition of the particular sequence and composition of the particular database against which the sequence of interest is being searched; however, the values may be adjusted to increase sensitivity. An additional useful algorithm is gapped BLAST as reported by Altschul, S. F. et al. (1997) Nucleic Acids Res. 25: 3389-402. Gapped BLAST uses BLOSSOM-62 substitution scores; threshold parameter set to 9; the two-hit method to trigger ungapped extensions; charges gap lengths of k at cost of 10+k; Xu set to 16, and Xg set to 40 for database search stage and to 67 for the output stage of the algorithms. Gapped alignments are triggered by a score corresponding to -22 bits. Speific programs have been developed to may and assemble NGS data, e.g. the program BOWTIE.

[0097] Effector proteins SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, AvrA and SptP from Salmonella and described in detail below.

[0098] Salmonella invasion protein A (SipA) plays a vital role in the invasion process, via actin binding and in inducing inflammation. The C-terminal entity of SipA is important for F-actin bundling, inhibits actin depolymerization and potentiates the activity of SipC, thereby supporting the invasion process. The protein is processed by Caspase3 between residues 431 to 434, resulting in two fragments. The C-terminal fragment contains the actin interacting domains, whereas the N-terminal fragment is important for the inflammatory response. SipA induces CXC chemokine and IL- 8 expression by phosphorylation of JUN and p38MAPK and activates NF_KB via NOD1/NOD2 signaling.

[0099] SipB is one of the two hydrophobic translocases. The secretion sequence was found to be at residues 3 to 8 of the SipB protein and the chaperone binding domain is located around somewhere between residues 80 and 100 (see for example Kim et al., "Analysis of functional domains present in the N-terminus of the SipB protein", Microbiology 153, 2998-3008 (2007)). The translocase has two transmembrane helices in the C-terminal region, which are inserted into the host membrane. Upon insertion of the transmembrane helices a hydrophilic loop of SipB is located in the mammalian cytoplasm. SipB has been shown to induce apoptosis in macrophages in a caspasel -dependent and a caspasel - independent manner.

[00100] SipC is the second hydrophobic translocase and exhibits actin nucleation, actin bundling and translocation functions. (Kim et al., "Molecular characterization of the InvE regulator in the secretion of type III secretion translocases in Salmonella enterica serovar Typhimurium," Microbiology 159:446-461 (2013)). The protein forms homooligomers in the translocation pore and interacts with SipB. Its secretion depends on the first 120 amino acids of SipC, where the secretion signal and the chaperone binding domain are located. The translocation activity of SipC depends on its interaction with SipB, however the mechanism for membrane insertion remains elusive. SipB and SipC have been shown to form extracellular complexes during secretion.

[00101 ] Salmonella Invasion Protein D (SipD) is part of the tip complex and is denoted the hydrophilic translocase protein. It has two functional domains, comprising an N-terminal secretion signal and a C-terminal functional domain.

[00102] SopA has a 45 N-terminal domain containing the secretion signal and the chaperone binding domain (CBD). The C-terminal domain contains the HECT-like motif, composed of a substrate binding-helix, an extended central domain, the N-lobe and a globular C-terminal C-lobe. The mammalian ubiquitin system uses a triade of enzymes to transfer the small post-translational modification to proteins. The first enzyme is an E1 ubiquitin-activating enzyme, which transfers the protein to an E2 ubiquitin conjugating enzyme. From E2 the ubiquitin is transferred to an E3 ubiquitin-ligase, which subsequently transfers the modification to a lysine of the target protein. SopA is an E3- ubiquitin-protein ligase, which can transfer ubiquitin to target proteins. SopA uses UbcH5a, UbcH5c and UbcH7 as E2 ubiquitin-conjugating enzymes. These E2s have been shown to be involved in inflammation. SopA induces inflammation and polymorphonuclear transmigation, thereby promoting enteritidis.

[00103] The N-terminus of Salmonella outer protein B (SopB) is characterized by the presence of a secretion signal and the chaperone binding domain (CBD), which is followed by a guanine dissociation inhibitor (GDI) domain comprising residues 1 17 to 168. The C- terminal moiety carries two inositol-4-phosphatase domains and a very C-terminal synaptojanin-homologous region (residues 357 - 561 ). SopB has sequence homology to mammalian inositol polyphosphate 4-phosphatases and that recombinant SopB has inositol phosphate phosphatase activity in vitro. SopB mediates virulence by interdicting inositol phosphate signaling pathways.

[00104] Salmonella outer protein D (SopD) is a 317 amino acid polymer and has a weight of 36.141 kDa. SopD is involved in membrane fission and macropinosome formation. It is recruited to membranes in a SopB-dependent manner and thus these two effectors act cooperatively to encourage host-cell membrane internalization and sealing. Loss of SopD leads to delayed membrane fission kinetics, but has little impact on the uptake efficiency of Salmonella.

[00105] The N-terminal approximately 100 amino acids of the guanine nucleotide exchange factor SopE serve as secretion signal and chaperone binding domain (CBD) for the chaperone InvB, whereas the enzymatic activity is lying in the C-terminal part (from residues 78 to 240).

[00106] The Avirulence Protein A (AvrA) is a ubiquitin-like protein cystein protease and exhibits acetyltransferase activity. It enhances proliferation and inhibits inflammation by stabilizing β-catenin and inhibiting JNK. The latter one is a consequence of the acetylation of upstream activators (MKKs) on serines or threonines, which inhibits their activation. Additionally AvrA deubiquitinizes Wnts and therefore dampens inflammatory responses. Recently AvrA has been shown to be important for the establishment of chronic infections.

[00107] The Salmonella protein tyrosine phosphatase (SptP) has an N-terminal secretion signal followed by a chaperone binding domain of 139 residues, which is bound by Salmonella invasion chaperone P (SicP). The central domain of the protein has GTPase activating protein (GAP) function from residues 174 to 290, which induces small GTPases to hydrolyse their bound GTP and renders them inactive. The very C-terminal residues 340 to 543 have tyrosine-protein phosphatase (PTPase) activity.

[00108] SptP has GAP activity for Rac1 and Cdc42 and thereby antagonizes the effect of SopE, reconstituting the cytoskeleton. Additionally, the inactivation of the Rho GTPases leads to the downregulation of JNK phosphorylation, which is induced by activated ho GTPases and leads to the expression of inflammatory genes. As SptP can be detected in cells up to eight hours post infection, it is also involved in the formation of the intracellular niche Salmonella uses to survive (SCV). SptP interacts with the AAA+ ATPase Valosin- containing protein (VCP) and dephosphorylates it. This interaction is catalysed by the PTPase domain, but it also needs the GAP domain, to stabilize the interaction. SptP targets VCP on membranes, distinct from the SCV, dephosphorylates it, which in turn allows the interaction of VCP with adaptor proteins and cofactors leading to the formation of Sa/mone//a-induced filaments. The function of SptP is therefore important for the maintenance of the SCV and allows the acquisition of nutrients.

[00109] The complete polypeptide length preceding the anchoring sequence (including the secretion tag, protein of interest, effector domains and any other sequences) may have a length of more than 864 amino acid residues, such as more than 865, 866, 867, 868, 869, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1 100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000 residues or more.

Globular structure

[001 10] A protein having a globular structure is expressed C-terminally to the one or more effector domains for anchoring. Globular structure can be determined by X-ray crystallography or NMR spectroscopy. For example, GFP is a globular protein of 27 kDa with a diameter of about 35x50 A. Ubiquitin is another example of globular protein which is about 8 kDa and has a diameter of about 20x30 A. The globular structure should have a minimum diameter of at least 10 A, such as at least 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30 A, so it will anchor the sequence to the needle complex at the basal side of the complex. The term globular protein refers to proteins which has a generally round structure. [001 1 1 ] The present invention provides a method for identifying a cell secreting a protein of interest via T3SS comprising (a) providing one or more cells on which proteins are partially secreted via T3SS, ; and (b) identifying said protein of interest with a detection moiety which recognizes said protein of interest. Identification can be performed by using any known means in the field of cell surface display.

Zinc finger domain

[001 12] A zinc finger is a small protein structural motif that is characterized by the coordination of one or more zinc ions to form a stable compact fold. Zinc fingers coordinate zinc ions with a combination of cysteine and histidine residues. Classes of zinc fingers include Cys2His2, Gag knuckle, Treble clef, Zinc ribbon, Zinc ribbon, Zn2/Cys6, and TAZ2 domain like zinc fingers. The term "zinc finger domain" as used within the present invention refers to a protein domain that comprises a zinc ion and is capable of binding to a specific three basepair DNA sequence. Proteins comprising zinc fingers are known in the art.

Which cells for display

[001 13] Any cells can be used for cell surface displayed employing T3SS as described herein. Pathogens which inherently express T3SS can advantageously employed, although non-pathogenic cells may be preferred for reduced harmful effects due to the pathogenicity.

[001 14] Cells suitable for cell surface display of proteins of interest include Salmonella typhimurium, Bacillus subtilis, Pseudomonas aeruginosa, Vibrio cholerae, Klebsiella pneumonia, Neisseria gonorrhoeae, Neisseria meningitidis, Bacteroides nodosus, Moraxella bovis, and Escherichia coli. Examples

Example 1

[001 15] We designed a series of substrates to demonstrate properties allowing or precluding entry into and secretion from the needle complex. Overview of the design can be seen in Set-1 of Fig. 1 a. We used tandem-repeat-design of substrates based on the natural effector protein SptP. We fused one, three, or five tandemly repeated effector domains from the natural substrate SptP (SptP1 , SptP3, SptP5) downstream of a secretion signal.

[001 16] Secretion tests were performed in the Salmonella typhimurium strain SB905 monitoring the presence of designed substrates and also the native substrates SptP and SipA in the cell culture medium.

[001 17] In more detail, the non-flagellated Salmonella typhimurium strain SB905 was derived from SJW2941 as described in Sukhan et al. J Bacterid 183, 1 159-1 167 (2001 ). This strain carried the plasmid pSB3291 (AmpR) expressing the transcriptional regulator hilA under the araBAD promoter for overexpression of needle complex proteins.

[001 18] Substrates were based on SptP, containing an N-terminal signal (35 amino acids), a chaperone binding domain and one or multiples of the effector domain organized as tandem repeats (SptP1 , SptP3, SptP5). Sequences for 3x-FLAG-tags were introduced at the C-terminal end of the proteins. To maintain a balanced ratio of designed substrates and the chaperone SicP, co-expression was carried out by cloning both constructs into a single plasmid (pACYCDuet-1 (CmR) (Merck Chemicals Ltd.).

[001 19] Secreted proteins were detected directly from bacterial cell culture supernatants using polyclonal rabbit antibodies raised against domains of SptP or SipA. Briefly, seed cultures were grown overnight at 37 °C in LB medium supplemented with 0.3 M NaCI and antibiotic selection. Cultures were diluted 1 :10 in 50 mL of the same medium without antibiotics and the expression of hilA was induced by adding arabinose to a final concentration of 0.012% (w/v) for another growth period of 4 hours to allow for the assembly of functional injectisomes. Expression of designed substrates was induced by addition of 1 mM IPTG and cultures were subsequently incubated for another 2 hours. Harvested cultures were adjusted to an of OD (600nm) of 1.0, and cells were separated from supernatant by centrifugation at 5000xg for 15 minutes (Multifuge3,S- , Heraeus, Thermo).

[00120] Cell culture supernatants were separated from pelleted cells and filtered (pore size 0.2 μηι). The cell pellets were resuspended in 5 mL 1x PBS (phosphate buffered saline), diluted 1 :1 with 87% (v/v) glycerol and both samples were directly used for protein separation on 4-20% gradient SDS-polyacrylamid gels followed by Western blotting using rabbit anti needle complex, rabbit anti SptP, rabbit anti SipA, mouse anti FLAG mAb (Sigma) (all 1 :10000) and anti rabbbit and anti mouse secondary antibodies (1 :10000) conjugated with horseradish peroxidase for chemiluminescent detection (Sigma).

[00121 ] Referring to Fig. 1 b upper panel under secretion, strains transformed with plasmids encoding for SptP1 (theoretical molecular weight 64.9 kDa), SptP3 (149.2 kDa), or SptP5 (232.5 kDa) substrates (set-1 ) were able to secrete all three designed substrates (left). While full length SptP1 and SptP3 could also be detected using both anti SptP and anti FLAG-tag antibodies, detection of SptP5 was possible using anti SptP antibodies, but not anti FLAG-tag.

[00122] Natural, wild-type SptP (60.1 kD) served as a positive control for secretion. Position of migration of various SptP constructs in 4-20% SDS-PAgels is indicated on the right-hand side of the individual gels.

[00123] Expression of needle complex proteins (InvG, PrgH, and PrgK) were also monitored by immuno-blotting using rabbit anti needle complex (NC) antibodies show similar amounts present in all samples.

[00124] Thus, like wild-type SptP, SptP1 , SptP3 and SptP5 were all secreted into the cell culture medium. Fig. 1 d summarizes the result as well as the size of the substrates in terms of residue and the molecular weight. This example demonstrates that the protein size does not dictate T3SS secretion and substrate of large size and concatemeric design can be transported through T3SS.

Example 2

[00125] The following shows that globular proteins fused at the N terminus of the substrate such as GFP (26kD; about 35x50 A) or ubiquitin (8kD; about 20x30 A) are able to hinder substrate secretion.

Example 2.1 Fusion to ubiquitin

[00126] Overview of the substrate design in this example is shown in Fig. 1c. The construct contains an N-terminal signal (35 amino acids), a chaperone binding domain (cbd) and sequences for ubiquitin (A5JUZ1_mouse) or ubiquitin mutant (I3G/I13G). Sequences were transferred into pACYCDuet-1 (CmR). In addition, these constructs also contained a C-terminally encoded 3x FLAG-tag. Methods and materials as described in Example 1 were used. As negative control (Δ) secretion in a needle complex assembly mutant strain (AprgH) was tested (sup. = supernatant; cbd = chaperone binding domain). Expression of the structural needle complex (NC) components InvG, PrgH, and PrgK in cells was monitored by rabbit anti NC Western blotting.

[00127] The result in Fig. 1 c and 1d shows that only the substrate fused to ubiquitin mutant (I3G/1 13G) but not wild type ubiquitin was specifically secreted by the T3SS. In both samples, natural SptP (see anti SptP in Fig. 1 c) was secreted, indicating that the type-3 secretion machinery is functional. The stability of wild-type ubiquitin prevented the substrate secretion from the T3SS. However, the conformationally-destabilized mutant (I3G/I13G) was secreted by the T3SS. Example 2.2 Fusion to GFP

[00128] In addition to Set-1 in Example 1 , a second set of substrates was generated by fusing green fluorescent protein (GFP) to the C-terminal ends of the SptP1 , SptP3 (SEQ ID NO: 4), SptP5 constructs (Set-2 of Fig. 1 a) before the 3x FLAG-tag

[00129] The result shows that all GFP-containing constructs (Set-2: SptP1 -GFP (91.9 kD), SptP3-GFP (176.2 kD), and SptP5-GFP (260.4 kD; SEQ ID NO: 5) were retained in the cells and could not be detected in the cell culture supernatant (see pSptP-GFP in Fig. 1 b; Fig. 1 d). This demonstrates that C-terminally fused GFP was able to prevent substrate secrete out of the cell.

Example 3 Substrate loading and entry/exit point on injectisomes in situ

[00130] We performed cryo electron tomography to show unperturbed structural details of T3SS in the presence of wild-type substrates or the GFP-fused substrate SptP3-GFP constructed from previous examples.

[00131 ] Cryo electron tomography was performed on Salmonella typhimurium SB905 wild type and SptP3-GFP expressing cells. Five μΙ_ of osmotically shocked bacteria were applied on holey carbon grids (Quantifoil, R2/1 , Mo, 400 mesh) and then mixed with 10 nm gold particles used as fiducial markers and subsequently vitrified automatically in liquid ethane using the Leica grid plunger. Vitrified samples were transferred under liquid nitrogen temperatures into a 300 kV field emission gun (FEG) Polara transmission electron microscope (FEI). Tilt series were acquired using the software SerialEM covering an angular range of -60 to +60 degrees and with tilt increments of 2 degrees. Data were collected set at a defocus of 9 μηι and the total electron dose for the entire tilt series was below 100 electrons/A². For data collection, the magnification was set to 23000 (30120 at specimen level) to obtain a pixelsize of 9.94 A px at specimen level (binning=2). Subsequently, tilt series were binned using the software package IMOD, aligned and tomograms were reconstructed by weighted back-projection. The pixel size in the final tomograms was 19.88 A/px.

[00132] The results are shown in Fig. 2 where nine slices of the tomogram covering 179 A were averaged (n=9) (B = basal body of the T3SS, NC = needle complex, peptidoglycan layer indicated with black dots in sketches, bar = 50nm).

[00133] Referring to Fig 2, the upper panel shows the reconstructed cryo electron tomogram of the cell containing T3SSs in wild type cells. The lower panel shows cells expressing SptP3-GFP (lower panel)). The T3SSs in the presence of the GFP-fused substrates were strikingly similar. However, SB905 cells expressing SptP3-GFP show additional densities at the needle tip of injectisomes (see arrow). We also isolated needle complexes by isopycnic followed by rate-zonal centrifugation to confirm that the GFP-fused substrates can be isolated with the needle complex (data not shown) and only the GFP-fused substrates, but not wild-type substrate (SptP), cofractionated with the needle complexes. This indicates that the extra density at the needle tip is indeed the exit point of the substrate.

[00134] The data demonstrate that fusion of GFP at the C terminus did not preclude entry into the T3SS or lead to repulsion back into the bacteria. Despite the fusion, the substrates were able to be loaded into and transported across injectisomes and became trapped within the secretion path. The roundish appearance of the extra density in cryo electron microscopy tomograms indicates that substrate becomes partially or fully folded after exiting the needle filament.

Example 4 Competition between designed substrates and natural substrates

[00135] In the following example we showed that it is possible to block bacterial effector transport by plugging the needle complex.

[00136] Overview of the design can be seen in Fig. 3a. The substrates used are wildtype SptP, Spt3, and Spt3-GFP. Methods and materials used are as described under Example 1 and 2, except that IPTG was added after 4 hours arabinose induction (time point Omin) for a another growth period of 60, 120, or 180 minutes before harvesting to allow measurement for incremental secretion (one hour before harvesting fresh and pre-warmed (37^°C) LB medium (+ 0.3mM NaCI) was exchanged). We measured the inhibitory effect on the secretion of natural substrates SptP and SipA (Fig. 3b) and SipB (Fig. 3c). Expression of needle complex proteins (InvG, PrgH, and PrgK) monitored by immunoblotting using rabbit anti needle complex (NC) antibodies.

[00137] As shown in Fig. 3b, increasing expression of SptP3-GFP over time (180min) in cells correlated with a decrease of the natural substrates SptP and SipA found in the cell culture supernatant. Overexpression of SptP3 resulted also in an increased level of secreted protein (SptP3) found in the cell culture supernatant, indicating that this protein is able to successfully engage with the secretion machinery.

[00138] Fig. 3c shows that SptP3-GFP also has an inhibitory effect on the secretion of the natural substrate SipB. Increasing expression of SptP3-GFP over time (180min) in cells correlated with a decrease of SipB secreted in the cell culture supernatant.

[00139] This example demonstrates that there is consistent with competition between SptP3-GFP and the endogenous substrates for access to secretion machineries. The GFP fused substrates effectively occupy the secretion path and even competed with natural substrates for secretion. It is therefore possible to attenuate a pathogenic bacterium having T3SS by expressing heterologously substrates which plugs the T3SS.

Example 5 Substrate loading and entry/exit point on injectisomes in vitro

[00140] We isolated needle complexes (with GFP-fused substrate or wild type substrate) by isopycnic followed by rate-zonal centrifugation. Only the GFP-fused substrates, but not wild-type substrate (SptP), cofractionated with the needle complexes. The isolation of substrate-free and substrate-trapped needle complexes was performed essentially as described in Marlovits et al. Nature 441 , 637-640 (2006) and Schraidt et al. PLoS Pathog 6, e1000824 (2010). The same growth conditions were applied as for the secretion assays.

[00141 ] Substrate trapped complexes were then immuno-purified from the isolated needle complexes by using anti-FLAG M2 magnetic beads (Sigma-Aldrich). In buffer A (10 mM Tris, 500 mM NaCI, 5 mM EDTA, 0.1 % (w/v) LDAO (n-Dodecyl-N,N-Dimethylamine-N- Oxide), pH 8.0) equilibrated beads (250 μΙ_, 50% (v/v) slurry) were transferred into a 0.7 ml_ tube and equilibrated 3x10 min with 500 μΙ_ buffer A. CsCI fractions 2-5 of previously purified needle complexes were pooled (about 120 μΙ_), added to the equilibrated beads and the volume adjusted to 700 μΙ_ using buffer A. The solution was kept at room temperature under gentle agitation for 2-3 hours. Thereafter, beads were washed 10x using 400 μΙ_ buffer A each under gentle agitation for 10 min. Bound complexes were eluted with 150 μΙ_ 3x FLAG peptide (400 ng/μί) in buffer A for 45 min at room temperature. The elution fractions were subsequently analyzed by Western blotting.

[00142] As shown in Fig. 4a, right panel bottom, SptP3-GFP containing complexes were immunopurified using a 3x FLAG-tag fused to the C terminus of SptP3-GFP ('S' input sample, 'FT' flow-through, Έ' elution). As a control substrate-free wild type (w.t.) needle complex preparations were not withheld on anti FLAG-beads as analyzed by Western blotting (anti NC) monitoring for the presence of the NC structural proteins InvG, PrgH, and PrgK.

Exit point

[00143] The elution fractions were also analyzed by transmission electron microscopy. Referring to Fig. 4b, substrate-trapped complexes (+pSptP3-GFP) showed an extra density at the terminal end of the needle filament visible on single complexes (white arrow) (2% (w/v) phosphotungstic acid (PTA), 300keV, bar = 30nm). This additional densities at the tip of the needle corresponds to the portion of the injectisome we found protruding from bacterial cells in situ (see Fig. 2a), confirming that this is the exit point of substrates. Entry point

[00144] To visualize the blockage of GFP at the entry point of the needle complex, we conjugated anti GFP antibody to colloidal gold (Bioss) for immune-gold labeling on the isolated, substrate-trapped needle complexes. Model for substrate-trapped needle complexes is illustrated in Fig. 5a.

[00145] Prior to GFP-labeling, measurements showed that the average gold-size visible by electron microscopy was about 20 nm. Labeling was performed on glow-discharged carbon-coated 400 mesh hexagonal Cu/Pd-grids.

[00146] Five μΙ_ of substrate-trapped needle complexes purified by immunoprecipitation were incubated on the grid for 2 min. Afterwards, the grid was treated for 15 min with 5 μΙ_ 5% (w/v) bovine serum albumin in buffer A (10 mM Tris, 500 mM NaCI, 5 mM EDTA, 0.1 % (w/v) LDAO, pH 8.0) to decrease unspecific binding. Subsequently, the grid was incubated with anti GFP-gold diluted in buffer A (1 :100) and incubated for 30 min under a humid atmosphere. Then, the grid was washed 10x with double distilled water and stained with 2% (w/v) phosphotungstic acid (pH 7.0) for transmission electron microscopy.

[00147] Fig. 5b shows trapped SptP3-GFP labeled using an anti-GFP antibody conjugated to colloidal gold. Electron microscopy showed the gold labels exclusively located at the basal side (2% (w/v) PTA, 80keV, bar = 20nm) which in a cellular context is oriented towards the cytoplasm, indicating that this is the substrate entry position.

[00148] The simultaneous demarcation of the substrate entry and exit positions for the substrate shows that N-terminal SptP3-portion of SptP3-GFP enters needle complex and continues until the C-terminally fused GFP domain reaches the entry position because the transport is blocked by GFP due to the folded state. The substrate spanning the needle and occupying the entire secretion path (over a distance of about 800 A) confirms that fully unfolded proteins are passed through T3SS. Furthermore, this also shows that the length of at least three SptP effector domains prior to the globular protein allows for the preceding polypeptides to exit and extend outside the needle complex.

Example 6 Structure study of the substrate-loaded injectisomes

[00149] To examine structural changes associated with protein transport, we investigated the structure of substrate-trapped (SptP3-GFP) injectisomes by cryo electron microscopy and single particle analysis. In total 82403 particles from substrate-trapped needle complexes were used to reconstitute the three-dimensional volume by single particle analysis. For comparison, a similar number of particles of substrate-free particles (74964) was used, obtained by combining a newly collected dataset with previously recorded images.

[00150] In comparison to the structure of the substrate-free needle complex, difference maps with the substrate-trapped (SptP3-GFP) needle complex structure showed an additional density present within the needle filament (Fig. 6a and 6b), which extended far into the inner rod.

[00151 ] Because trapped substrates are expected to be present in vertically slightly different positions within the individual needle complexes used for single particle reconstruction, the difference volume obtained reflects the space that is occupied by the substrate (Fig. 6c, Asub), and these dimensions imply the presence of an unfolded substrate within injectisomes. Difference mapping also revealed small changes within the larger inner ring-1 (IR1 ) of needle complexes (A^IR1), which is distant from the secretion path.

[00152] The secretion path within the injectisome is characterized by areas of different diameters (Fig. 6d): the funnel-shaped 'portal' at the cytoplasmic site tapers from about a 15 to 10 A wide opening to continue into an approximately 10x10 A constriction ('channel') connected to a about 40 A wide space ('atrium'). This means the protein which is used to plug the opening should have a minimal diameter of more than 10-15 A so it would not pass the opening and/or the channel. From the atrium, the secretion path continues with an about 20 A wide 'tunnel', which is defined by the inner rod and the needle filament. The 'portal', 'channel', and 'atrium' are determined by the centrally located cup and socket substructures (Fig. 6d), Surprisingly, we observed that the dimensions of the 'channel', the narrowest part of the secretion path, stay largely invariant during substrate transport. Therefore, the constricting dimensions of the 'channel' can be plugged by an unfolded domain which is located C-terminally from the substrate.

[00153] We also compared the needle length from substrate-free (w.t.) and substrate- trapped (+pSptP3-GFP) needle complexes. We observed that substrate-trapped sucrose- gradient needle complexes display a broader distribution of needle lengths than substrate- free needle complexes. This suggests that the substrate-trapped sample is composed of a mixture of empty and substrate-trapped needle complexes.

[00154] Fig. 7 shows that the most frequent needle length for substrate-trapped NCs is 41.4 nm and for substrate-free NCs 29.9 nm. It is surprisingly found that substrate-containing complexes have longer needles.

Example 7 Secretion of Nanobodies Monobodies and Trascription Activator-Like Effectors.

[00155] The following proteins of interest were fused to first 167 amino-terminal residues of the wildtype SptP as described earlier for T3SS secretion:

Molecular

weight

Constructs name AA (kDA) function SEQ ID NO

vgfp 323 36.194 binds eGFP SEQ ID NO . 6

D NO

Nanobodies vamy 330 37,069 binds Amylase SEQ I ^• 7

ega l 338 37.941 binds EGF SEQ ID NO . 8

HEL 321 35.520 binds Lysozyme SEQ ID NO ^• 9

Adnl 315 35.324 binds EGFR SEQ ID NO 10

Monobodfes HA4 308 34.200 binds Bcr-Abl SEQ ID NO 11

HA4-7cl 2 420 46.025 binds Bcr-Abl SEQ ID NO 12

AvrBs3 1341 142.050 TAL-TF SEQ ID NO 13

Transcription AvrBs3-VP64 1337 141.91 1 TAL-TF

activator SEQ ID NO 14

like endonucleoases AvrBsJ-Fokl 1471 157.665 TAL-TF SEQ ID NO 15

Ubiquitin 283 31.91.2 mammalian, protein SEQ ID NO 16

Ubiquitin niu.ta.nt 283 31.799 mammalian protein SEQ ID NO 17

Granzyme B 462 51.938 mammalian protein SEQ ID NO 18

Others Sox 2 527 57.884 mamma lian TF SEQ ID NO 19

Oct 4 560 61.694. mammalian TF SEQ ID NO 20

cMy 641 71.850 mammalian TF SEQ ID NO 21

Klf4 682 74.360 mammalian TF SEQ ID NO 22

Table 3: constructs for T3SS secretion

[00156] Constructs containing sequences encoding the proteins were transformed into Salmonella Typhimurium strain M2406 by electroporation. Salmonella typhimurium Type Three Secretion System is induced in a high salt (0:3M NaCI) environment, which this leads to the secretion of TTS effectors into the cell culture supernatant, where they can be detected by Western Blotting or the supernatant can be used for further experiments. To detect the secreted proteins a 3-Flag tag was fused C-terminally to the monobodies, nanobodies and other proteins (ubiquitins, Granzyme B, Sox2 and Oct4). AbrBs3 and variants were detected by precipitation, SDS page and immunoblotting. The transgenic Salmonella strains were subsequently used for a secretion assay. During this assay Salmonella Typhimurium is grown in T3SS inducing medium (LB supplemented with 0.3M NaCI) for six hours, allowing the bacteria to secrete T3S proteins into the cell culture supernatant. To detect the secreted proteins, the cells are separated from the supernatant by centrifugation and subsequently the supernatant and cells are subjected to SDS-PAGE and immunoblotting. To check for correct expression and secretion of the artificial substrates, supernatant and cells were probed with a polyclonal SptP antibody (aSptP) and with a monoclonal Flag antibody (aFlag). The SptP antibody allows the direct detection of the SptP moiety. Flag blot additionally allows the detection of the carboxyterminal end of the secreted effectors and therefore the combination ensures that the correct constructs are expressed and secreted. Scheme of the assay is provided in Fig. 10a and the result of the secretion is shown in Fig. 10b.

[00157] The blots in Fig. 10B indicate that all tested nanobody effectors (vgfp, vamy, egal and HEL; bands around 36kDa) and all tested monobodies (Adn1 and HA4; 34:5/ Da) were secreted. Furthermore, T3SS also successfully secreted the tandem monobody (HA4- 7c12, tandem), which gives a band at 46/ Da. In a separate experiment it was found that the secreted nanobodies and monobodies are functional after secretion into cell supernatant.

[00158] The thermodynamically less stable mutant ubiquitin (I3G/I13G, Ubi3,13) which could be unfolded was secreted. In contrast, the stably folded wild-type ubiquitin (Ubi) was not secreted. The proteins Granzyme B (GrmB) and Oct4 also showed no bands in the supernatant and are therefore considered to be not secreted.

[00159] Granzyme B is a trypsin-like serine proteinase, which is the major activator of apoptotic pathways. It is secreted by natural killer (NK) cells and cytotoxic T leukocytes (CTL) to kill transformed or infected cells and activate the apoptotic program. The protein has a stable complex structure forming two β barrels connected by loops as in GFP: As shown above GFP cannot be secreted via Salmonella Typhimurium T3SS. Furthermore, Granzyme contains basic amino acid stretches which diminish its secretion efficiency.

[00160] Oct4 is a transcription factor of the POU (Pit, Oct, Unc) homeobox family. The expression of Oct4 is tightly regulated and high expression is usually found in pluripotent stem cells, where it is responsible to keep the cells in an undifferentiated state. Oct4 contains two helix-loop-helix DNA binding motifs which forms dimers by packing the second helix against the same helix of another molecule. It is also believed that the non-secretion by Salmonella T3SS is due to the in the basic helix establishing a DNA contact.

[00161 ] Fig. 10B bottom shows the cells probed with a polyclonal needle complex antibody (aNC) used as a loading control to estimate the amount of loaded needle complexes (NCs) and to allow the evaluation of needle complex induction and check if cell lysis appeared by probing the supernatant. Multiple prominent bands, corresponding to the major building blocks of the needle complex, are detected by the antibody such as InvG (62/fDa), PrgH (45/ Da) and PrgK (28/cDa). The aNC in the supernatant shows no bands, meaning that no cells were lysed.

[00162] Because of the highly repetitive DNA sequence, the transcription activator-like endonucleases (AbrBs3 and variants) cannot be amplified by PCR during cloning and was therefore sub-cloned to pCASP(MCS2) (Fig. 1 1 a and b) by gateway cloning and the restriction site used for the excision of the full-length construct lies behind a stop codon. The soluble fraction of the AbrBs3 and variants was precipitated by salting out with ammonium sulfate and subjected to SDS-page and immunoblotting. The blots for the AvrBs3 variants are shown in Figure 10c for SptP and NC, respectively. The needle complex blots served as loading and induction control.

[00163] It was found that AvrBs3 and AvrBs3-VP64 TAL-TF, 142kDa, see Fig. 10c) can be secreted, whereas AvrBs3-Fokl (TALEN, 157kDa) was not secreted. The endonuclease domain of Fokl from Flavobacterium okeanokoites is known to form dimers. The lack of secretion observed for AvrBs3-Fokl demonstrates that an oligomerization interface interferes with T3SS secretion.

Claims

1 . A non-pathogenic cell secreting a therapeutic protein via a type III secretion system (T3SS), preferably that encoded by Salmonella Pathogenicity Island 1 (SPI-1 ), said cell comprising:

(i) a nucleotide sequence encoding a type III secretion system,

2. The non-pathogenic cell of claim 1 or 2, wherein said T3SS is from Salmonella typhimurium, Salmonella enterica, Pseudomonas aeruginosa, Yersinia pseudotuberculosis, Yersinia pestis, Shigella felxneri, Citrobacter rodentium, Escherichia coli EHEC, Escherichia colia EPEC, Pseudoman syringae, Ralstonia solanacearum, Xanthominas campestris, or Erwinia amylovora.

3. The non-pathogenic cell of any one of the preceding claims, wherein the following N-terminal secretion tag and chaperone binding domain/chaperones are applied SipA/lnvB, SipC/SicA, SopA lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE.

4. The non-pathogenic cell of any one of the preceding claims,

- wherein nucleotide sequence (ii) and (iii) are contained in a plasmid,

5. The non-pathogenic cell of any one of the preceding claims, wherein the therapeutic protein comprises an antibody, polyclonal antibody, monoclonal antibody, recombinant antibody, antibody fragment, Fab', F(ab')₂, Fv, scFv, di-scFvs, bi-scFvs, tandem scFvs, bispecific tandem scFvs, sdAb, V_H, V_L, humanized antibody, chimeric antibody, IgA antibody, IgD antibody, IgE antibody, IgG antibody, IgM antibody, intrabody, minibody or monobody.

6. A non-pathogenic cell of any one of claims 1-5 for use in a method of treating a disease.

7. A pharmaceutical or diagnostic composition comprising a non-pathogenic cell of any one of the preceding claims.

8. Use of a non-pathogenic cell of any one of claims 1-6 for the production and secretion of a therapeutic protein.

9. A method for the production of a protein of interest, comprising culturing a non-pathogenic cell of any one of claims 1 -20 and harvesting said therapeutic protein from the culture.

10. A cell comprising a type III secretion system (T3SS) and a protein comprising from N- to C-terminus an N-terminal secretion tag recognized by said T3SS, optionally a chaperone binding domain, a protein of interest, a spanning sequence having a length in amino acids of at least three SptP effector domains, and a protein having a globular structure or zinc finger domain.

1 1. The cell of claim 10, wherein the spanning sequence comprises one or more effector domains a T3SS effector protein, and optionally wherein the domain is selected from SipA, SipB, SipC, SipD, SopA, SopB, SopD, SopE2, SptP and AvrA.

12. The cell of claim 10 or 1 1 , wherein the T3SS is encoded by genes from the Salmonella typhimurium pathogenicity island 1 (SPI-1 ) locus.

13. The cell of any one of claims 10 to 12, wherein the following N-terminal secretion tag and chaperone binding domain/chaperones are applied SipA/lnvB, SipC/SicA, SopA lnvB, SopE2/lnvB, SptP/SicP, SopB/SigE.

14. Use of the cell as defined in any one of claims 10-13 or plurality cells thereof for screening a protein of interest that is secretable by said T3SS.

15. A method for identifying a cell secreting a protein of interest via a T3SS comprising

(a) providing a cell as defined in any one of claims 10-13; and