Detection of pathogenic e. coli strains

Abstract

The invention relates to compositions of binding peptides, and individual binding peptides, capable of specific binding to either of two variants of OmpA occurring in pathogenic E. coli, and to methods employing these binding peptides for detection or isolation of pathogenic E. coli.

Description

[0001]Detection of Pathogenic E. Coli Strains This application claims the right of priority of European Patent Applications EP24174412.7 filed 06 May 2024, and EP24175857.2 filed 14 May 2024, which are both incorporated by reference herein. Field The present invention relates to single-chain antibodies that specifically recognize particular E. coli OmpA epitope variants, and to compositions comprising such single-chain antibodies. The invention further relates to methods of sample analysis making use of the single-chain antibodies according to the invention. Background Escherichia coli (E. coli) is an abundant and highly diverse Gram-negative bacterial species that naturally inhabits the gut of humans and animals but is also frequently found in the environment. Despite being an important commensal bacterium, it is a leading cause for bloodstream infections, urinary tract infections, and severe intestinal diseases. Antimicrobial resistance (AMR) is increasingly prevalent in pathogenic E. coli strains. In particular, resistance to cephalosporins and carbapenems caused by extended spectrum β-lactamases (ESBL), carbapenemases, and loss of porin function are of high concern. This explains why this species has been designated as a priority pathogen by the World Health Organization. The clinical manifestation of E. coli can be highly diverse and is strain- and lineage dependent. Accordingly, the species of E. coli is subdivided into eight “pathotypes”, which fall into three main categories, namely enteric and diarrheal disease, urinary tract infections (UTIs) as well as bloodstream infections (BSIs), and meningitis. Characteristic sets of virulence genes equip pathogenic bacteria to cause specific pathotypes. A hallmark of diarrhoeagenic E. coli strains is the secretion of toxins, which can be life-threatening. Of note, large genome studies have clearly shown that Shigella, despite being historically assigned as a separate species, is a specialized diarrheagenic E. coli strain. Based on large sets of genome data, E. coli has been classified into eight phylogroups, and a more fine-grained analysis subdivides the species into 50 main lineages. The E. coli species displays a large diversity of O-antigens that are part of the lipopolysaccharides (LPS) and exposed at the cell surface. Historically, different E. coli strains have been classified according to their serotype; that is a group of E. coli strains being recognized by the same polyclonal antibodies generated in mice or rabbits upon immunization with various strains. According to a recent review, there are 185 known serogroups known for E. coli, each having a different O-antigen structure. The extraordinary O-antigen variability at the surface of E. coli complicates the use of antibodies for diagnostic purposes because an impractically large number of binders would be needed to achieve reasonable species coverage. In contrast, outer membrane proteins (OMPs) are considerably more conserved in terms of their sequence. By occupying more than 50 % of the outer membrane surface, OMPs are an integral part of the asymmetric LPS bilayer, and they are among the most highly expressed proteins of the entire proteome. With an estimated 100’000 copies per cell, OmpA is the most abundant OMP of E. coli and occupies around 6–20 % of the bacterial surface alone, depending on the cell size. However, in intact cells, the surface-accessible loops of OmpA (and other OMPs) are shielded by dense LPS O-antigens, explaining the paucity of reported monoclonal antibodies against E. coli OmpA. Of note, these previously identified OmpA antibodies have not been demonstrated to reliably recognize a larger set of clinical strains. E. coli is a leading cause of bloodstream and urinary tract infections, and in light of the increasing global AMR burden, novel diagnostic tests that would allow for the rapid detection of frequent resistance mechanisms are urgently needed. Especially bloodstream infections (BSIs) pose a diagnostic challenge. If treated with an inadequate antibiotic, the survival rate of BSI patients drops by around 8 % every hour after the onset of recurrent or persistent hypotension, which marks the start of a septic shock. Yet, in patients suffering from BSI, the bacterial load is estimated in the range of only 1 - 1000 cells per millilitre of blood, depending on whether the count is determined by conventional plating or the number of genome copies. Therefore, blood culturing is required to first multiply the bacterial numbers for subsequent diagnostic analysis. For E. coli, the median time to positivity of blood cultures is around 11 hours. Hence, valuable time is lost while waiting for the blood culture to turn positive. There have been recent technical developments tackling the challenge of accelerating the diagnostic testing of BSIs using DNAzyme technology or chemically functionalized magnetic nanoparticles for bacterial capture. However, when applying these new methods, the bacterial cells are lysed during the process of analysis, thereby impeding phenotypic antibiotic susceptibility testing (AST). Based on the above-mentioned state of the art, the objective of the present invention is to provide means and methods to detect or separate E. coli in samples. This objective is attained by the subject-matter of the independent claims of the present specification, with further advantageous embodiments described in the dependent claims, examples, figures and general description of this specification. Summary of the Invention The inventors developed high-quality affinity reagents for the staining and capture of clinical E. coli strains by generating nanobodies recognizing E. coli OmpA in the native context of viable cells. Nanobodies are the variable domain of heavy chain-only antibodies found in camelids such as alpacas (Hamers-Casterman et al., Nature, 1993, vol.363, pp.446–448). A hallmark of nanobodies is their small size; with 15 kDa they are approximately 10 times smaller than classical antibodies. The invention relates to compositions of binding peptides, and individual binding peptides, capable of specific binding to either of two variants of OmpA occurring in pathogenic E. coli, and to methods employing these binding peptides for detection or isolation of pathogenic E. coli. Terms and definitions General For purposes of interpreting this specification, the following definitions will apply and whenever appropriate, terms used in the singular will also include the plural and vice versa. In the event that any definition set forth below conflicts with any document incorporated herein by reference, the definition set forth shall control. The terms “comprising”, “having”, “containing”, and “including”, and other similar forms, and grammatical equivalents thereof, as used herein, are intended to be equivalent in meaning and to be open-ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the listed item or items. For example, an article “comprising” components A, B, and C can consist of (i.e., contain only) components A, B, and C, or can contain not only components A, B, and C but also one or more other components. As such, it is intended and understood that “comprises” and similar forms thereof, and grammatical equivalents thereof, include disclosure of embodiments of “consisting essentially of” or “consisting of.” Where a range of values is provided, it is understood that each intervening value, to the tenth of the unit of the lower limit, unless the context clearly dictates otherwise, between the upper and lower limit of that range and any other stated or intervening value in that stated range, is encompassed within the disclosure, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the disclosure. Reference to “about” a value or parameter herein includes (and describes) variations that are directed to that value or parameter per se. For example, description referring to “about X” includes description of “X.” As used herein, including in the appended claims, the singular forms “a”, “or” and “the” include plural referents unless the context clearly dictates otherwise. "And / or" where used herein is to be taken as specific recitation of each of the two specified features or components with or without the other. Thus, the term "and / or" as used in a phrase such as "A and / or B" herein is intended to include "A and B," "A or B," "A" (alone), and "B" (alone). Likewise, the term "and / or" as used in a phrase such as "A, B, and / or C" is intended to encompass each of the following aspects: A, B, and C; A, B, or C; A or C; A or B; B or C; A and C; A and B; B and C; A (alone); B (alone); and C (alone). Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art (e.g., in cell culture, molecular genetics, nucleic acid chemistry, hybridization techniques and biochemistry, organic synthesis). Standard techniques are used for molecular, genetic, and biochemical methods (see generally, Sambrook et al., Molecular Cloning: A Laboratory Manual, 4th ed. (2012) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. and Ausubel et al., Short Protocols in Molecular Biology (2002) 5th Ed, John Wiley & Sons, Inc.) and chemical methods. The term OmpA in the context of the present specification relates to outer membrane protein A of E. coli (UniProt ID P0A910 – primary accession; P02934 – secondary accession). The terms single-chain antibody and single-domain antibody in the context of the present specification are used interchangeably. These terms refer to a type of antibody fragment consisting of a single variable domain, also known as a VHH domain, derived from the heavy chain of camelid antibodies. The term PBS in the context of the present specification relates to Phosphate Buffered Saline: NaCl: 8.0 g / L; KCl: 0.2 g / L; Na2HPO4: 1.44 g / L KH2PO4: 0.24 g / L, pH 7.4. Sequences In the context of the present specification, the terms sequence identity and percentage of sequence identity refer to a single quantitative parameter representing the result of a sequence comparison determined by comparing two aligned sequences position by position. Methods for alignment of sequences for comparison are well-known in the art. Alignment of sequences for comparison may be conducted by the local homology algorithm of Smith and Waterman, Adv. Appl. Math. 2:482 (1981), by the global alignment algorithm of Needleman and Wunsch, J. Mol. Biol.48:443 (1970), by the search for similarity method of Pearson and Lipman, Proc. Nat. Acad. Sci.85:2444 (1988) or by computerized implementations of these algorithms, including, but not limited to: CLUSTAL, GAP, BESTFIT, BLAST, FASTA and TFASTA. Software for performing BLAST analyses is publicly available, e.g., through the National Center for Biotechnology-Information (http: / / blast.ncbi.nlm.nih.gov / ). One example for comparison of amino acid sequences is the BLASTP algorithm that uses the default settings: Expect threshold: 10; Word size: 3; Max matches in a query range: 0; Matrix: BLOSUM62; Gap Costs: Existence 11, Extension 1; Compositional adjustments: Conditional compositional score matrix adjustment. One such example for comparison of nucleic acid sequences is the BLASTN algorithm that uses the default settings: Expect threshold: 10; Word size: 28; Max matches in a query range: 0; Match / Mismatch Scores: 1.-2; Gap costs: Linear. Unless stated otherwise, sequence identity values provided herein refer to the value obtained using the BLAST suite of programs (Altschul et al., J. Mol. Biol. 215:403-410 (1990)) using the above identified default parameters for protein and nucleic acid comparison, respectively. Reference to identical sequences without specification of a percentage value implies 100% identical sequences (i.e. the same sequence). Particular embodiments make use of the sequences as disclosed herein (i.e.100% identical). General Biochemistry: Peptides, Amino Acid Sequences The term polypeptide in the context of the present specification relates to a molecule consisting of 50 or more amino acids that form a linear chain wherein the amino acids are connected by peptide bonds. The amino acid sequence of a polypeptide may represent the amino acid sequence of a whole (as found physiologically) protein or fragments thereof. The term "polypeptides" and "protein" are used interchangeably herein and include proteins and fragments thereof. Polypeptides are disclosed herein as amino acid residue sequences. The term peptide or oligopeptide in the context of the present specification relates to a molecule consisting of up to 50 amino acids, in particular 8 to 30 amino acids, more particularly 8 to 15 amino acids, that form a linear chain wherein the amino acids are connected by peptide bonds. Amino acid residue sequences are given from amino to carboxyl terminus. Capital letters for sequence positions refer to L-amino acids in the one-letter code (Stryer, Biochemistry, 3rded. p. 21). Lower case letters for amino acid sequence positions refer to the corresponding D- or (2R)- amino acids. Sequences are written left to right in the direction from the amino to the carboxy terminus. In accordance with standard nomenclature, amino acid residue sequences are denominated by either a three letter or a single letter code as indicated as follows: Alanine (Ala, A), Arginine (Arg, R), Asparagine (Asn, N), Aspartic Acid (Asp, D), Cysteine (Cys, C), Glutamine (Gln, Q), Glutamic Acid (Glu, E), Glycine (Gly, G), Histidine (His, H), Isoleucine (Ile, I), Leucine (Leu, L), Lysine (Lys, K), Methionine (Met, M), Phenylalanine (Phe, F), Proline (Pro, P), Serine (Ser, S), Threonine (Thr, T), Tryptophan (Trp, W), Tyrosine (Tyr, Y), and Valine (Val, V). Markers, ligands The terms detectable label or detectable marker in the context of the present specification relate to a molecule with distinctive properties that is conjugated or attached to a carrier molecule, such as a ligand or a molecular probe, more specifically to an antibody or nucleic acid hybridization probe, in order to facilitate the detection and quantification of the ligand’s target in a biological sample. It allows to visually or instrumentally identify the presence or concentration of the target molecule. Common types of detectable labels used in procedures such as ELISA and other diagnostic methods include enzymes, fluorescent molecules, radioactive isotopes, colloidal gold particles, and biotin, among others. These labels emit signals or exhibit distinct properties that can be measured or visualized to determine the presence and quantity of the target molecule in the sample. The term fluorescent dye in the context of the present specification relates to a small molecule capable of fluorescence in the visible or near infrared spectrum. Examples for fluorescent labels or labels presenting a visible color include, without being restricted to, fluorescein isothiocyanate (FITC), rhodamine, allophycocyanine (APC), peridinin chlorophyll (PerCP), phycoerithrin (PE), Alexa Fluors (Life Technologies, Carlsbad, CA, USA), dylight fluors (Thermo Fisher Scientific, Waltham, MA, USA) ATTO Dyes (ATTO-TEC GmbH, Siegen, Germany), BODIPY Dyes (4,4- difluoro-4-bora-3a,4a-diaza-s-indacene based dyes) and the like. In certain embodiments, the fluorescent dye is a triarylmethane or xanthene type fluorophore. In certain particular embodiments, the fluorescent dye is selected from a rhodamine, a silicon rhodamine, a fluorescein, a Janelia Fluor dye, an olefinic silicon rhodamine derivative with an exocyclic double bond, a cell permeable (MaP) xanthene fluorophore dye, a carbopyronine, a carbocyanine (particularly a Cy3, or a Cy5 dye), a pyrene, a Bodipy fluorophore, a coumarine, a rhodol, and an Alexa dye. In certain embodiments, the fluorescent dye is a rhodamine selected from carboxytetramethylrhodamine (TAMRA), tetramethylrhodamine (TMR) and the isothiocyanate derivative TRITC, sulforhodamine 101, Texas Red, and Rhodamine Red. In certain embodiments, the fluorescent dye is a silicon rhodamine, a rhodamine wherein the central oxygen atom is replaced by SiR’2(with R’ selected from C1 to C4 alkyl). In certain embodiments, the fluorescent dye is a fluorescein selected from 3′,6′- dihydroxyspiro[isobenzofuran-1(3H),9′-[9H]xanthen]-3-one, fluorescein isothiocyanate (FITC) and, 6-FAM phosphoramidite. The Janelia Fluor family of molecules comprises rhodamine-type dyes having an azetidine moiety formed around the nitrogen atoms the outer rings. In certain embodiments, the fluorescent dye is a Janelia Fluor dye selected from JF646, JF635, JF585, JF549, JF525, and JF503. In certain embodiments, the fluorescent dye is a Bodipy fluorophore is selected from BODIPY FL, BODIPY R6G, BODIPY TMR, BODIPY 581 / 591, BODIPY TR, BODIPY 630 / 650, and BODIPY 650 / 665. In certain embodiments, the Alexa dye is selected from Alexa Fluor 350, Alexa Fluor 350, Alexa Fluor 405, Alexa Fluor 430, Alexa Fluor 488, Alexa Fluor 500, Alexa Fluor 514, Alexa Fluor 532, Alexa Fluor 546, Alexa Fluor 555, Alexa Fluor 568, Alexa Fluor 594, Alexa Fluor 610, Alexa Fluor 633, Alexa Fluor 635, Alexa Fluor 647, Alexa Fluor 660, Alexa Fluor 680, Alexa Fluor 700, Alexa Fluor 750, and Alexa Fluor 790. In the context of the present specification, the term labelled antibody is used for an antibody being covalently bound to a detectable label. It is understood that any useful antibody-like ligand may be used instead of a labelled antibody. Labelled antibodies are particularly advantageous as they are in many cases commercially available or easily obtained by labelling commercially available antibodies with off-the-shelf labelling kits. Detectable labels include for example, without limitation, octadecyl rhodamine B, 7-nitro-2-1,3-benzoxadiazol-4-yl, 4- acetamido-4′-isothiocyanatostilbene-2,2′ disulfonic acid, acridine and derivatives, 5-(2′- aminoethyl)aminonaphthalene-1-sulfonic acid (EDANS), 4-amino-N-(3-[vinylsulfonyl]phenyl)- naphthalimide-3,6-disulfonate dilithium salt, N-(4-anilino-1-naphthyl)maleimide, anthranil- amide, BODIPY, Brilliant Yellow, coumarin and derivatives, cyanine dyes, cyanosine, 4′,6- diaminidino-2-phenylindole (DAPI), bromopyrogallol red, 7-diethylamino-3-(4′- isothiocyanatophenyl)-4-methylcoumarin, diethylenetriamine pentaacetate, 4,4′- diisothiocyanatodihydro-stilbene-2,2′-disulfonic acid, 4,4′-diisothiocyanatostilbene-2,2′- disulfonic acid, dansylchloride, 4-dimethylaminophenylazophenyl-4′-isothiocyanate (DABITC), eosin and derivatives, erythrosin and derivatives, ethidium, fluorescein, 5-carboxyfluorescein (FAM), 5-(4,6-dichlorotriazin-2-yl)aminofluorescein (DTAF), 2′,7′-dimethoxy-4′5′-dichloro-6- carboxyfluorescein, fluorescein isothiocyanate, X-rhodamine-5-(and 6)-isothiocyanate (QFITC or XRITC), fluorescamine, IR-144 (2-[2-[3-[[1,3-dihydro-1,1-dimethyl-3-(3-sulfopropyl)-2H- benz[e]indol2-ylidene]ethylidene]-2-[4-(ethoxycarbonyl)- 1-piperazinyl]-1-cyclopenten-1- yl]ethenyl]-1,1-dimethyl-3-(3-sulforpropyl)-1H-benz[e]indolium hydroxide, inner salt, compound with n,n-diethylethanamine(1:1), CAS No.: 54849-69-3), 5-chloro-2-[2-[3-[(5-chloro- 3-ethyl-2(3H)-benzothiazol-ylidene)ethylidene]-2-(diphenylamino)-1-cyclopenten-1- yl]ethenyl]-3-ethyl benzothiazolium perchlorate (IR140), malachite green isothiocyanate, 4- methylumbelliferone, ortho cresolphthalein, nitrotyrosine, pararosaniline, phenol red, B- phycoerythrin, o-phthaldialdehyde, pyrene, pyrene butyrate, succinimidyl 1-pyrene, butyrate quantum dots, Reactive Red 4 (Cibacron Brilliant Red 3B-A), rhodamine and derivatives, 6- carboxy-X-rhodamine (ROX), 6-carboxyrhodamine (R6G), lissamine rhodamine B sulfonyl chloride rhodamine (Rhod), rhodamine B, rhodamine 123, rhodamine X isothiocyanate, sulforhodamine B, sulforhodamine 101, sulfonyl chloride derivative of sulforhodamine 101 (Texas Red), N,N,N′,N′tetramethyl-6-carboxyrhodamine (TAMRA) tetramethyl rhodamine, tetramethyl rhodamine isothiocyanate (TRITC), riboflavin, rosolic acid, terbium chelate derivatives, Cyanine-3 (Cy3), Cyanine-5 (Cy5), Cyanine-5.5 (Cy5.5), Cyanine-7 (Cy7), IRD 700, IRD 800, Alexa 647, La Jolta Blue, phthalo cyanine, and naphthalo cyanine. The term specific binding in the context of the present invention refers to a property of ligands that bind to their target with a certain affinity and target specificity. The affinity of such a ligand is indicated by the dissociation constant of the ligand. A specifically reactive ligand has a dissociation constant of ≤ 10-8mol / L (particularly ≤ 10-9mol / L) when binding to its target, but a dissociation constant at least three orders of magnitude higher in its interaction with a molecule having a globally similar chemical composition as the target, but a different three-dimensional structure. Detailed Description of the Invention The invention relates to compositions of binding peptides, and individual binding peptides, capable of specific binding to either of two variants of OmpA occurring in pathogenic E. coli, and to methods employing these binding peptides for detection or isolation of pathogenic E. coli. One aspect of the invention relates to a composition comprising: a. a first binding agent comprising a first single-chain polypeptide capable of specific binding to a first epitope variant of E. coli OmpA (SEQ ID NO 054; OmpA-short) (MKKTAIAIAVALAGFATVAQAAPKDNTWYTGAKLGWSQYHDTGFINNNGPTHENQLGAGAFGGYQVNP YVGFEMGYDWLGRMPYKGSVENGAYKAQGVQLTAKLGYPITDDLDIYTRLGGMVWRADTKSNVYGKNH DTGVSPVFAGGVEYAITPEIATRLEYQWTNNIGDAHTIGTRPDNGMLSLGVSYRFGQGEAAPVVAPAP APAPEVQTKHFTLKSDVLFNFNKATLKPEGQAALDQLYSQLSNLDPKDGSVVVLGYTDRIGSDAYNQG LSERRAQSVVDYLISKGIPADKISARGMGESNPVTGNTCDNVKQRAALIDCLAPDRRVEIEVKGIKDV VTQPQA) and b. a second binding agent comprising a second single-chain polypeptide capable of specific binding to a second epitope variant of E. coli OmpA (SEQ ID NO 055; OmpA-long) (MKKTAIAIAVALAGFATVAQAAPKDNTWYTGAKLGWSQYHDTGFIPNNGPTHENQLGAGAFGGYQVN PYVGFEMGYDWLGRMPYKGDNINGAYKAQGVQLTAKLGYPITDDLDIYTRLGGMVWRADTKANVPGGA SFKDHDTGVSPVFAGGVEYAITPEIATRLEYQWTNNIGDAHTIGTRPDNGMLSLGVSYRFGQGEAAPV VAPAPAPAPEVQTKHFTLKSDVLFTFNKATLKPEGQAALDQLYSQLSNLDPKDGSVVVLGYTDRIGSD AYNQALSERRAQSVVDYLISKGIPADKISARGMGESNPVTGNTCDNVKQRAALIDCLAPDRRVEIEVK GIKDVVTQPQA). In certain embodiments, first and second binding agent are linked to a detectable label as set forth below. Nb01 and Nb39 can cover 79 % or 91 % of the total available OmpA sequences found in the Swiss or 661k database, respectively. In certain embodiments, first and second binding agent are attachable, or attached, to a surface, facilitating isolation of a bound E. coli bacterium. The two isoforms of OmpA have been described before. The fact, however, that a majority of OmpA sequences related to clinically relevant strains of E. coli can all be classified into these two main groups / isoforms and that two binding agents suffice to reach very high sequence coverage has, to the inventor’s best knowledge, not been published previously. The sequences given as SEQ ID NO 054 and 055 contain the leader sequence, which is later cleaved off and not present in the folded OMP. The binding agents of the invention are not specific for the leader sequence, but for the sequence of the OmpA β-barrel including four extracellular loops. The term binding agent denotes a polypeptide, particularly an OmpA-specific single domain antibody or arrangement of single domain antibodies, and a moiety for detection or isolation of the OmpA-specific single domain antibody. The binding agent may be entirely composed of amino acids, or may be a construct that combines a single domain antibody with a covalently attached modification, such as a dye, a biotin, or other functionalities that enable direct or indirect detection or isolation of the single domain antibody and its cognate antigen. In certain embodiments, the binding agent may be attached to a surface, for example the surface of a microsphere. A microsphere to which the binding agent, or both binding agents, are attached may be susceptible to magnetic separation, for example by comprising a ferromagnetic core. This will result in the particle being retained in a magnetic field, along with any bacterium attached thereto by means of binding to the first or second binding agent, while the rest of the sample may be washed off. Likewise, the microsphere may be fluorescently labelled. In particular embodiments, the microsphere may comprise both a core that renders the particle susceptible to magnetic fields, and a fluorescent label. In certain embodiments of the composition according to the invention, the first single-chain polypeptide does not bind to the second epitope variant, and the second single-chain polypeptide does not bind to the first epitope variant. This allows individual determination of the variants for epidemiological or other purposes. In certain embodiments of the composition according to the invention, the first single-chain polypeptide is capable of binding specifically to the first epitope variant of OmpA (SEQ ID NO 054) when said first epitope variant is present on the surface of a clinically relevant E. coli bacterium, and the second single-chain polypeptide is capable of binding specifically to the second epitope variant of OmpA (long, SEQ ID NO 055) when said second epitope variant is present on the surface of a clinically relevant E. coli bacterium. Binder pairs with close sequence similarity to Nb01 and Nb39: In particular embodiments of the composition according to the invention, a. the first single chain polypeptide comprises a CDR1, CDR2 and CDR3 sequence according to the Kabat notation and wherein CDR1 is selected from SEQ ID NO 014 [GTGFTFSKSPMS] and SEQ ID NO 015 [GSGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is selected from SEQ ID NO 017 [RRLGKTTYDY] and SEQ ID NO 018 [RRLGKRTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to one of SEQ ID NO 014, 015, combined with SEQ ID NO 016 and with one of SEQ ID NO 017 and 018, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01), and / or b. the second single chain polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 043 [GSIFNFNPMG], SEQ ID NO 044 [GSIFNFNLMG], SEQ ID NO 045 [GGIFNFNIMG], and SEQ ID NO 046 [GSIFNFNIMG] and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to either SEQ ID NO 043, 044, 045 or 046, and to 047, and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). The notation applied to annotate CDR regions has been: CDR1: Kabat Residues 26-35; CDR2: Kabat Residues 50-58; CDR3: Kabat Residues 95-102. In more particular embodiments of the composition according to the invention, a. the first single chain polypeptide comprises a CDR1, CDR2 and CDR3 sequence according to the Kabat notation and wherein CDR1 is SEQ ID NO 014 [GTGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is SEQ ID NO 017 [RRLGKTTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 014, 016 and 017 respectively, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01), and / or b. the second single chain polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is SEQ ID NO 043 [GSIFNFNPMG], and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to SEQ ID NO 043 and 047 and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). Within the first single-chain polypeptides (capable of specific binding to the first, short, epitope variant of E. coli OmpA, SEQ ID NO 054) identified in the course of the studies underlying the present invention, seven individual binders characterized herein comprised identical or almost (one amino acid varied) identical CDRs: CDR1 GTGFTFSKSPMS (SEQ ID NO 014) is present identically in Nb01, 02, 03, 04, 06. A variant having one amino acid exchange T2S is present in NB05 and 08. CDR2 AIFADSSTY (SEQ ID NO 016) is present identically in Nb01, 02, 03, 04, 05, 06, 08 (SEQ ID NO 001, 002.003, 004, 005, 006, and 008, respectively). CDR3 RRLGKTTYDY (SEQ ID NO 017) is present identically in Nb01, 03, 05, 06, 08. A variant having one amino acid exchange T6R is present in Nb02 and 04. Within the second single-chain polypeptides (capable of specific binding to the second, long, epitope variant of E. coli OmpA, SEQ ID NO 055) identified in the course of the studies underlying the present invention, four individual binders characterized herein comprised identical or almost (one or two amino acids varied) identical CDRs: CDR1 GSIFNFNPMG (SEQ ID NO 043) is present identically in Nb39, and variants having one amino acid exchange are present in NB38 (P8L) and 40 (P8I), Nb42 shows a two-AA conversion mutant (S2G and P8I). CDR3 NYRIGRNDLPV (SEQ ID NO 047) is present identically in all of NB38, 39, 40 and 42 (SEQ ID NO 038, 039, 040 and 042, respectively). The “same biological activity” in the context of the first single chain polypeptide refers to the ability of the first single chain polypeptide to bind to SEQ ID NO 054 as assayed in Example 3 of the present specification. The “same biological activity” in the context of the second single chain polypeptide refers to the ability of the second single chain polypeptide to bind to SEQ ID NO 055 as assayed in Example 3 of the present specification. “The same biological activity” means in both contexts that the polypeptide thus qualified binds to its target, OmpA-short or OmpA-long, respectively, with at least the same specificity as the reference antibody. In certain embodiments, the biological activity is the same if the polypeptide thus qualified binds to its target, OmpA-short or OmpA-long, respectively, with approximately the same specificity as the reference antibody. This means, the biological activity is the same if the polypeptide of a sequence having e.g.90% sequence identity with a polypeptide of sequence SEQ ID NO 054 binds to the OmpA short target with approximately the same specificity as the polypeptide of SEQ ID NO 054. The same holds true for SEQ ID NO 055. In certain embodiments, the biological activity is assayed via a method comprising the steps: a. labelling the polypeptide directly with AlexaFluor 647 (AF647) via a cysteine introduced at the C-terminus; and b. observing binding of the polypeptide to its target, OmpA-short or OmpA-long, respectively, by fluorescence microscopy. In particular embodiments, the sequence of CDR1, CDR2 and CDR3 of the first single chain polypeptide are identical to SEQ ID NO 014, 016 and 017 except for a maximum total of two conversions within the CDR sequences. In more particular embodiments, the residues depicted in bold face above (K and P in CDR1; F, A, D, the second S and Y in CDR2, the first six residues in CDR3) are preserved. In more particular embodiments, the sequence of CDR1, CDR2 and CDR3 of the first single chain polypeptide are identical to SEQ ID NO 014, 016 and 017 except for a maximum total of one conversion within the CDR sequences. In particular embodiments, the sequence of CDR1 and CDR3 of the second single chain polypeptide are each identical to SEQ ID NO 043 and 047 except for a maximum total of two conversions within the CDR sequences. In more particular embodiments, the sequence of CDR1 and CDR3 of the second single chain polypeptide are each identical to SEQ ID NO 043 and 047 except for a maximum total of one conversion within the CDR sequences. In particular embodiments, the second single chain polypeptide comprises a CDR2 sequence selected from SEQ ID NO 048 [TLTRDGVEN], SEQ ID NO 049 [TLTRDGSEN] and SEQ ID NO 050 [TMTRDGSAS]. In more particular embodiments, the second single chain polypeptide comprises the CDR2 sequence SEQ ID NO 048 [TLTRDGVEN]. In even more particular embodiments of the composition according to the invention, a. the first single chain polypeptide comprises a sequence selected from: i. SEQ ID NO 001 (Nb01) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSVRGR FTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS; ii. SEQ ID NO 002 (Nb02) SQMQFVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSV KGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGQGTPVTVS; iii. SEQ ID NO 003 (Nb03) SQGQLVESGGGLVPPGGSLRLSCAVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSV RGRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS; iv. SEQ ID NO 004 (Nb04) SQRQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSV KGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGKGTPVTVS; v. SEQ ID NO 005 (Nb05) SQRQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSV KGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVTVS; vi. SEQ ID NO 006 (Nb06) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSV RGRFTISRDNAKNTVYLEMNNVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVTVS; vii. SEQ ID NO 008 (Nb08) SQMQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSV KGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGYRRLGKTTYDYRGQGTPVTVS; or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 001, 002, 003, 004, 005, 006 or 008, and having the same biological activity as SEQ ID NO 001 (Nb01), and / or b. the second single chain polypeptide comprises a sequence selected from: i. SEQ ID NO 039 (Nb39) ISRDSAKNTLYLQMTDVKPGDAAVYICHANYRIGRNDLPVWGKGTPVTVS; ii. SEQ ID NO 038 (Nb38) RFTISRDSGKNTMYLQMTDVKPSDTAVYICHANYRIGRNDLPVWGKGTRVTVS; iii. SEQ ID NO 040 (Nb40) SQRQLVESGGGLVQPGGSLRLSCVPNGSIFNFNIMGWYRQNAGNQRELVATMTRDGSASYSDSVKG RFTISRDVDKNTIYLQLDSVKPEDTAVYICHANYRIGRNDLPVWGRGTRVTVS; iv. SEQ ID NO 042 (Nb42) SQLQLVESGGGLVQPGGSLRLSCVPNGGIFNFNIMGWYRQNAGNQRELVATMTRDGSASYSDSVKG RFTISRDVDKNTIYLQMDSVEPEDTAVYICHANYRIGRNDLPVWGQGTPVTVS; or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 039, 038, 040, or 042, and having the same biological activity as SEQ ID NO 039 (Nb39). In particular embodiments, the sequence of the first single chain polypeptide is each at least (≥) 90%, particularly ≥95%, more particularly ≥98% identical to one of SEQ ID NO 001 (Nb01), and the residues depicted in bold face above in SEQ ID NO 001 are preserved, and the single chain polypeptide has the same biological activity as SEQ ID NO 001. In particular embodiments, the sequence of the second single chain polypeptide is each at least 90% identical to SEQ ID NO 001 depicted in bold face above in SEQ ID NO 001 are preserved, and the single chain polypeptide has the same biological activity as SEQ ID NO 001. Single binders to OmpA-short with close sequence similarity to Nb01: Another aspect of the invention relates to a binding agent comprising a first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli, said first single- chain polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is selected from SEQ ID NO 014 [GTGFTFSKSPMS] and SEQ ID NO 015 [GSGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is selected from SEQ ID NO 017 [RRLGKTTYDY] and SEQ ID NO 018 [RRLGKRTYDY]. In some embodiments, the first single-chain polypeptide may be characterized by a sequence of CDR1, CDR2 and CDR3 showing a total (among the three sequences) of one or two conversions relative to one of SEQ ID NO 014, 015, combined with SEQ ID NO 016 and with one of SEQ ID NO 017 and 018, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01). In particular embodiments thereof, CDR1 is SEQ ID NO 014 [GTGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is SEQ ID NO 017 [RRLGKTTYDY], or the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 014, 016 and 017, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01). In more particular embodiments thereof, the first single-chain polypeptide comprises a sequence selected from: i. SEQ ID NO 001 (Nb01) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSVRGRF TISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS; ii. SEQ ID NO 002 (Nb02) SQMQFVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSVK GRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGQGTPVTVS; iii. SEQ ID NO 003 (Nb03) SQGQLVESGGGLVPPGGSLRLSCAVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSVR GRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS; iv. SEQ ID NO 004 (Nb04) SQRQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSVK GRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGKGTPVTVS; v. SEQ ID NO 005 (Nb05) SQRQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSVK GRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVTVS; vi. SEQ ID NO 006 (Nb06) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSVR GRFTISRDNAKNTVYLEMNNVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVTVS; vii. SEQ ID NO 008 (Nb08) SQMQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYADSVK GRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGYRRLGKTTYDYRGQGTPVTVS; or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO SEQ ID NO 001, 002, 003, 004, 005, 006 or 008, and having the same biological activity as SEQ ID NO 001 (Nb01). Single Binders to OmpA-short different from the Nb01 family of binders: Similarly encompassed by the invention disclosed herein are binders to SEQ ID NO 054 (OmpA- short): In one embodiment, a binding agent comprises a first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli, said first single-chain polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 019 [GLAFSSYASG], CDR2 is SEQ ID NO 020 [AMTARGGFTN], and CDR3 is SEQ ID NO 021 [DPRRYGSQVY], or the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 019, 020 and 021, and the single chain polypeptide has the same biological activity as SEQ ID NO 007 (full sequence of Nb07). In a particular embodiment, the binding agent comprises the first single-chain polypeptide comprises SEQ ID NO 007 (Nb07) SQGQLLESGGGLVQPGGSLRLSCAASGLAFSSYASGWYRQAPGKERELVAAMTARGGFTNYADSVKGRFTI SRDNGKNTVYLQMNSLKPEDTAVYYCNADPRRYGSQVYWGQGTPVTVS, or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 007, and having the same biological activity as SEQ ID NO 007 (Nb07). In another embodiment, the first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli comprises a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 022 [GFRFSSVALS], CDR2 is SEQ ID NO 023 [LITNDHKSR], and CDR3 is SEQ ID NO 024 [SLLGSNYGRY], or the sequences of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 022, 023 and 024, and the single chain polypeptide has the same biological activity as SEQ ID NO 009 (full sequence of Nb09). In another particular embodiment, the binding agent comprises the first single-chain polypeptide comprises SEQ ID NO 009 (Nb09) SQLQFVESGGGMVQPGGSLRLSCVASGFRFSSVALSWYRQAPGRDRELVALITNDHKSRYGDFVKDRFTISR DNNKNTVYLQMNNLKPEDTAVYSCGVSLLGSNYGRYWGKGTRVTVS, or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 009, and having the same biological activity as SEQ ID NO 009 (Nb09). In another embodiment, the first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli comprises a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 025 [GFTFSSYAMT], CDR2 is SEQ ID NO 026 [LITNDARTR], and CDR3 is SEQ ID NO 027 [SLLGRNYGQH], or the sequences of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 025, 026 and 027, and the single chain polypeptide has the same biological activity as SEQ ID NO 010 (full sequence of Nb10). In another particular embodiment, the binding agent comprises the first single-chain polypeptide comprises SEQ ID NO 010 (Nb10) SQGQLVESGGGLVQPGGSLRLSCTASGFTFSSYAMTWHRQAPGKERELVALITNDARTRYGDFVKGRFTIS RDNAKNTIYLQMNTLAPEDTALYYCGVSLLGRNYGQHWGKGTRVTVS, or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 010, and having the same biological activity as SEQ ID NO 010 (Nb10). In another embodiment, the first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli comprises a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 028 [GRTFNPYAMG], CDR2 is SEQ ID NO 029 [AIRWSGGSIN], and CDR3 is SEQ ID NO 030 [AKSLGVWAREYDD], or the sequences of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 028, 029 and 030, and the single chain polypeptide has the same biological activity as SEQ ID NO 011 (full sequence of Nb11). In another particular embodiment, the binding agent comprises the first single-chain polypeptide comprises SEQ ID NO 011 (Nb11) SQGQLVESGGGLVQAGGSLRLSCAASGRTFNPYAMGWFRQAPGKEREFVAAIRWSGGSINYADSVKGRFTI SRDNAKNTVYLQMNSLKPEDTAVYYCNAAKSLGVWAREYDDWGKGTPVTVS, or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 011, and having the same biological activity as SEQ ID NO 011 (Nb11). In another embodiment, the first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli comprises a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 031 [GSTSSIDTMS], CDR2 is SEQ ID NO 032 [SITKGGGRPY], and CDR3 is SEQ ID NO 033 [RDSNSGFYY], or the sequences of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 031, 032 and 033, and the single chain polypeptide has the same biological activity as SEQ ID NO 012 (full sequence of Nb12). In another particular embodiment, the binding agent comprises the first single-chain polypeptide comprises SEQ ID NO 012 (Nb12) SQRQLVESGGGLVQAGGSLRLSCAASGSTSSIDTMSWYRQTTGNEREMVASITKGGGRPYYDFSVKGRFTI SRDNDKNTMSMQMNSLKPEDTAVYYCNARDSNSGFYYWGKGTPVTVS, or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 012, and having the same biological activity as SEQ ID NO 012 (Nb12). In another embodiment, the first single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli comprises a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 034 [GFTFSSALMR], CDR2 is SEQ ID NO 035 [YVNNKDLYTF], and CDR3 is SEQ ID NO 036 [ASTPNT], or the sequences of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 034, 035 and 036 respectively, and the single chain polypeptide has the same biological activity as SEQ ID NO 013 (full sequence of Nb13). In another particular embodiment, the binding agent comprises the first single-chain polypeptide comprises SEQ ID NO 013 (Nb13) SQGQFVESGGALVQPGGSLRLSCAASGFTFSSALMRWYRQAAGKERELVAYVNNKDLYTFYVDSVKGRFTIS RDNAKNTVYLQMNSLKPEDTAVYYCNAASTPNTWGQGTRVTVS, or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 013, and having the same biological activity as SEQ ID NO 013 (Nb13). Single binders to OmpA-long with close sequence similarity to Nb39: In another embodiment, the first single-chain polypeptide capable of selectively binding to OmpA (SEQ ID NO 055) of a clinical isolate of E. coli comprises a CDR1 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 043 [GSIFNFNPMG], SEQ ID NO 044 [GSIFNFNLMG], SEQ ID NO 045 [GGIFNFNIMG], and SEQ ID NO 046 [GSIFNFNIMG] and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or the sequences of CDR1 and CDR3 show a total of one or two conversions relative to either SEQ ID NO 043, 044, 045 or 046, and to 047, and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). In particular embodiments thereof, the second single chain polypeptide comprises a CDR2 sequence selected from the group of SEQ ID NO 048, 049 and 050. In certain embodiments thereof, CDR1 is SEQ ID NO 043 [GSIFNFNPMG], and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV], or the sequences of CDR1 and CDR3 show a total of one or two conversions relative to SEQ ID NO 043 and 047 respectively and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). In particular embodiments the second single chain polypeptide comprises a sequence selected from: viii. SEQ ID NO 039 (Nb39) SQRQLVESGGGLVHTGGSLKLSCVPNGSIFNFNPMGWYRQVSGQQRELVATLTRDGVENYASSVKGRFTISR DSAKNTLYLQMTDVKPGDAAVYICHANYRIGRNDLPVWGKGTPVTVS; ix.SEQ ID NO 038 (Nb38) SQRQLVESGGGTVQTGGSLRLSCVPNGSIFNFNLMGWYRQSSGQQRELVATLTRDGSENYAEFVKGRF TISRDSGKNTMYLQMTDVKPSDTAVYICHANYRIGRNDLPVWGKGTRVTVS; x. SEQ ID NO 040 (Nb40) SQRQLVESGGGLVQPGGSLRLSCVPNGSIFNFNIMGWYRQNAGNQRELVATMTRDGSASYSDSVKGRF TISRDVDKNTIYLQLDSVKPEDTAVYICHANYRIGRNDLPVWGRGTRVTVS; xi.SEQ ID NO 042 (Nb42) SQLQLVESGGGLVQPGGSLRLSCVPNGGIFNFNIMGWYRQNAGNQRELVATMTRDGSASYSDSVKGRF TISRDVDKNTIYLQMDSVEPEDTAVYICHANYRIGRNDLPVWGQGTPVTVS; or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 039, 038, 040 or 042, and having the same biological activity as SEQ ID NO 039 (Nb39). Single Binders to OmpA-long different from the Nb39 family of binders: In another embodiment, a binding agent comprises a (second, in the parlance of this specification, to reflect the fact that it binds to the (second) OmpA-long isoform) single-chain polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli (OmpA-long, SEQ ID NO 055), wherein the single chain polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is SEQ ID NO 051 [DNILQFGNMG], and CDR3 is SEQ ID NO 053 [QYVIGRNRLDV], or the sequences of CDR1 and CDR3 show a total of one or two conversions relative to SEQ ID NO 051 and 053 respectively and the single chain polypeptide has the same biological activity as SEQ ID NO 041 (full sequence of Nb41). In particular embodiments, the polypeptide comprises, in addition to CDR1 of SEQ ID NO 051 and CDR3 of SEQ ID NO 053, CDR2 of SEQ ID NO 052. In particular embodiments, the second single chain polypeptide comprises SEQ ID NO 041 (Nb41): SQRQLVESGGGSMQPGESLTLSCEASDNILQFGNMGWYRQSPGTQRELVARIHKRGDSDYGDFAKGRFTIS RDTVKNKVYLQMTDLKPEDSANYICNGQYVIGRNRLDVWGQGTPVTVS; or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 041, and having the same biological activity as SEQ ID NO 041 (Nb41). Binder pairs with close sequence similarity to Nb01 and Nb39: Also encompassed within the present invention is a composition comprising a first binding agent characterized by a first single-chain polypeptide capable of specific binding to E. coli OmpA-short (SEQ ID NO 054) and a second single-chain polypeptide capable of specific binding to E. coli OmpA-long (SEQ ID NO 055), wherein the first single-chain polypeptide is selected from any of SEQ ID NO 001, 002, 003, 004, 005, 006, 007, 008, 009, 010, 011, 012, or 013 (Nb01, Nb02, Nb03, Nb04, Nb05, Nb06, Nb07, Nb08, Nb09, Nb10, Nb11, Nb12, or Nb13) and the second single-chain polypeptide is selected from any of SEQ ID NO 038, 039, 040, 041 or 042 (Nb38, Nb39, Nb03, Nb40, Nb41, Nb42). In particular embodiments of the compositions according to any one of the previously disclosed aspects of the invention, or of the individual binding agents specific for the OmpA-short or long isoform, the binding agents are covalently attached to a fluorescent dye, i.e. the first binding agent and, where applicable, the second binding agent are linked to a fluorescent dye. The crystal structures depicted in the Figures and discussed in the Examples clearly denote which residues are involved in binding to which loops of OmpA-short / long. The CDRs as based on the Kabat numbering scheme (CDR1: Kabat Residues 26-35; CDR2:Kabat Residues 50-58; CDR3: Kabat Residues 95-102) comprise most of the contacting residues, and can thus be designated as the main contributor to the specificity observed. Some framework residues, however, which may not be identical in all binding sequences, also interact with OmpA, according to the structure. The inventors did not perform exhaustive mutation studies, but have characterized a number of binders for each target. For OmpA-short, they have identified Nb01-Nb13. Among these, Nb1-Nb6 + Nb8 are very similar and have some differences in CDR1 and CDR3 and in the framework. In certain embodiments, different dyes may be attached to the binding agents specifically recognizing the two isoforms, i.e. a first fluorescent dye is attached to the first binding agent and a second, optically discernible dye is attached to the second binding agent. For conducting comparative infection experiments (to differentiate two different E. coli strains), differential labelling may be helpful. Clinically, it might be (under certain circumstances) interesting to differentiate E. coli strains with long or short OmpA, e.g. in the surveillance of an outbreak (e.g. EHEC E. coli). Some advantage may thus be attributed to different labelling in the context of basic science or epidemiologic surveillance. In certain embodiments, the same dye may be attached to the first and second binding agents specifically recognizing the two isoforms. This will make detection easier and lower the cost of the system, which may be an influencing factor in making the system available to a maximum number of patients. Since one cannot generally say that E. coli with OmpA-long are more “dangerous” than E. coli with OmpA-short (or vice versa), such simple system may be sufficient to facilitate translation of the invention into the clinic as quickly as possible. In particular embodiments, the methods and compositions provided herein are used in the testing of water samples for the presence of pathogenic E. coli strains, which is of particular utility to the public health. In particular embodiments, the first binding agent consists of the first polypeptide and a fluorescent dye, the second binder where present, also consists of the second polypeptide and a fluorescent dye. The dye is attached to the first or second polypeptide, respectively through a linker having a molecular weight of 1200 u or less, particularly through a linker having a molecular weight of ≤1000 u, ≤800u, ≤600u or ≤400u. A broad range of linkers with different functionalities facilitating their attachment to a dye and polypeptide, is commercially available, ranging in size from 1200 to ≤400u. One such source is BroadPharm, San Diego, CA, USA, but many alternatives exist. The inventors have shown in the examples that a short (1-3 amino acid and / or oligo-ethylenglycol linker) brings excellent results. In certain embodiments, the composition according to the invention further comprises a chelator of divalent cations. The examples show the use of EDTA (ethylenediaminetetraacetic acid). Several alternatives to EDTA may be used to chelate divalent cations, which include: - EGTA (ethylene glycol tetraacetic acid): EGTA is similar to EDTA but has a higher affinity for calcium ions over magnesium ions. - DTPA (diethylenetriaminepentaacetic acid): DTPA is a chelating agent with a structure similar to that of EDTA but has five carboxymethyl groups instead of four. - NTA (nitrilotriacetic acid): NTA is a weaker chelator than EDTA but is biodegradable and often used in detergents and cleaning agents. It can chelate a variety of divalent cations. - Citrates: Citric acid and its salts can chelate metal ions, and additionally serve as a buffer. - Phosphates: Phosphate ions can form complexes with divalent cations, especially with calcium. Phosphates are used as buffering agents in biological systems. - Oxalates: Oxalic acid and its salts are effective in chelating calcium and other divalent cations. Oxalates are used in analytical chemistry for the determination of calcium and other metal ions. - HEDTA (hydroxyethyl ethylenediaminetriacetic acid): HEDTA is a less commonly used chelator that can bind to divalent cations. It is used in certain specific applications where EDTA might not be suitable. In certain embodiments, the composition according to the invention further comprises an anticoagulant agent, particularly an anticoagulant agent selected from citrate phosphate dextrose, sodium polyanethole sulfonate (SPS) and heparin. The composition according to any one of the preceding claims 1 to 6, 27 to 30, comprising additionally a third binding agent comprising a third single-chain polypeptide capable of specific binding to E. coli OmpF. In certain embodiments, the third single-chain polypeptide comprises a sequence selected from the group consisting of SEQ ID NO 056, 057, 058, 059, 060, 061, 062, 063. Another aspect of the invention relates to a method for analysis of sample of an aqueous fluid (water, cell culture media, bacteria culture media, blood, urine, homogenized food samples, resuspended swabs, drinking water, sewage a water sample taken from a swimming location; blood plasma; extracellular body fluids (ECF), including, but not limited to: cerebrospinal fluid that bathes the brain and spinal cord, lymph, the synovial fluid in joints, intestinal fluids), comprising a. contacting the sample with a composition according to any one of the aspects or embodiments disclosed herein, wherein the first binding agent and the second binding agent contained in the composition are attached to a detectable label, and b. detecting the presence of E. coli in said sample by determining the distribution of the label within the sample. In particular embodiments, the detectable label is a fluorescent dye. In particular embodiments, the distribution of the label within the sample is determined by flow cytometry. Yet another aspect of the invention relates to a method for isolating E. coli bacteria from a sample, said method comprising a. contacting a sample with a composition or binding agent that comprises a polypeptide capable of selectively binding to OmpA as described in any one of the preceding claims, wherein the binding agent comprises an attachment moiety capable of selectively binding to a surface, and b. contacting the sample with a surface to which the attachment moiety can bind; c. removing the sample and isolating E. coli bacteria from the surface. The method according to claim 36, wherein the attachment moiety is biotin, and the surface comprises streptavidin moieties. In certain embodiments, the polypeptide capable of selectively binding to OmpA and the attachment moiety are separated by a linker having a molecular weight of 600 u or less, particularly through a linker having a molecular weight of ≤1200 u ≤1000 u, ≤800u, ≤600u or ≤400u. In certain other embodiments, the polypeptide capable of selectively binding to OmpA and the attachment moiety are separated by a peptide linker of ≥5 nm length, particularly of ≥10 nm. In certain embodiments, the linker is characterized by alpha-helical structure. The LEA-linker, with its 46 residues, is of alpha-helical structure and covers a distance of approximately 7 nm. Assuming fully extended peptides, adjacent residues are 0.35 nm apart. Therefore, the PAPA-linker with 33 residues, of mainly rigid, unstructured secondary structure, is maximally 11.5 nm long. A fully extended GS-linker would cover 7 nm, but its flexible nature likely makes it shorter. In certain embodiments, the linker is a LEA linker comprised of 46 residues (alpha helix) (see SEQ ID NO 069). In certain embodiments, the linker is a PAPA linker comprised of 33 residues (see SEQ ID NO 070). In certain embodiments, the linker is a heavy chain only antibody containing the FC part and five fused single-domain antibodies comprising 209 residues. In particular embodiments, the polypeptide capable of selectively binding to OmpA and the attachment moiety are separated by an alpha-helical peptide linker of 20 or more amino acids, particularly 30-240 amino acids. The invention further encompasses the following embodiments: 1. A composition comprising: a. a first binding agent comprising a first single-chain polypeptide capable of specific binding to a first epitope variant of OmpA (SEQ ID NO 054; OmpA-short) and b. a second binding agent comprising a second single-chain polypeptide capable of specific binding to a second epitope variant of OmpA (SEQ ID NO 055; OmpA- long) wherein said first and second binding agent are: i) linked to a detectable label; or ii) attachable, or attached, to a surface. 2. The composition according to claim 1, wherein the first single-chain polypeptide does not bind to the second epitope variant, and the second single-chain polypeptide does not bind to the first epitope variant. 3. The composition according to embodiment 1 or 2, wherein - the first single-chain polypeptide is capable of binding specifically to the first epitope variant of OmpA (SEQ ID NO 054) when said first epitope variant is present on the surface of a E. coli bacterium, and wherein - the second single-chain polypeptide is capable of binding specifically to the second epitope variant of OmpA (long, SEQ ID NO 055) when said second epitope variant is present on the surface of a E. coli bacterium. 4. The composition according to any one of the preceding embodiments 1 to 3, wherein c. the first single chain polypeptide comprises a CDR1, CDR2 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 014 [GTGFTFSKSPMS] and SEQ ID NO 015 [GSGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is selected from SEQ ID NO 017 [RRLGKTTYDY] and SEQ ID NO 018 [RRLGKRTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to one of SEQ ID NO 014, 015, combined with SEQ ID NO 016 and with one of SEQ ID NO 017 and 018, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01), and / or d. the second single chain polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 043 [GSIFNFNPMG], SEQ ID NO 044 [GSIFNFNLMG], SEQ ID NO 045 [GGIFNFNIMG], and SEQ ID NO 046 [GSIFNFNIMG] and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to either SEQ ID NO 043, 044, 045 or 046, and to 047, and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). 5. The composition according to embodiment 4, wherein c. the first single chain polypeptide comprises a CDR1, CDR2 and CDR3 sequence and wherein CDR1 is SEQ ID NO 014 [GTGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is SEQ ID NO 017 [RRLGKTTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 014, 016 and 017 respectively, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01), and / or d. the second single chain polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is SEQ ID NO 043 [GSIFNFNPMG], and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to SEQ ID NO 043 and 047 and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). 6. The composition according to embodiment 4, wherein c. the first single chain polypeptide comprises a sequence selected from: i. SEQ ID NO 001 (Nb01) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYY SDSVRGRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS; ii. SEQ ID NO 002 (Nb02) SQMQFVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGQGTPVT VS iii. SEQ ID NO 003 (Nb03) SQGQLVESGGGLVPPGGSLRLSCAVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYSDSVRGRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVT VS iv. SEQ ID NO 004 (Nb04) SQRQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGKGTPVT VS v. SEQ ID NO 005 (Nb05) SQRQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVT VS vi. SEQ ID NO 006 (Nb06) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYSDSVRGRFTISRDNAKNTVYLEMNNVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVT VS vii. SEQ ID NO 008 (Nb08) SQMQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGYRRLGKTTYDYRGQGTPVT VS or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 001, 002, 003, 004, 005, 006 or 008, and having the same biological activity as SEQ ID NO 001 (Nb01), and / or d. the second single chain polypeptide comprises a sequence selected from: i. SEQ ID NO 039 (Nb39) SVKGRFTISRDSAKNTLYLQMTDVKPGDAAVYICHANYRIGRNDLPVWGKGTPVTVS] ii. SEQ ID NO 038 (Nb38) SQRQLVESGGGTVQTGGSLRLSCVPNGSIFNFNLMGWYRQSSGQQRELVATLTRDGSEN YAEFVKGRFTISRDSGKNTMYLQMTDVKPSDTAVYICHANYRIGRNDLPVWGKGTRVTV S iii. SEQ ID NO 040 (Nb40) SQRQLVESGGGLVQPGGSLRLSCVPNGSIFNFNIMGWYRQNAGNQRELVATMTRDGSAS YSDSVKGRFTISRDVDKNTIYLQLDSVKPEDTAVYICHANYRIGRNDLPVWGRGTRVTV S iv. SEQ ID NO 042 (Nb42) SQLQLVESGGGLVQPGGSLRLSCVPNGGIFNFNIMGWYRQNAGNQRELVATMTRDGSAS YSDSVKGRFTISRDVDKNTIYLQMDSVEPEDTAVYICHANYRIGRNDLPVWGQGTPVTV S or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 039, 038, 040, or 042, and having the same biological activity as SEQ ID NO 039 (Nb39). 7. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is selected from SEQ ID NO 014 [GTGFTFSKSPMS] and SEQ ID NO 015 [GSGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is selected from SEQ ID NO 017 [RRLGKTTYDY] and SEQ ID NO 018 [RRLGKRTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to one of SEQ ID NO 014, 015, combined with SEQ ID NO 016 and with one of SEQ ID NO 017 and 018, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01). 8. The binding agent according to embodiment 7, wherein CDR1 is SEQ ID NO 014 [GTGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is SEQ ID NO 017 [RRLGKTTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 014, 016 and 017, and the single chain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01). 9. The binding agent according to embodiment 7 or 8, wherein the first single-chain polypeptide comprises a sequence selected from: i. SEQ ID NO 001 (Nb01) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSDSVR GRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS; ii. SEQ ID NO 002 (Nb02) SQMQFVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYAD SVKGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGQGTPVTVS iii. SEQ ID NO 003 (Nb03) SQGQLVESGGGLVPPGGSLRLSCAVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSD SVRGRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS iv. SEQ ID NO 004 (Nb04) SQRQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYAD SVKGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGKGTPVTVS v. SEQ ID NO 005 (Nb05) SQRQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYAD SVKGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVTVS vi. SEQ ID NO 006 (Nb06) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYSD SVRGRFTISRDNAKNTVYLEMNNVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVTVS vii. SEQ ID NO 008 (Nb08) SQMQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYYAD SVKGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGYRRLGKTTYDYRGQGTPVTVS or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO SEQ ID NO 001, 002, 003, 004, 005, 006 or 008, and having the same biological activity as SEQ ID NO 001 (Nb01). 10. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 019 [GLAFSSYASG], CDR2 is SEQ ID NO 020 [AMTARGGFTN], and CDR3 is SEQ ID NO 021 [DPRRYGSQVY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 019, 020 and 021, and the single chain polypeptide has the same biological activity as SEQ ID NO 007 (full sequence of Nb07). 11. The binding agent according to embodiment 10, wherein the first single-chain polypeptide comprises SEQ ID NO 007 (Nb07) [SQGQLLESGGGLVQPGGSLRLSCAASGLAFSSYASGWYRQAPGKERELVAAMTARGGFTNYADSV KGRFTISRDNGKNTVYLQMNSLKPEDTAVYYCNADPRRYGSQVYWGQGTPVTVS], or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 007, and having the same biological activity as SEQ ID NO 007 (Nb07). 12. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 022 [GFRFSSVALS], CDR2 is SEQ ID NO 023 [LITNDHKSR], and CDR3 is SEQ ID NO 024 [SLLGSNYGRY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 022, 023 and 024, and the single chain polypeptide has the same biological activity as SEQ ID NO 009 (full sequence of Nb09). 13. The binding agent according to embodiment 12, wherein the first single-chain polypeptide comprises SEQ ID NO 009 (Nb09) [SQLQFVESGGGMVQPGGSLRLSCVASGFRFSSVALSWYRQAPGRDRELVALITNDHKSRYGDFVK DRFTISRDNNKNTVYLQMNNLKPEDTAVYSCGVSLLGSNYGRYWGKGTRVTVS], or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 009, and having the same biological activity as SEQ ID NO 009 (Nb09). 14. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 025 [GFTFSSYAMT], CDR2 is SEQ ID NO 026 [LITNDARTR], and CDR3 is SEQ ID NO 027 [SLLGRNYGQH], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 025, 026 and 027, and the single chain polypeptide has the same biological activity as SEQ ID NO 010 (full sequence of Nb10). 15. The binding agent according to embodiment 14, wherein the first single-chain polypeptide comprises SEQ ID NO 010 (Nb10) [SQGQLVESGGGLVQPGGSLRLSCTASGFTFSSYAMTWHRQAPGKERELVALITNDARTRYGDFVK GRFTISRDNAKNTIYLQMNTLAPEDTALYYCGVSLLGRNYGQHWGKGTRVTVS], or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 010, and having the same biological activity as SEQ ID NO 010 (Nb10). 16. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 028 [GRTFNPYAMG], CDR2 is SEQ ID NO 029 [AIRWSGGSIN], and CDR3 is SEQ ID NO 030 [AKSLGVWAREYDD], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 028, 029 and 030, and the single chain polypeptide has the same biological activity as SEQ ID NO 011 (full sequence of Nb11). 17. The binding agent according to embodiment 8, wherein the first single-chain polypeptide comprises SEQ ID NO 011 (Nb11) [SQGQLVESGGGLVQAGGSLRLSCAASGRTFNPYAMGWFRQAPGKEREFVAAIRWSGGSINYADSV KGRFTISRDNAKNTVYLQMNSLKPEDTAVYYCNAAKSLGVWAREYDDWGKGTPVTVS], or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 011, and having the same biological activity as SEQ ID NO 011 (Nb11). 18. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 031 [GSTSSIDTMS], CDR2 is SEQ ID NO 032 [SITKGGGRPY], and CDR3 is SEQ ID NO 033 [RDSNSGFYY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 031, 032 and 033, and the single chain polypeptide has the same biological activity as SEQ ID NO 012 (full sequence of Nb12). 19. The binding agent according to embodiment 8, wherein the first single-chain polypeptide comprises SEQ ID NO 012 (Nb12) [SQRQLVESGGGLVQAGGSLRLSCAASGSTSSIDTMSWYRQTTGNEREMVASITKGGGRPYYDFSV KGRFTISRDNDKNTMSMQMNSLKPEDTAVYYCNARDSNSGFYYWGKGTPVTVS], or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 012, and having the same biological activity as SEQ ID NO 012 (Nb12). 20. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is SEQ ID NO 034 [GFTFSSALMR], CDR2 is SEQ ID NO 035 [YVNNKDLYTF], and CDR3 is SEQ ID NO 036 [ASTPNT], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 034, 035 and 036 respectively, and the single chain polypeptide has the same biological activity as SEQ ID NO 013 (full sequence of Nb13). 21. The binding agent according to embodiment 8, wherein the first single-chain polypeptide comprises SEQ ID NO 013 (Nb13) [SQGQFVESGGALVQPGGSLRLSCAASGFTFSSALMRWYRQAAGKERELVAYVNNKDLYTFYVDSV KGRFTISRDNAKNTVYLQMNSLKPEDTAVYYCNAASTPNTWGQGTRVTVS], or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 013, and having the same biological activity as SEQ ID NO 013 (Nb13). 22. A binding agent comprising a polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli, wherein the polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 043 [GSIFNFNPMG], SEQ ID NO 044 [GSIFNFNLMG], SEQ ID NO 045 [GGIFNFNIMG], and SEQ ID NO 046 [GSIFNFNIMG] and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to either SEQ ID NO 043, 044, 045 or 046, and to 047, and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). 23. The binding agent according to embodiment 22, wherein CDR1 is SEQ ID NO 043 [GSIFNFNPMG], and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to SEQ ID NO 043 and 047 respectively and the single chain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39). 24. The binding agent according to embodiment 22 or 23, wherein the second single chain polypeptide comprises a sequence selected from: i.SEQ ID NO 039 (Nb39) RFTISRDSAKNTLYLQMTDVKPGDAAVYICHANYRIGRNDLPVWGKGTPVTVS] ii.SEQ ID NO 038 (Nb38) SQRQLVESGGGTVQTGGSLRLSCVPNGSIFNFNLMGWYRQSSGQQRELVATLTRDGSENYAEFV KGRFTISRDSGKNTMYLQMTDVKPSDTAVYICHANYRIGRNDLPVWGKGTRVTVS iii.SEQ ID NO 040 (Nb40) SQRQLVESGGGLVQPGGSLRLSCVPNGSIFNFNIMGWYRQNAGNQRELVATMTRDGSASYSDSV KGRFTISRDVDKNTIYLQLDSVKPEDTAVYICHANYRIGRNDLPVWGRGTRVTVS iv.SEQ ID NO 042 (Nb42) SQLQLVESGGGLVQPGGSLRLSCVPNGGIFNFNIMGWYRQNAGNQRELVATMTRDGSASYSDSV KGRFTISRDVDKNTIYLQMDSVEPEDTAVYICHANYRIGRNDLPVWGQGTPVTVS or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 039, 038, 040 or 042, and having the same biological activity as SEQ ID NO XXX (Nb39). Single Binders to OmpA-long different from the Nb39 family of binders: 25. A binding agent comprising a polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli, wherein the polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is SEQ ID NO 051 [DNILQFGNMG], and CDR3 is SEQ ID NO 053 [QYVIGRNRLDV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to SEQ ID NO 051 and 053 respectively and the single chain polypeptide has the same biological activity as SEQ ID NO 041 (full sequence of Nb41). 26. The binding agent according to embodiment 25, wherein the second single chain polypeptide comprises SEQ ID NO 041 (Nb41) [SQRQLVESGGGSMQPGESLTLSCEASDNILQFGNMGWYRQSPGTQRELVARIHKRGDSDYGDFAK GRFTISRDTVKNKVYLQMTDLKPEDSANYICNGQYVIGRNRLDVWGQGTPVTVS]; or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 041, and having the same biological activity as SEQ ID NO 041 (Nb41). 27. The composition according to any one of embodiments 1 to 6, or the binding agent according to any one of embodiments 7 to 26, wherein the binding agent is linked to a fluorescent dye. 28. The composition or binding agent according to embodiment 27, wherein the binding agent consists of the polypeptide and a fluorescent dye, the dye being attached to the polypeptide through a linker having a molecular weight of 1200 u or less, particularly through a linker having a molecular weight of ≤1000 u, ≤800u, ≤600u or ≤400u. 29. The composition according to any one of the preceding embodiments 1 to 6, 27 or 28, further comprising a chelator of divalent cations. 30. The composition according to any one of the preceding embodiments 1 to 6, 27 to 29, further comprising an anticoagulant agent, particularly an anticoagulant agent selected from citrate phosphate dextrose, sodium polyanethole sulfonate (SPS) and heparin. 31. The composition according to any one of the preceding embodiments 1 to 6, 27 to 30, comprising additionally a third binding agent comprising a third single-chain polypeptide capable of specific binding to OmpF. 32. The composition according to embodiment 31, wherein the third single-chain polypeptide comprises a sequence selected from the group consisting of SEQ ID NO 056, 057, 058, 059, 060, 061, 062, 063. 33. A method for analysis of sample, said method comprising a. contacting the sample with a composition according to any one of embodiments 1 to 6, 27 to 32, wherein the first binding agent and the second binding agent are attached to a detectable label, and b. detecting the presence of E. coli in said sample by determining the distribution of the label within the sample. 34. The method according to embodiment 33, wherein the detectable label is a fluorescent dye. 35. The method according to embodiment 33 or 34, wherein the distribution of the label within the sample is determined by flow cytometry. 36. A method for isolating E. coli bacteria from a sample, said method comprising a. contacting a sample with a composition or binding agent that comprises a polypeptide capable of selectively binding to OmpA as described in any one of the preceding embodiments, wherein the binding agent comprises an attachment moiety capable of selectively binding to a surface, and b. contacting the sample with a surface to which the attachment moiety can bind; c. removing the sample and isolating E. coli bacteria from the surface. 37. The method according to embodiment 36, wherein the attachment moiety is biotin, and the surface comprises streptavidin moieties. 38. The method according to any one of embodiments 36 or 37, wherein the polypeptide capable of selectively binding to OmpA and the attachment moiety are separated by a linker having a molecular weight of 600 u or less, particularly through a linker having a molecular weight of ≤1200 u ≤1000 u, ≤800u, ≤600u or ≤400u. 39. The method according to any one of embodiments 36 or 37, wherein the polypeptide capable of selectively binding to OmpA and the attachment moiety are separated by a peptide linker of ≥5 nm length, particularly of ≥10 nm. 40. The method according to any one of embodiments 36 or 37, wherein the polypeptide capable of selectively binding to OmpA and the attachment moiety are separated by an alpha-helical peptide linker of 20 or more amino acids, particularly 30-240 amino acids. Wherever alternatives for single separable features are laid out herein as “embodiments”, it is to be understood that such alternatives may be combined freely to form discrete embodiments of the invention disclosed herein. The invention is further illustrated by the following examples and figures, from which further embodiments and advantages can be drawn. These examples are meant to illustrate the invention but not to limit its scope. Description of the Figures Fig.1 shows species coverage of OmpA nanobodies. a. Cellular binding assays were carried out using AlexaFluor647-labeled nanobodies. b. Cellular binding assays using Nb01 and Nb39. Strains expressing OmpA-short and the OmpA-long isoforms are indicated in the legend. ND: not determined. c. and d. Sequence analysis of extracellular loop regions of OmpA in the Swiss (c) or 661k (d) dataset. Unique OmpA sequences are listed according to their frequency found in the respective database. The reference OmpA-short and OmpA-long sequences used to generate Nb01 and Nb39, respectively, are the most frequent ones in both datasets. Detected, undetected, and unknown detection (grey) indicates whether the respective OmpA variant is recognized by either Nb01 or Nb39 Fig.2 shows nanobody characterization using flow cytometry. a. and b. Flow cytometry histograms showing binding of Nb01-AF647 (a) or Nb39-AF647 (b) to the indicated bacterial strains. For these experiments, 33 nM of fluorescently labeled nanobodies were used. Measurements were performed on an Aurora flow cytometer (Cytek) using unfixed cells. c. and d. Concentration-dependent binding of Nb01-AF647 against the indicated strains (c) or against CS #2 in the presence of NaCl or different anticoagulants (CPD, EDTA, or Heparin) (d). Data were fitted with the Hill equation to determine EC50 values. Cells were fixed and samples measured in triplicates on the CytoFlex S flow cytometer (Beckman). e. Binding of Nb01-AF647 (10 nM) to the indicated strains grown in Luria Broth (LB) or Lactose Broth (LacB) media. Measurements were performed in triplicates on an Aurora flow cytometer (Cytek) using unfixed cells. For all data shown, earth mover’s distances (EMDs) were calculated relative to binding of an unrelated nanobody at 10 nM (see methods). Representative data of three biological replicates are shown. Errors correspond to standard deviations of three biological replicates, which correspond to three independently inoculated cultures grown and processed in parallel. CS, clinical strain; wt, wildtype; EC50, half maximal effective concentration. Fig.3 shows immunomagnetic capture of E. coli. a. Experimental scheme of capture assay. In procedure 1, magnetic streptavidin beads are saturated with biotinylated nanobodies in the first step. Next, they are added to a solution with bacteria for capture. In procedure 2, bacteria are first incubated with biotinylated nanobodies and in a second step, magnetic streptavidin beads capture the tagged bacteria. b. Schematic illustration of nanobody constructs. Nb01 is colored light grey, Nb18 dark grey and non-related nanobodies in grey. Biotin is depicted as chemical structure. PEG11 is shown as zagged line. Stiff linkers are shown as α-helices, flexible glycine-serine linkers as lines. Fc-part of antibodies are shown as grey, rounded rectangles. c. and d. Time-dependent capture of E. coli lab strain MC1061 according to procedure 1 using Nb01-PEG11-Biotin and 1. Error bars correspond to standard deviations of biological triplicates (except the indicated data point). e. Capture of E. coli cells (ca.500 CFU) cultured in LB or LacB using Nb01-PEG11- Biotin (see icon). Capture was performed for 2 h according to procedure 1. f. Capture of E. coli cells (ca.500 CFU) cultured in LB or LacB using Nb01-GS-Nb18- LEA-Biotin (see icon). Capture was performed for 2 h according to setup 1. g. Capture of E. coli cells (ca.500 CFU) cultured in LB using 4Nb01-hcAb-Biotin (see icon) or 4Nb39-hcAb-Biotin, as indicated. Capture was performed for 1 h according to procedure 2. EDTA (0.5 mM) was supplemented in capture buffers for the indicated sample. Examples Example 1: Selection of OmpA as target protein We identified OmpA as a promising surface-accessible target to generate nanobodies for staining and capture of living E. coli cells. OmpA is among the most abundant E. coli proteins, and fairly conserved at the protein level. In addition, it is non-essential, which is relevant for the live capture of E. coli for phenotypic antimicrobial susceptibility testing. OmpA has also been identified as a target for several bacteriophages showing that it can be bound by large molecules. To gain insights into the prevalence of OmpA in E. coli genomes, we performed a bioinformatics analysis of the 661k database, which contains approximately 90,000 assembled E. coli genomes. Our bioinformatics pipeline (see methods) identified the ompA gene in 95 % of these E. coli genomes. The respective OmpA protein sequences clustered into two main groups in the phylogenetic tree, as well as in a hierarchical clustering of the loop variants, corresponding to the two isoforms described earlier. In this study, we refer to these isoforms as OmpA-short and OmpA-long, because their extracellular loop 3 differs in length. Within these two isoforms, the OmpA sequences are highly conserved. Example 2: Generation of nanobodies recognizing OmpA in the native context of E. coli The most frequent OmpA-short and OmpA-long sequences were cloned, expressed, and purified to immunize two alpacas (see methods). After RNA extraction from lymphocytes, two consecutive rounds of phage display were carried out (Zimmermann et al., 2018 ibid), (Zimmermann et al., Nat Protoc, Vol.15, No.5, pp. 1707-1741, 2020), followed by deep screening using the flycode technology (Egloff et al., Nat Methods, Vol.16, No.5, pp. 421-428, 2019) to identify nanobodies that bind OmpA in the native context of E. coli. Flycodes are genetically encoded peptide barcodes optimized for detection by mass spectrometry. Approximately 2,000 nanobodies pre-enriched by phage display were tagged with flycodes, deep-sequenced and screened against different E. coli strains. Our strain set included the laboratory E. coli K-12 strain and its isogenic ompA knockout, both devoid of O-antigens, as well as six clinical E. coli isolates shielded by different O-antigens and expressing either the short (SEQ ID NO 054) or long (SEQ ID NO 055) OmpA isoform. Thereby, we identified several nanobodies, most notably among them those assigned our internal codes Nb01 and Nb39, which according to our quantitative flycode analysis, exhibit high binding affinity and specificity for OmpA-short or OmpA-long, respectively, in the cellular context of clinical E. coli strains. Example 3: Validation of Nb01 and Nb39 in cellular binding assays Nb01 and Nb39 were cloned and purified as individual proteins and their ability to specifically bind to their respective OmpA isoforms was analyzed in a cellular binding assay using fluorescently labeled nanobodies against the same set of E. coli strains as used for flycode screening. For the cellular binding experiments, we either labeled the nanobody directly with AlexaFluor 647 (AF647) via a cysteine introduced at the C-terminus or indirectly via an NTA-Biotin linker bound to Atto565- labeled streptavidin, which recognizes the C-terminal His-tag of the respective nanobody. Since E. coli MC1061 expresses OmpA-short, we complemented the E. coli MC1061 ΔompA strain with a plasmid expressing OmpA-long to characterize Nb39. When directly labeled with AF647, both Nb01 and Nb39 recognized all E. coli strains including the clinical isolates expressing the respective OmpA isoform, showing that these nanobodies can stain a diverse set of strains (Fig.1a and b). In contrast, when performing the same assay with indirect labeling via streptavidin, only the lab strain E. coli MC1061 devoid of O-antigen and clinical strain #1 (CS#1) were recognized. This shows that most clinical strains possess a dense O-antigen layer that prevents the bulky nanobody-streptavidin complex from binding to OmpA, thereby directly demonstrating the advantage of using small nanobodies for the staining and capture of clinical E. coli strains. Intriguingly, the nanobodies were found to be highly specific. Nb01 did not recognize any of the strains carrying the OmpA-long isoform, and conversely, Nb39 did not recognize strains expressing OmpA-short. E. coli MC1061 ΔompA was not recognized by any of the nanobodies (Fig.1b). Example 4: Structural characterization of nanobodies To gain insights into the binding geometry of Nb01 and Nb39 and to rationalize their specificity and species coverage, we determined crystal structures of Nb01-OmpA-short and Nb39-OmpA-long, both resolved at a resolution of around 2.3 Å. The structures revealed that both nanobodies bind to the surface-accessible epitope of OmpA, consistent with their ability to recognize OmpA in the cellular context. A closer inspection of the binding interface by PDBePISA (https: / / www.ebi.ac.uk / pdbe / pisa / ) revealed that the nanobodies establish interactions with all four extracellular loops of OmpA, including extensive interactions with the variable loop 3, thereby explaining the high specificity for the respective isoform. Nb01 establishes contacts with all its three CDRs, whereas Nb39 interacts via CDR1 and CDR3 only. The complex interfaces are 548 Å2and 658 Å2for Nb01-OmpA-short and Nb39-OmpA-long, respectively. Example 5: Sequence conservation analysis reveals high species coverage To obtain detailed insights into the species coverage of our nanobodies, we performed a comprehensive sequence analysis of OmpA variants in clinical E. coli isolates. The analysis is based on two databases. The first database stems from the University Hospital Basel (USB) and the Institute of Medical Microbiology at the University of Zurich (IMM), representing the German- speaking part of Switzerland. It is hereafter named Swiss dataset and contains 2’034 E. coli strains encoding a complete OmpA sequence (Fig. 1c). Importantly, we had physical access to these strains for experimental validation. The second source of sequences is the 661k database (Fig.1d) (Blackwell et al., PLos Biol, vol.19, no.1, 2021), which contains over 661,000 assembled bacterial genomes and 85,680 E. coli strains containing ompA. We performed a protein sequence analysis of the extracellular loops (underlined sequences in SEQ ID NO 054 and 055): (SEQ ID NO 054; OmpA-short) MKKTAIAIAVALAGFATVAQAAPKDNTWYTGAKLGWSQYHDTGFINNNGPTHENQLGAGAFG GYQVNPYVGFEMGYDWLGRMPYKGSVENGAYKAQGVQLTAKLGYPITDDLDIYTRLGGMVWR ADTKSNVYGKNHDTGVSPVFAGGVEYAITPEIATRLEYQWTNNIGDAHTIGTRPDNGMLSLG VSYRFGQGEAAPVVAPAPAPAPEVQTKHFTLKSDVLFNFNKATLKPEGQAALDQLYSQLSNL DPKDGSVVVLGYTDRIGSDAYNQGLSERRAQSVVDYLISKGIPADKISARGMGESNPVTGNT CDNVKQRAALIDCLAPDRRVEIEVKGIKDVVTQPQA SEQ ID NO 055; OmpA-long) (MKKTAIAIAVALAGFATVAQAAPKDNTWYTGAKLGWSQYHDTGFIPNNGPTHENQLGAGAFG GYQVNPYVGFEMGYDWLGRMPYKGDNINGAYKAQGVQLTAKLGYPITDDLDIYTRLGGMVWR ADTKANVPGGASFKDHDTGVSPVFAGGVEYAITPEIATRLEYQWTNNIGDAHTIGTRPDNGM LSLGVSYRFGQGEAAPVVAPAPAPAPEVQTKHFTLKSDVLFTFNKATLKPEGQAALDQLYSQ LSNLDPKDGSVVVLGYTDRIGSDAYNQGLSERRAQSVVDYLISKGIPADKISARGMGESNPV TGNTCDNVKQRAALIDCLAPDRRVEIEVKGIKDVVTQPQA We identified 32 unique OmpA loop combinations in the Swiss dataset (Fig.1c), and 179 unique OmpA loop combinations in the 661k database (Fig.1d). By combining this sequence information with testing representative strains in a high-throughput binding assay based on flow cytometry, we categorized the strains into three groups: recognized and thus covered by one of our two nanobodies (detected), not recognized (not detected), and unknown (when no representative strain was tested or available). This analysis was conducted for both our Swiss dataset and the 661k dataset, respectively (Fig.1c and d). Remarkably, with our two OmpA nanobodies, we achieved a species coverage of 79 % of the total 2,034 available E. coli strains from the Swiss database and 91 % of the 85,680 E. coli strains from the 661k database. We identified five OmpA variants which collectively account for 97 % of the undetected strains of the Swiss database and 92 % of the undetected E. coli strains of the 661k database (largest five Not Detected bars in Fig.1c and d respectively). For simplicity, we call these non-detected variants OmpA-ND#1-5. A closer inspection of the respective loop regions of these non-detected variants revealed that the deviating residues map to positions which interact with the nanobodies in the crystal structures. For example, the most frequent non-detected OmpA variant (OmpA-ND#1), an OmpA-short isoform variant which accounts for 54 % of the non-detected strains in the Swiss collection and 46 % of the non-detected strains in the 661k dataset, features a NFDG motif in loop 3 instead of the more frequent NVYG motif. As the crystal structure revealed, the entire NVYG motif is recognized by Nb01 (SEQ ID NO 001), thereby explaining why deviations in this core region of the OmpA-nanobody interface are not tolerated. OmpA-ND#5 is quite similar to OmpA-ND#1 and only contains a conservative asparagine to aspartate mutation in the first loop, a mutation which was shown to be tolerated by Nb01. A binder against OmpA-ND#1 is therefore expected to detect also strains expressing OmpA-ND#5. Example 6: Nb01 and Nb39 do not cross-react with other bacterial species We performed a sequence analysis using hidden Markov models (see methods) to assess potential cross-reactions of our nanobodies by screening for protein sequences similar to each extracellular OmpA loop i) within other E. coli proteins and ii) within proteins of other bacteria. Within E. coli, no other matches were found. However, for other bacterial taxa, while no matches were found for loop 3, multiple matches were detected in OmpA homologues in Enterobacteriaceae, in particular in Salmonella and Klebsiella strains. In an orthogonal approach, we used our high-throughput flow cytometry approach to probe Nb01 and Nb39 against 19 reference strains belonging to species other than E. coli, of which 10 contain an annotated OmpA sequence. In addition, we tested our nanobodies against eleven reference strains for E. coli as well as three Shigella strains (which technically are specialized diarrheagenic E. coli strains). Remarkably, our nanobodies demonstrated to bind all E. coli reference strains and the three Shigella strains. Of further note, the OmpA-short and OmpA-long isoforms appear to be distributed quite evenly over the tested reference strain set, showing that nanobodies against both isoforms are needed to reach satisfactory species coverage. No cross-reactive binding to any of the tested non-E. coli strains was observed. This finding correlates well with the bioinformatics analysis, as none of the available OmpA sequences matched within loop 3 of the two E. coli OmpA reference sequences. Example 7: Nanobodies bind their target protein with high affinity To characterize binding of Nb01 and Nb39 in the cellular context of the E. coli lab strain and clinical isolates, we employed flow cytometry using fluorescently labeled nanobodies (Fig. 2). Both nanobodies showed strong OmpA-specific binding within the native bacterial cell environment, even in the presence of O-antigens found on clinical strains (Fig. 2a and b). Notably, both nanobodies exhibited high specificity for the targeted OmpA isoform. While Nb01 did not show any binding to strains expressing OmpA-long (CS#6 and CS#11), its binding to OmpA-short (expressed on CS#1 and CS#2) was robust (Fig.2a). Conversely, Nb39 showed binding to CS#6 and CS#11, but not to CS#1 and CS#2 (Fig.2b). To obtain values for the apparent binding affinity of Nb01, we determined EC50 values, i.e. the nanobody concentration at which the half-maximal binding signal in the flow cytometry analysis is reached. For the E. coli lab strain MC1061, an EC50 of 12.95 ± 3.24 nM was determined (Fig.2c). In contrast, the EC50 values were around 2-5 times weaker for clinical isolates CS#1 and CS#2. Since the extracellular loops of OmpA-short of CS#1 and CS#2 do not differ from the reference OmpA-short sequence present in E. coli MC1061, the weaker EC50 values can be attributed to the presence of the O-antigen layer in these strains. To test this hypothesis, we pre-treated CS#2 with EDTA, which destabilizes the LPS by chelating divalent cations and which is also used as anticoagulant in blood samples. Indeed, supplementation with EDTA during staining resulted in a ten-fold enhancement of the EC50 values for Nb01 targeting CS#2 (Fig.2d). Additionally, we tested the effects of the anticoagulants citrate phosphate dextrose (CPD) and heparin, and all of which improved EC50 values for Nb01 on CS#2 (Fig.2d). We reasoned that in the case of clinical strains not treated with anticoagulants, the nanobodies need more time to reach their binding epitope on OmpA. This would manifest in slower binding on- rates as compared to the lab strain. To test this hypothesis, we pre-incubated the nanobodies for variable durations (10 – 120 min) against CS#2 prior to FACS analysis. However, the resulting EC50 values did not significantly change, showing that the weaker EC50 values observed in case of clinical strains versus lab strain cannot be explained by slower binding on-rates. Example 8: LPS density depends on growth medium By serendipity, we discovered that when clinical E. coli strains were grown in Lactose Broth (LacB) instead of Luria Broth (LB), staining with AF647-labeled nanobody is drastically enhanced, particularly for CS#2 (Fig.2e). We suspected that the change of growth medium alters the LPS structure. To validate this hypothesis, we extracted the LPS from three clinical strains grown either in LacB or LB in the presence or absence of EDTA, separated them on an SDS-PAGE gel, and stained the O-antigen sugars with Emerald dye. We used the lab strain E. coli MG1655 as a reference strain devoid of O-antigens. As positive control, we included the E. coli MG1655 wbbL+ strain, whose O-antigen production was genetically restored (Liu et al., Microbiology, vol.140, pp. 49-57,1994). When grown in LB, we noted a large variability among the clinical isolates in terms of the amount of O-antigen produced, with clinical strain #1 clearly having the least O-antigen presented on the cellular surface. When including EDTA in the growth medium for clinical strain #1, the density of the O-antigen was somewhat reduced, and shorter sugar chains become more abundant. When grown in LacB, the O-antigen density was strongly reduced for E. coli MG1655 wbbL+ and clinical strains #1 and #2, and to a lesser degree also for clinical strain #6. In addition, a prominent O-antigen ladder was observed for clinical strains #1 and #2. Hence, we noted that the extent of O-antigen surface decoration is highly strain-dependent, and that the growth conditions have a profound effect on the O-antigen density and length distribution. Together with the observation that nanobody staining heavily depends on the growth medium (Fig. 2e), we demonstrated that LPS density strongly impacts OmpA accessibility and thus, nanobody binding. Example 9: Nanobody-mediated capture of E. coli lab strain To establish a bacterial capture assay at a bacterial load that is realistic in patients suffering from BSI, we first diluted the lab strain E coli MC1061 to a cell density of approximately 500 CFU / mL in PBS. For magnetic extraction, we initially immobilized nanobodies via a short maleimide-PEG11- biotin linker (Fig.3b) on magnetic streptavidin beads and then used these functionalized beads to capture bacteria (Fig. 3a, procedure 1). Thereby, we achieved complete capture of the E. coli MC1061 lab strain within two hours, while the corresponding ΔompA strain was not captured (Fig. 3c). Complete capture was also achieved upon lowering the bacterial cell number to 50 CFU / mL (Fig.3d). The large error bars are attributed to the small number of cells involved in this assay. Disappointingly, we failed to capture any of the clinical strains under the same experimental conditions, even upon loosening the LPS layer by adding EDTA to the growth medium or growing the cells in Lactose Broth (Fig.3e). Example 10: Targeting trimeric OmpF with nanobodies With the aim to overcome target accessibility problems encountered when capturing clinical strains decorated with a dense O-antigen layer, we generated nanobodies against the outer membrane protein OmpF. While OmpA exists as monomer in the outer membrane, crystal structures of OmpF as well as biochemical experiments demonstrated its homotrimeric assembly. In addition, the OmpF monomer is a β-barrel consisting of sixteen β-sheets, as opposed to eight β-sheets in case of OmpA. According to a recent molecular simulation study, the footprint of OmpF displacing LPS within the outer membrane is around six times larger than the one for OmpA in Samonella spp., a bacterial species that is closely related to E. coli. Using our nanobody generation pipeline, we targeted purified OmpF, identifying strong binder candidates as based on NestLink analysis (SEQ ID NO 056-068). One of the OmpF nanobodies, called Nb18, was further characterized using flow cytometry, thereby determining an EC50 of 3 nM against the E. coli K12 strain. Binding was specific, as no background binding signal was observed against the E. coli MC1061 ΔompF strain. The maximal binding signal was found more than 10 times lower than that for OmpA, which can be explained by the lower expression level of OmpF compared to OmpA. In attempts to capture E. coli MC1061 with Nb18 immobilized via a short maleimide-PEG11-biotin linker on magnetic streptavidin beads, capture was only partial. This stands in contrast to the near 100 % capture using Nb01 directed against OmpA under identical experimental conditions (Fig.3c), indicating that higher OMP densities are beneficial to achieve efficient capture and that the encounter of functionalized beads with the bacterial cell is likely the rate-limiting factor. Regrettably, capture of clinical strain CS#1 via Nb18-functonalized beads was not successful. Example 11: Linker engineering to bridge the LPS Since our nanobodies efficiently bind to clinical E. coli strains (Fig.2), but fail to capture them (Fig. 3e), we reasoned that steric hindrance caused by the LPS layer is the underlying problem impeding capture of clinical E. coli strains. To gain further experimental insights into the steric hindrance problem, we generated Avi-tagged nanobody constructs with the Avi-tag being separated with linkers of variable lengths and flexibility from the nanobody core (Fig. 3b). The linkers included flexible glycine and serine linkers (GS-linkers), helical structures, containing the leucine-glutamate- alanine sequence(LEA-linker), and stiff linkers containing proline-alanine-proline-alanine repeats (PAPA linker). As one turn with 3.6 amino acids in an α-helix covers a distance of 0.54 nm, the LEA-linker, with its 46 residues, covers a distance of approximately 7 nm. Assuming fully extended peptides, adjacent residues are 0.35 nm apart. Therefore, the PAPA-linker with 33 residues is maximally 11.5 nm long. A fully extended GS-linker would cover 7 nm, but its flexible nature likely makes it shorter. Avi-tagged nanobodies were then purified and enzymatically biotinylated to be recognized by Atto565-labeled streptavidin. Streptavidin is bulky compared to a small AF647 dye, and prevents the nanobodies from binding to OmpA in clinical strains due to steric hindrance as opposed to nanobodies labeled with AF647. Hence, size enlargement via streptavidin binding impedes the penetration of nanobodies through the O-antigen layer to reach their target. For our binding assays, we included the lab strain MC1061 (devoid of O-antigen), CS#1 featuring a loose and partially permeable O-antigen layer and CS#2 being shielded with a dense and thus challenging O-antigen layer. Lab strain MC1061 lacking the respective OMPs served as negative control in all assays performed. In a first experiment, constructs having Nb01 as binding module extended with the short PEG11- Biotin linker or the longer polypeptide linkers GS, LEA or PAPA were complexed with Atto565- labeled streptavidin and tested. As expected, for the lab strain MC1061, equally strong binding signals were observed for all linker designs, whereas the signal was absent when probing the MC1061ΔompA strain. In case of CS#1, the construct with the short PEG11-Biotin linker gave rise to the maximal binding signal, showing that in this clinical strain steric hindrance does not appear to be an issue. In contrast, the construct with the short PEG11-Biotin linker did not give rise to any binding signal in case of CS#2, indicating a strong shielding effect of the dense LPS. Interestingly, while the construct with a flexible GS linker did not reach OmpA in CS#2, the constructs bearing the more rigid LEA and PAPA linkers were partially able to penetrate and bridge the dense LPS of CS#2. The same experiment was also conducted with constructs wherein the binding module was Nb18, hence targeting OmpF. Quite unexpectedly, we observed only very weak binding for CS#1 as well as CS#2 as compared to the lab strain MC1061, independently of the linker used. Finally, we generated a construct series wherein Nb01 is fused to Nb18 with a flexible glycine linker, followed by either LEA, PAPA or GS linkers and a biotinylated Avi-tag (Fig. 3b). Overall, the performance of this construct in the cellular binding assay with Atto565-labeled streptavidin was basically identical to the construct bearing only Nb01 as binding module, showing that the increased length gained by the additional nanobody in the construct as well as a bi-paratopic binding module expected to increase binding strength by avidity does not improve cellular staining in a discernible manner. In conclusion, these experiments showed that i) OmpA is better suited as a target to stain and capture E. coli than OmpF, ii) in particular relatively stiff linkers are potentially suited to bridge the LPS layer of clinical E. coli strains featuring dense O-antigen decoration as found in CS#2 and iii) the linker lengths used in our nanobody constructs are likely too short to efficiently bridge the LPS of most clinical strains. Example 12: Capture of clinical E. coli strains using engineered heavy chain-only antibodies Based on our systematic assessment of linker designs, we concluded that achieving an even greater separation between the nanobody and the magnetic bead would likely be required for a reliable capture of clinical strains. Further, dense LPS layers such as present in CS#2 pose a challenging obstacle preventing efficient capture. Therefore, alterations to the growth conditions and capture buffers, including the use of Lactose Broth and / or EDTA, are likely to be required for successful immunomagnetic capture of E. coli. Finally, we reasoned that the experimental setup how the binding assay is performed could potentially influence capture efficiency (Fig.3a); this is why we compared bacterial capture with magnetic beads functionalized with nanobody constructs (procedure 1 in Fig.3a) with a procedure wherein the biotinylated nanobody construct is added in solution and allowed to bind the cells, followed by capture using streptavidin-coated magnetic beads (procedure 2 in Fig.3a). To assess the potential benefits of LPS trimming and thinning, we grew the cells in LB or LacB, and captured the cells using Nb01 modified with the small PEG11- linker according to procedure 1 (Fig. 3e). However, LPS thinning using LacB did not result in improved capture efficiencies of the two clinical strains as compared to cells grown in LB. Of note, we observed considerable background binding to MC1061 ΔompA when cells were grown in LacB. This was, however, not observed in other capture experiments, and thus likely is an outlier. Next, we performed the same assay with the Nb01-Nb18-LEA-biotin construct, which is considerably longer than Nb01-PEG11-biotin. Intriguingly, in this case we saw robust capture of CS#1 and CS#2 when the cells were grown in LacB, but not when grown in LB (Fig. 3f). This experiment clearly showed that both LPS trimming and linker extension are required to achieve successful capture. While the use of LacB-cultured strains provided important insights in the context of capture assay development, it is not a practical solution for diagnostic purposes. Therefore, and based on the findings made above indicating that the linkers we employed thus far are likely too short, we decided to engineer even longer constructs. We realized that the nanobody constructs containing the various linkers gave rather low protein yields, impeding the generation of constructs with even longer linkers fused to the nanobodies. This is why we constructed heavy chain-only antibodies (hcAbs), wherein we fused up to four nanobodies as “pearls on a string”, starting with Nb01 as the outermost molecule (Fig.3b). Again, we introduced an Avi-tag for enzymatic biotinylation at the C- terminus of the Fc-part. To avoid growth in LacB, we transitioned to culturing strains with 0.5 mM EDTA supplementation, as EDTA serves as an anticoagulant. In test experiments, we assessed different capture formats (Fig.3a) and found the procedure 2 (adding biotinylated 4Nb01-hcAb in solution to cells at a 10-fold molar excess over the estimated OmpA concentration in the assay, followed by magnetic capture) to be more efficient in capturing clinical strains than procedure 1. Next, we assessed which molar ratio of 4Nb01-hcAb to OmpA is optimal for capture according to procedure 2. We found that a 100-fold molar excess gives rise to complete capture of CS#2, whereas lower ratios resulted in incomplete or failed capture. In subsequent experiments using the optimized capture procedure 2, we sought to evaluate capture of cells cultured in LB in the absence of EDTA (Fig. 3g). Using the 4Nb01-hcAb construct, we attempted to capture CS#2. CS#6, which expresses the OmpA-long version, served as negative control. We also tested a 4Nb39-hcAb construct targeting OmpA-long (expressed on CS#6) and used CS#2 as a negative control, as it expresses the OmpA-short version not recognized by Nb39. Without EDTA supplementation in the growth medium, capturing CS#2 remained challenging, even with the long linker provided by 4Nb01-hcAb. However, the addition of 0.5 mM of EDTA to the capture buffers resulted in capture efficiency of around 25 % of CS#2. Finally, CS#6, which features a slightly less dense O-antigen than CS#2, was captured effectively by 4Nb39-hcAb, even without the supplementation of EDTA in the capture buffer (Fig.3g). Example 13: Discussion Molecular tools for the reliable and rapid detection, enrichment and isolation of live E. coli cells offer broad applications in routine diagnostics, water surveillance as well as clinical research. However, the high sequence diversity within the versatile E. coli species have thus far prevented the development of antibodies or other capture molecules, that would reach satisfactory species coverage. In this study, we took advantage of the small nanobody scaffold to target the conserved and abundant outer membrane protein OmpA on the cellular surface of E. coli. In clinical isolates, access to OmpA is shielded by the dense O-antigen sugars, as we demonstrated in experiments in which we attached the nanobodies to larger moieties such as streptavidin. Hence, binding of classical antibodies to OmpA is at least partially impeded, and explains the small number of biomolecules thus far generated against E. coli OmpA that would reliably work in the context of intact cells. Only few nanobodies targeting surface structures on pathogenic bacteria have been described in the literature. In previous studies, immunizations were performed using heat-inactivated Campylobacter species or by fixed Acinetobacter baumannii cells . While in case of Campylobacter, the identified nanobodies were shown to recognize the major outer membrane protein (MOMP), the target structure remained unidentified in case of Acinetobacter baumannii. In both studies, cellular staining with the nanobodies required cell fixation, indicating that the targeted epitopes are not accessible in the context of the living cell. In case of the A. baumannii nanobodies, the authors demonstrated that the peptide corresponding to CDR3 can reach the target epitope in the context of intact cells. Our approach taken here and in a previous work was to first raise nanobodies against a purified OMP, thereby achieving strong enrichment of the nanobody pool against the specific target. In a subsequent step, we employed the flycode technology (Egloff et al., 2019 ibid) to deep-screen the nanobody pool against the respective OMP in intact cells, thereby identifying nanobodies that perform exceptionally well in the cellular context of a broad set of strains of the same species. Our approach comes with the important advantage that we knew the target structure prior to the selection. Hence, using bioinformatics analyses we could rationally plan the selection campaign to reach high species coverage with a low number of nanobody binders. As we demonstrated in this study, our targeted approach allowed us to reach a species coverage of E. coli of above 90 % using only two nanobodies. In addition, our bioinformatics analysis allowed us to identify the expected gaps we would have to fill in order to further improve species coverage. Firstly, we realized that the ompA gene appears to be lacking in around 5 % of the genomes found in the 661k database. The reasons behind this might be manifold, including general errors in the sequencing and assembly of genomes. Nanobody binding experiments on strains apparently lacking OmpA would be needed to confirm this in silico finding. The lack of OmpA in clincal isolates is particularly puzzling, because OmpA has been described to be an immune evasin. Should the proportion of E. coli strains without ompA indeed amount to around 5 % as estimated, the nanobodies raised against OmpF, such as Nb18, can be used to fill the gap. Among the E. coli strains in which we found an ompA gene, our two nanobodies reach a coverage of 91 % in the 661k database. By analyzing the OmpA sequences that escaped detection by Nb01 and Nb39, we surmise that another 3 – 4 nanobodies are likely required to achieve a coverage of close to 100 % of the strains producing OmpA. To the best of our knowledge, no other studies exist wherein species coverage of antibodies or nanobodies raised against bacteria has been systematically analyzed and tested. The identified nanobodies were thoroughly characterized at the biochemical and molecular level to rationalize their binding mechanism in the cellular context. The crystal structures of Nb01 and Nb39 revealed that they recognize all four extracellular loops of the respective OmpA isoform, thereby explaining their specificity for OmpA-short and OmpA-long, respectively. Intriguingly, in both structures the nanobodies extend the OmpA barrel like a “flame on a candle”, and hence prevent strong steric clashes with the LPS sugars. Flow cytometry analysis confirmed nanobody binding in the context of clinical isolates and showed that their apparent binding affinities are in the single to double digit nanomolar range, depending on the strain used. While the nanobodies performed very well for cellular staining, immunomagnetic capture of clinical strains turned out to be more challenging, indicating that the O-antigen layer forms a steric barrier. In a series of systematic experiments, we could show that this barrier permits the passage of nanobodies but obstructs bulkier molecules like streptavidin and even larger magnetic beads from penetrating effectively. To enable immunomagnetic capture of clinical strains, we extended the distance between the nanobodies and the bulky magnetic beads by engineering nanobody constructs with long linkers. In addition, we discovered by accident that E. coli cells cultured in lactose broth produce a much thinner sugar layer and are much easier to capture. The molecular reasons underlying this impactful phenomenon are worth being addressed in future studies. Finally, addition of the metal chelating agent EDTA to growth medium and capture buffers consistently increased capture efficiency. While immunomagnetic capture of clinical E. coli strains directly from patient samples or from early blood cultures that did not yet turn positive remains to be demonstrated, we provided here an experimental rationale to tackle this challenge. Our nanobodies hold promise to be used for diagnostic applications. The enrichment of live bacteria allows for seamless integration with traditional phenotypic antimicrobial susceptibility testing (AST), expediting diagnostics by eliminating or shortening blood culture. In molecular diagnostics, the identified nanobodies can be used to enrich bacteria directly from patient samples, thereby improving the ratio of bacterial DNA over human DNA and rendering whole genome sequencing more efficient. Further, our nanobodies are expected to perform well for immunostaining in the context of diagnostic digital microscopy. Beyond clinical diagnostics, the described E. coli nanobodies are suited for culture-independent water and food surveillance. Last but not least, we are confident that our nanobodies will find broad application in clinical research as a highly sensitive tool to live-stain clinical E. coli isolates in the context of infected cells, tissues or organs-on-a-chip. Example 14: Materials and Methods Strains used in this study In this study we used E. coli K-12 strain MC1061, for which we generated unmarked gene deletions as outlined below. E. coli K-12 strain MG1655 wt and wbbL+ were kindly provided by Prof. Dr. Sebastian Hiller. Clinical E. coli strains used in this study are whole genome sequenced clinical isolates collected at the Institute of Medical Microbiology, University of Zurich. Generation of gene deletions To generate gene deletions in E. coli MC1061, a two-step recombination technique was used as described (Murphy et al., Journal of Bacteriology, Vol.180, no.8, 1998) (Murphy et al., BMC Mol Biol, Vol.4, no.4, 2003), (Jensen et al., Sci Rep, Vol.5, 2015). Firstly, the target gene was replaced by homologous recombination with a FRT-flanked kanamycin cassette using the red recombinase from the pKM208 plasmid (Murphy et al., 1998, ibid) (Murphy et al.. 2003, ibid). Secondly, the kanamycin cassette was removed by expressing a flippase from the pSIJ8 plasmid (Jensen et al., Sci Rep, 2015, ibid). To allow for homologous recombination, the kanamycin resistance gene in between two FRT sites was flanked with homologous DNA stretches upstream and downstream of the ompA or ompF. To generate this kanamycin cassette, we generated three PCR products which we ligated together and integrated into a pINIT vector containing an Chloramphenical resistance (Addgene: #46858) using FX cloning (Cull et al., methods Enzymol., Vol.326, 2000) resulting in pINIT_KO_ompA and pINIT_KO_ompF. The three PCR products composed of the 5’ region upstream (primer pairs of ompA_EC_FX_5_FW / RV and ompF_EC_FX_5_FW / RV), the Kanamycin cassette (primer pair Kan_cassette_FW / RV), and the 3’ region downstream (primer pairs of ompA_EC_FX_3_FW / RV and ompF_EC_FX_3_FW / RV), of the region of interest. Finally, PCRs of the generated pINIT_KO_ompA and pINIT_KO_ompF were carried out, resulting in double stranded PCR products encoding for upstream homologous region-FRT-KanR-FRT-downstream homologous region using primer pairs ompA_EC_DKO_5_FW and ompA_EC_DKO_3_FW, and ompF_EC_DKO_5_FW and ompF_EC_DKO_3_FW. E. coli MC1061 cells were transformed with pKM208 (carrying an ampicillin resistance marker) and grown to OD600nm = 0.1 at 30 °C and 160 rpm, at which point recombinase expression was induced with 1 mM IPTG. After 4 h expression, cells were made electrocompetent by washing three times with ice-cold water. Subsequently, electroporation was performed using 3 ug of the gene deletion PCR product, followed by recovery in LB and selection on LB / Kan50 plates. Kanamycin resistant clones were selected, and integration confirmed by colony PCR. To remove the heat-sensitive pKM208 plasmid, the selected colonies were cultured at 42 °C. In the second step, Kanamycin-resistant but Ampicillin-sensitive clones were made electrocompetent and transformed with pSIJ8 (carrying an ampicillin resistance marker). Again, cells were grown to OD600nm = 0.1 at 30 °C and 160 rpm, and flippase expression was induced by adding 10 mM Rhamnose. After 6h of expression at 30 °C, the cells were plated on LB Agar plates. Clones were picked and grown at 42 °C and 160 rpm for 4 h to lose the heat- sensitive pSIJ8 plasmid and plated again on LB Agar plates. Gene deletions of Kanamycin- and Ampicillin-sensitive clones were confirmed by Sanger sequencing. Protein expression and purification and biotinylation Genes encoding ompA-short and ompF were amplified from E. coli K-12 MC1061, while OmpA- long was amplified from clinical isolate #11 (CS#11) of IMM using conventional primer sequences. One primer pair ompA_FW / ompA_RV was used for amplifying full length ompA-short and ompA- long, another primer pair ompA_FW / ompA_TMD_RV for transmembrane domain of ompA-short and ompA-long, and a third primer pair ompF_FW / ompF_RV for full length ompF. For OmpF, a point mutation (S75C) allowing for site-specific biotinylation was introduced (OmpF_S75C_FW &RV primers). Full length OmpA-short and OmpA-long were expressed in p7XC3H (Addgene: #47065, containing a C-terminal 3C-cleavage site and deca-His-tag) for immunization or p7XCA3H (like p7XC3H but with an Avi-tag before the 3C cleavage site) for phage display selection. Transmembrane domain of OmpA-short and OmpA-long for X-ray crystallography and full-length OmpF_S75C for immunization and selection were expressed in p7X (tagless). All constructs were expressed as inclusion bodies in E. coli C43 cells. Cells were grown in TB / Kan50 to OD600nm of 0.9 - 1.5 at 37 °C and 90 rpm in baffled flasks. Expression was induced with 1 mM IPTG, followed by an additional 3 h growth before harvesting by centrifugation (6000 × g, 10 min, 4 °C). Pellets were resuspended in 20 mM Tris, pH 8 and cells were disrupted by passing 4 times through a microfluidizer at 30 kpsi. Unbroken cells were removed by low centrifugation (2000 × g, 15 min, 4 °C). Inclusion bodies were harvested (8000 × g, 10 min, 4 °C), washed once with 20 mM Tris, pH 8, 1 % Triton X-100 and twice with 20 mM Tris, pH 8. Inclusion bodies were solubilized in 20 mM Tris, pH 8 and 6 M guanidine hydrochloride for two hours at room temperature. For full-length and transmembrane domain OmpA (short and long isoforms), aggregates were spun down (8000 × g, 30 min, 4 °C), and the supernatant was added dropwise to 20 mM Tris, pH 8, 5 % C8POE to finally reach a 1:12 dilution, and incubated for 2 h at room temperature for folding. For full-length OmpA, refolded proteins were subjected to Ni-NTA column, washed with 15 column volumes (CV) 20 mM Tris, pH 8, 5 % C8E4, 50 mM imidazole, and eluted with 4 CV 20 mM Tris, pH 8, 5 % C8E4, 250 mM imidazole.3C protease was added and His-tag was cleaved over night at room temperature while dialyzing against 20 mM Tris, pH 8, 0.5 % C8E. Cleaved protein was further purified via reverse IMAC, concentrated, and applied to size- exclusion chromatography (SEC) using a Superdex S200 increase 10 / 300 GL column in TBS, pH 7.4, 0.5 % C8E4. In case of the tag-less transmembrane constructs of OmpA expressed fromp7X, refolded proteins were concentrated, and supernatants were directly run on SEC. The Avi-tagged versions of OmpA-short and OmpA-long were biotinylated in vitro using purified BirA protein (Cull et al., Methods Enzymol., 2000, ibid). The biotinylation reaction was carried out after reverse IMAC in TBS pH7.5, 0.5 % C8E4, 5 mM ATP, 10 mM MgOAc and two-fold molar excess of Biotin overnight at 4°C, followed by SEC. For OmpF_S75C, solubilized inclusion bodies were applied to a PD10 column and eluted with 50 mM Tris, pH 8, 6 M Urea. The protein was diluted to 15 mM Tris, pH 8, 6 M Urea and subjected to anion exchange chromatography using a Resource™ Q, 1 mL column and eluted via a gradient from 15 mM bis-Tris, pH 7, 6 M Urea to 1 M NaCl, 15 mM bis-Tris, pH 7, 6 M Urea. Fractions corresponding to (unfolded) OmpF_S75C were combined and added dropwise to 50 mM Tris, pH 8, 1 mM DTT, 0.1 M EDTA, 0.2 % β-DDM (1:20 dilution) and incubated at 37 °C overnight. The protein was then concentrated and applied for SEC using a Superdex S200 increase 10 / 300 GL column in TBS, pH 7.5, 0.05 % β-DDM. For OmpF_S75C, following refolding, the protein was concentrated, and buffer was exchanged to TBS, pH 7.5, 0.1 % β-DDM using PD10 columns. A five-fold molar excess of biotin-maleimide (Sigma, B1267) was added, and the reaction proceeded for 1 h at room temperature before concentration and application of SEC using a Superdex S200 increase 10 / 300 GL with TBS, pH 7.5, 0.05 % β-DDM. For the expression of OmpA in the outer membrane for cellular binding assays or flow cytometry experiments, E. coli MC1061 ΔompA were transformed with plasmids pBXNPH3 containing either ompA-short or ompA-long. For expression, cells were grown in TB / Amp120 at 37 °C and 90 rpm in baffled flasks to an OD600nm of about 0.5. The temperature was lowered to 20 °C and cells were grown for an additional 2 h. Expression was induced with 0.05 % L-arabinose at an OD600nm of 0.8-1.2. Expression was carried out overnight. Alpaca immunizations Alpacas were immunized by a total of four subcutaneous injections of about 100 ug purified OmpA- short, OmpA-long or OmpF at two-week intervals. Small blood samples were taken to follow the immune response by ELISA. Two weeks after the final injection, larger blood samples were drawn for lymphocyte RNA extraction. Specifically, an Alpaca named Waikuri received injections of both OmpA-short and OmpF, while Thurbo was immunized with OmpA-long. Nanobody Selections Lymphocyte RNA extracted from immunized alpacas served as template for reverse transcription and subsequent amplification of the VHH / nanobody region (Pardon et al., Nat Protoc, 2014 ibid)to generate a phagemid library. Two rounds of phage display were performed against purified and biotinylated OmpA-short, OmpA-long and OmpF in TBS, pH 7.5 and 0.03 % β-DDM exactly as described elsewhere (Zimmermann et al., Nat Protoc, 2020 ibid). Phages were produced in E. coli SS320 ΔompA cells, which were generated by mating MC1061 ΔompA with XL1-blue. The second phage display selection resulted in a 764-fold, 7163-fold, and 1925-fold enrichment against OmpA- short, OmpA-long, or OmpF, respectively, as determined by qPCR compared to a control protein AcrB. From the enriched pools, 2000-2500 colony forming units (CFU) were subcloned into pNLx and nested with an approximately 30-fold excess of flycodes for NestLink analysis (Egloff et al., Nat Methods, 2019 ibid). Deep sequencing and flycode assignment Deep sequencing of the nested libraries was conducted following the previously described method (Egloff et al., Nat Methods, 2019 ibid). In brief, flycoded nanobodies in the plasmid pNLx were excised and ligated with compatible double-stranded Illumina adapter oligonucleotides. The sample was then subjected to sequencing on an Illumina MiSeq Sequencer using a 600-cycle v3 MiSeq Reagent Kit, resulting in 2x300 base pair (bp) paired-end reads. Bioinformatic analysis of the deep- sequenced libraries using the previously published filtering steps (Egloff et al., Nat Methods, 2019 ibid)revealed the following unique nanbodies and flycodes: for OmpA-short, 1’040 unique nanobodies were nested with 29’746 unambiguous flycodes; for OmpF, 824 unique nanobodies were nested with 15’320 unambiguous flycodes; for OmpA-long, 1’739 unique nanobodies were nested with 35’718 unambiguous flycodes. Based on this analysis, a database for MS / MS ion search was generated, containing information about the association of each unique nanobody with its corresponding unambiguously assignable flycodes. Expression and selection of nested libraries For NestLink selections, nested pools were produced in E. coli MC1061 carrying genomic deletions of ompA (for nanobodies raised against OmpA-short and OmpA-long) or ompF (for nanobodies raised against OmpF) (Egloff et al., Nat Methods, 2019 ibid) . Upon purification of the flycoded nanobodies via Ni-NTA chromatography, they were separated on a SRT SEC-300 column (Sepax) using PBS, pH 7.4 and the peak corresponding to monomeric nanobodies was pooled, diluted to 0.02 mg / ml in PBS pH7.4 supplemented with 0.5% BSA, and subsequently added to various bacterial strains. For OmpA-short and OmpF, the flycoded nanobody pools were added to E. coli K-12 strain MC1061, the respective isogenic knockout strain along with a set of clinical strains (CS#1, CS#2 , CS#3 , and CS#4 all expressing the OmpA-short isoform, as well as CS#5, CS#6 , CS#8 and CS#10 expressing OmpA-long). In the case of OmpA-long, lab strains MC1061 ΔompA- short::ompA-long and the isogenic control MC1061 ΔompA-short::ompA-short (both cases involved complementation by expressing the respective OmpA encoded on plasmid pBXNPH3) were used, along with clinical CS#8 and CS#11, both expressing OmpA-long. 50 ml cells with an OD600nm of 2 were harvested and washed once with 25 mL PBS pH 7.4, 0.5 % BSA, and incubated for 20 min at room temperature before pelleting again. 25 ml diluted nanobody pools (0.65 mg, final concentration of 26 µg / mL; determined by measuring A280 and assuming that 1 mg / ml flycoded nanobody has an A280 = 5.0) were added to each cell pellet and incubated for 20 min at room temperature. Each cell pellet was washed three times with 25 ml PBS pH 7.4. Flycode extraction and purification To isolate flycodes, we followed the previously established procedure (Egloff et al., Nat Methods, 2019 ibid). Briefly, the cell pellet was solubilized in 25 ml 4.8 M guanidinium chloride. For quantification in MS / MS a purified nanobody fused to 28 flycodes of known sequence was spiked in and insoluble components were separated by centrifugation. Ni-NTA resin was incubated with His-tagged flycodes under denaturing conditions for 2 h at room temperature, and then transferred to a Mini Bio-Spin chromatography column (BioRad) and washed (3x 500 µL TH-Im buffer (20 mM TEAB, pH 8.0, 150 mM NaCl, 2.5 mM CaCl2, 30 mM Imidazole, pH 8.0), 2x 500 µL TH buffer (20 mM TEAB, pH 8.0, 150 mM NaCl, 2.5 mM CaCl2)). The resin was incubated with 100 µL TH buffer containing 2.4 U thrombin and incubated overnight at room temperature. The resin was washed (5x 500 µL TH-Im buffer) and the flycodes eluted with 250 µL TRY-Im buffer (20 mM TEAB, pH 8.0, 50 mM NaCl, 2.5 mM CaCl2, 250 mM Imidazole, pH 8.0). The eluted flycodes underwent overnight digestion at 37 °C with 1 µg trypsin (2 µL of 0.5 ng / µL stock solution, Promega). The reaction was halted the following day by adding 20 µl of 5 % (v / v) TFA. Subsequently, the sample was diluted with 250 µL 3 % (v / v) ACN and 0.1 % (v / v) TFA before being purified according to the StageTip protocol (Rappsilber et al., Nat Protoc, vol.2, No.8, 2007). LC-MS / MS analysis For StageTip reverse-phase matrix preparation (using 3M™ C18 Extraction Disks), the following steps were performed: Activation was achieved by applying 150 µL of 100% methanol, followed by pre-elution with 150 µL of a solution composed of 60% acetonitrile (ACN) and 0.1% trifluoroacetic acid (TFA). Subsequently, the resin was prepared for peptide binding using 150 µL of a solution consisting of 3% ACN and 0.1% TFA. Trypsin-digested samples (as described above) were diluted with an equal amount of 250 µL 3% ACN and 0.1% TFA and loaded onto the StageTips. After loading, the StageTips were washed three times with 150 µL of 3% ACN and 0.1% TFA. Elution was carried out with 150 µL of 60% ACN and 0.1% TFA, followed by lyophilization. To reconstitute the flycodes, 15 µL of 3% ACN and 0.1% formic acid (FA) supplemented with indexed retention time (iRT) standard peptides (2xiRT kit, Biognosys) were added, and the samples were sonicated using a bath sonicator. Finally, 4 µL of the reconstituted samples were subjected to LC- MS / MS analysis. For OmpA-short and OmpF, sample analysis involved an ACQUITY M-class UPLC system (Waters AG) coupled with a Q-Exactive HF mass spectrometer (ThermoFisher). LC system equilibration used 99% solvent A (0.1% formic acid in water) and 1% solvent B (0.1% formic acid in ACN). Peptide trapping occurred on a Symmetry C18 trap column (5 µm, 180 µm × 20 mm, Waters AG) at a flow rate of 15 µL / min for 30 s. Subsequently, peptide separation utilized an HSS T3 C18 reverse-phase column (1.8 µm, 75 µm × 250 mm, Waters AG) with the following gradient: 8–20% solvent B in 60 min, 20–40% solvent B in 10 min, and 40-95% in 5 min. The flow rate remained constant at 0.3 µL / min, and the temperature was maintained at 50 °C. Mass spectra were recorded in a data-dependent acquisition mode on a Q-Exactive HF mass spectrometer. MS1 spectra were acquired using a mass range of 350–1,500 m / z at a resolution of 120’000 (at 200 m / z) with an automatic gain control (AGC) target of 3 × 106 and a maximum injection time of 50 ms. Peptide precursor with charge state between 2 and 7 were selected for fragmentation using quadropole isolation (1.6 m / z window), a resolution of 30’000 at 200 m / z, an AGC target value of 1 × 105 and a maximum injection time of 50 ms, with a normalized collision energy of 28 %. Dynamic exclusion was activated and set to 15 s with a mass tolerance of 10 p.p.m. For flycode analysis of nanobodies selected against OmpA-long, isolated and purified flycodes were analyzed using an ACQUITY M-class UPLC system (Waters AG) coupled to an Orbitrap Fusion Lumos Tribrid Mass Spectrometer (ThermoFisher). Trapping and elution of peptides were conducted as previously described. Acquisition of MS1 spectra was recorded in a data dependent mode using Orbitrap in the scan range of 300-1500 m / z, with an AGC target of 4 × 105, a resolution of 120,000 at 200 m / z, and a maximum injection time of 50 ms. For peptide precursors with charge states between 2 and 7, MS2 spectra were recorded with an IonTraputilizing a 1.6 m / z isolation window, an AGC target value of 8000, and a maximum injection time of 80 ms. High-energy collisional dissociation fragmentation (HCD) was set to 30 % collision energy. To prevent redundancy, dynamic exclusion was activated and configured with a 25-second interval, a mass tolerance of 10 p.p.m.. A minimum signal intensity of 5000, and a maximum cycle time of 3 s was set. LC-MS / MS data processing LC-MS / MS data was processed using Progenesis QI by Nonlinear Dynamics. Experiments with respective pools were aligned, and peaks with ion charges ranging from +2 to +4 were automatically identified. MS / MS fragment spectra, with a feature rank threshold of less than 5 and an ion fragment count limit of 1,000, were exported after applying deisotoping and charge deconvolution. Mascot 2.5 (Matrix Science) was utilized to match MS / MS features to flycodes present in the respective databases. For all experiments, the database 'p1875_db8' was employed, as it contained the spiked standard, in addition to a Swiss-Prot database ('fgcz_swissprot_S') containing common contaminants and decoys. In the case of the OmpA-short and OmpF library, database 'p3127_db1' was used, and for OmpA-long, 'p3127_db6' was used. Scaffold (Proteome Software Inc.) was applied for protein identification from the Mascot search as described. The spectrum report from Scaffold was integrated into Progenesis QI. Within Progenesis QI, protein abundance was normalized using the spiked control nanobody NB-Control. Protein abundance was calculated by summing the MS1 intensities of each flycode associated with the respective nanobody. Binder candidates were selected based on high MS1 intensities, ratios of MS1 intensities between wild- type and knockout strains, and the number of flycodes associated with them. Nanobody expression and purification Selected nanobody genes, codon-optimized for E. coli, were synthesized by Twist and cloned into pSBinit for purification (Zimmermann et al., Elife, 2018 ibid),(Zimmermann et al., Nat Protoc, 2020 ibid). The nanobodies were expressed in E. coli MC1061 knockout strains (ΔompA or ΔompAΔompF) depending on the target protein. Following periplasmic extraction, supernatants were subjected to Ni-NTA affinity chromatography (Qiagen) and SEC (SRT SEC-100 or SRT SEC- 300, Sepax). In order to attach fluorophores (Alexa Fluor 647, AF647, Catalogue #A20347, ThermoFisher or Dy-490 Maleimide, Catalogue #490-03 Dyomics) or linkers (PEG11-Biotin, Catalogue # 21911, ThermoFisher) using maleimide chemistry, a cysteine residue at the C- terminus of the nanobody was introduced by Quickchange mutagenesis using primer pair pSBinit_Cys_FW and RV. To avoid oxidation of the free thiol group, purification of these cysteine- containing nanobodies was performed in the presence of 2 mM DTT. After SEC, DTT was removed using a PD MidiTrap G-25 (28918008, Cytiva) desalting column equilibrated with degassed PBS, pH 7.0. Maleimide-functionalized AF647 or PEG11-Biotin, was added at 3.6-fold molar excess and the reaction was carried out in maximally 1 mL for 1 h at 4°C. Excess label was removed using another PD MidiTrap G-25 column, equilibrated with PBS, pH 7.4. Labelling efficiency was assessed by absorbance, MS, and SDS-PAGE. Nanobody constructs with peptide linkers were modified by replacing the myc-tag with either LEA (SEQ ID NO 069), PAPA (SEQ ID NO 070), or (tetraglycine-serine)x4 (SEQ ID NO 071) linker, followed by an Avi-tag for site-specific in vitro biotinylation with BirA and a 3C cleavage site. Genes for these linkers were obtained from GeneUniversal. For constructs containing two binders, the nanobodies were separated by a flexible (tetraglycine-serine)x4 (SEQ ID NO 071) linker. Nanobody constructs fused to a hIgG1 Fc part, codon-optimized for homo sapiens, were subcloned into pcDNA3.4-derived expression vectors under CMV promoter control, preceded by a mammalian Kozak sequence and an immunoglobulin secretion signal, synthesized by GeneUniversal. Transfection-grade plasmid DNA was prepared using a NucleoBond Xtra midiprep kit (Macherey- Nagel) and transfected into suspension Expi293 cells (ThermoFisher) using Expifectamine 293 transfection kit (ThermoFisher). Expression was carried out in 40-50 mL of 2-3 million cells / mL for 5 days at 37 °C in a humidified shaker maintained by 8 % CO2. Supernatants were collected, cleared by two centrifugation steps (500 × g for 5 min and 5000 × g for 20 min), and then subjected to Protein A Agarose (Abcam) capture using 2-4 mL bead slurry for 2-4 h at room temperature. The resin was transferred to a gravity flow column and washed with 50 mL PBS. Protein elution was performed with 0.1 M glycine, pH 3, and fractions were collected, rebuffered using 1 / 10 of the elution volume of 1 M Tris, pH 8.5, and concentrated with centrifugal spin filters with a MW cutoff of 100 kDa (Merck) and subjected for SEC using a Superose 6 column (Cytiva) after a centrifugal high spin of > 20000 × g for 10 min to remove potential aggregates. Purified HcAbs were finally run on Superose 6 column inPBS, and pure fractions, verified by SDS-PAGE, were pooled, concentrated, and stored at -80 °C after snap-freezing in liquid nitrogen until application. Nanobodies for X-ray crystallography were cloned into pBXNPHM3 (Addgene: 110099) and transformed into MC1061 ΔompA. Nb01 and Nb39 were expressed at 20 °C overnight, harvested, and purified via Ni-NTA affinity chromatography (Qiagen) with subsequent 3C protease cleavage and dialysis. Reverse Ni-NTA affinity chromatography was performed to remove 3C protease, His- tagged MBP fusion protein and uncleaved protein, followed by concentration and SEC purification (SRT10C 300, Sepax) using TBS. Purified nanobodies were concentrated to > 15 mg / mL. X-ray Crystallography The β-barrel domains of OmpA-short and OmpA-long were expressed and purified from plasmid p7X for tagless expression as described above. OMPs were concentrated to 20 mg / mL and mixed in a 1:1.1 molar ratio of OmpA to the respective targeting nanobody to a final concentration of 10 mg / mL. Protein mixtures were crystallized by the sitting drop vapor diffusion method at 20 °C. In case of OmpA-short in complex with Nb01, 100 nL of protein solution containing OmpA (10 mg / mL) and Nb01 (7.8 mg / mL) was mixed with 100 nL of reservoir solution (0.1 M sodium acetate pH 5.5, 0.2 M calcium acetate, 25 % [w / v] PEG MME 2K). The crystals were cryoprotected with cryoprotection solution (0.1 M sodium acetate pH 5.5, 0.2 M calcium acetate, 25 % [w / v] PEG MME 2K, 25 % [w / v] glycerol) and then flash-frozen in liquid nitrogen. Diffraction data were measured at the beamline X06DA (PXIII) of the Swiss Light Source at a temperature of 100 K (Paul Scherrer Institute, Villigen, Switzerland) and processed using autoPROC (Vonrhein et al., Acta Crystallogr D Biol Crystallogr, vol.67, No.4, 2011)in the space group P21. Phases were obtained by molecular replacement using the Phaser module of the Phenix package using the transmembrane domain of the OmpA PDB-ID: 1BXW and the Nb01 model generated by the AlphaFold2-based ColabFold as initial search model (Jumper et al., Nature, Vol.596, No.7873, pp.583-589, 2021), (Mirdita et al., Nat Methods, vol. 19, No.6, 2022). Two copies of the OmpA-Nb01 complex were present in the asymmetric unit. The model building was done manually in Coot (Emsley et al., Acta Crystallogr D Biol Crystallogr, vol.66, no.4, 2010). The model was refined using phenix.refine module (Afonine et al., Acta Crystallogr D Biol Crystallogr, vol.68, no.4, 2012) The crystals belonged to the P21 space group and contained two copies of the OmpA-Nb01 complex in the asymmetric unit. For OmpA-long in complex with Nb39, 100 nL protein complex mixture of OmpA-long (10 mg / mL) and Nb39 (7.5 mg / mL) was mixed with equal amount of reservoir solution (insert 0.1 M NaCl, 0.1 M Sodium citrate, pH 5.5, 12 % PEG 4K, 0.1 M LiSO4here). Crystals were cryoprotected in (0.1 M Sodium citrate, pH 5.5, 12 % PEG 4K, 0.1 M LiSO4, 30 % glycerolcryoprotectant) and snap-frozen in liquid nitrogen. Diffraction data were measured at the beamline X06SA (PXI) of the Swiss Light Source. The data were processed in the space group I222 using the Automatic Data Processing pipeline at the beamline (Wojdyla et al., Journal of Synchrotron Rad, Jan 2018). The structure was solved by molecular replacement using Phaser module of Phenix using the OmpA-short model processed by Scupltor (Bunkoczi et al., Acta Crystallogr D Biol Crystallogr, vol.67, no.4, 2011) and the Nb39 model generated by ColabFold. The model building was done in Coot and in ISOLDE (Croll et al., Acta Crystallogr D Struct Biol, vol.74, no.6, 2018) and the model refinement was done using phenix.refine. Cellular binding assays Unless otherwise specified, the indicated strains without an expression plasmid were cultured overnight in LB medium at 37 °C and 160 rpm. Strains containing plasmids for the expression of OmpA-long or OmpA-short (E. coli MC1061 ΔompA with either pBXNPH3-ompA-short or pBXNPH3-ompA-long) were grown and induced with 0.05 % L-arabinose according to the above- mentioned protein expression protocol. After expression and overnight growth, the OD600nm was determined and adjusted to 1.0 in 1 mL PBS at pH 7.4.1 mL cells of OD600nm 1 were pelleted and washed in 500 µL PBS at pH 7.4 with 0.5 % BSA. The cell pellets were then resuspended in 100 µL PBS with 0.5 % BSA containing 1 µM AF647-labeled nanobody. The mixture was incubated for 20 min at room temperature. Unbound nanobodies were removed by washing the cells twice with 500 µL PBS. Finally, the cells were resuspended in 100 µL PBS, and fluorescence was measured (excitation: 650 nm; emission: 675 nm) and normalized to the OD600nm using a plate reader (Cytation, BioTek). For the detection of biotinylated or His-tagged nanobodies using Atto565-labeled streptavidin, washed cells were incubated with 100 µl PBS with 0.5 % BSA containing 2.5 µM nanobodies for 20 min at room temperature. Unbound binders were removed by two washes with 500 µL PBS at pH 7.4 containing 0.5 % BSA. Cell pellets labelled with biotinylated nanobodies were then resuspended in 100 µL 1 µM streptavidin-Atto565 in PBS at pH 7.4 with 0.5 % BSA and incubated for 20 min. To detect nanobodies via the His-tag, previously coupled NTA-Biotin-streptavidin- Atto565 dye was used. Two washes with 500 µL PBS removed unbound Streptavidin. In the final step, cells were resuspended in 100 µL of PBS, and fluorescence was measured (excitation: 563 nm; emission: 592 nm) and normalized to the OD600nm using a plate reader (Cytation, BioTek). Analysis of Nb01 and Nb39 specificity by high-throughput flow cytometry E. coli (28 representative strains from the Swiss database and 12 strains from official culture collections) were cultured in Lactose Broth (LacB) overnight at 37 °C. Bacterial strains other than E. coli were cultured as recommended by the respective culture collection (DSMZ or ATCC). To determine cell densities, overnight cultures were stained with 1x SYBR Green I (S9430-.5ML, Sigma-Aldrich), 20 µg / mL propidium iodide (P1304MP, Molecular Probes) and quantified by flow cytometry using a CytoFLEX (Beckman Coulter), equipped with a 488 nm laser and filter sets of 525 / 40 (green channel) and 690 / 50 (red channel) at a flow rate of 100 µL / min. Based on this initial quantification, bacteria were diluted in PBS supplemented with 1 mM EDTA to approximately 100,000 cells per mL and stained for 1 h at room temperature with either 1 nM Dy490-labeled Nb01 or 1 nM Dy490-labeled Nb39 and 0.5 µg / mL propidium iodide. To obtain total cell counts independently of nanobody staining, the same bacterial dilutions were separately stained with 1x SYBR Green I, 20 µg / mL propidium iodide for 1 h at room temperature. To test for Nb01 and Nb39 specificity, the stained samples were analyzed by flow cytometry as described above. For evaluation of Nb01 and Nb39 specificity, intact bacteria signal with a green intensity brighter than 2,000 AU are quantified within a pre-defined gate and compared to the intact bacteria concentration of the SYBR Green reference. Strains with nanobody stained event counts higher than 50 % of the SYBR Green reference, were considered as detected. Strains with nanobody stained event counts less than 1 % of the SYBR Green reference were considered not detected. Nanobody characterization using flow cytometry – sample preparation To characterize nanobodies in detail by flow cytometry, the indicated E. coli strains were cultured overnight in LB media or induced for protein expression as per the previously described protocols when required. All reagents were sterile-filtered through a 0.22 µm pore filter. To begin, 1 mL of culture was taken and washed in 0.85 % NaCl or PBS, and the cell density was adjusted to 3 × 107 cells / mL assuming a concentration of 8 x 108cells / ml at OD600nm of 1. In the initial step, 50 µL of cells were stained with the indicated amounts of Alexa Fluor 647-labeled nanobody or nanobody construct. If not mentioned otherwise, cells were fixed after staining. To this end, cells were washed twice in 0.85 % NaCl and then resuspended in 2 % PFA. The samples were incubated at 4 °C for 45 min in the dark. After incubation, PFA was removed through another washing step. Finally, cells were resuspended in a solution containing 3.34 µM Syto9 and 20 µM PI (from the LIVE / DEAD™ BacLight™ Bacterial Viability and Counting Kit, ThermoFisher) for viability staining. For samples without fixation, cells were washed twice with PBS, and approximately 150,000 cells were added to each well.50 µL of labeled nanobody was added to the respectively indicated final concentration. Finally, 50 µL of Syto9 and PI were added to a final concentration of 3.34 µM and 20 µM, respectively. For each measurement, single stains and unstained controls were recorded for each strain used. To obtain a PI-positive control, cells were incubated in 70 % ethanol for 30 min and washed twice in 0.85 % NaCl or PBS before adding PI. Nanobody characterization using flow cytometry – measurements Data acquisition was performed using either a CytoFlex S (Beckman Coulter) or an Aurora (Cytek) flow cytometer, as indicated for the respective experiment. The Aurora flow cytometer seems to be more sensitive, collecting more signal, however, measurements on both cytometers results in a similar EC50 value. The CytoFlex S is equipped with 405 nm, 488 nm, 561 nm, and 640 nm lasers, while the Aurora additionally features a 355 nm laser. For the CytoFlex S, emitted fluorescence light was collected with 525 / 40 nm (blue channel), 610 / 20 nm (yellow-green channel), and 676 / 19 nm (red channel) bandpass filters. The Aurora utilized virtual filters set at 508 / 20 nm (blue), 615 / 20 nm (yellow-green), and 679 / 18 nm (red). To minimize background noise, a threshold of 650 was applied for FSC, and 750 for SSC during measurements on the CytoFlex. For the CytoFlex S, gains were set as follows: Syto9 = 100, PI = 600, AF647 = 750. For the Aurora, thresholds were adjusted within the range of 4000 - 6000 for FSC and 16000 - 21000 for SSC, without affecting the bacterial population. The gain settings for the Aurora were: FSC = 950, SSC = 735, B1 = 1200, YG3 = 920, R2 = 1500. In both cases, data was acquired at a flow rate of 100 µl / min. Data analysis was performed using R (v.4.0.2) and FlowJo (v.10). In R, we utilized the flowAI (v.1.18.5), flowCore (v. 2.0.1), flowWorkspace (v. 4.0.6), ncdfFlow (v. 2.34.0), and flowStats (v. 4.0.0) packages for evaluation. All datasets underwent quality control using the flow_auto_qc() function from the flowAI package with default settings. Fluorescence intensity was shown using the determined earth mover’s distance (EMD) instead of median fluorescence intensity (MFI). EMD quantifies biologically meaningful differences between a control sample, in our case a non-related control binder (SB-nr: a non-randomized synthetic nanobody (sybody) from the concave library (Zimmermann et al., Elife, 2018 ibid) ), and the investigated nanobody. MFI in contrast relies on normal distribution of the fluorescence intensity peaks and loses information about population distribution. To calculate the EMD, we use Sb-nr at a concentration of 33 nM and calculate its dissimilarity to different nanobody concentrations using the R package transport (v.012-2). Determined EMDs were plotted against nanobody concentration. Curves were fitted with GraphPad prism v9.3 and a variable slope model using the Hill equation to determine EC50 values. Standard deviations were determined from three EC50 values of biological replicates. LPS extraction and separation on Tricince-SDS-PAGE LPS was extracted by resuspending the bacterial pellet at OD600 of 10 in 1 mL distilled water.25 µL bacterial suspension was mixed with 25 µL 2× sample buffer (200 mM Tris-HCl (pH 6.8), 40 % glycerol, 2 % SDS, 0.04 % Coomassie Blue, and 2 % β-mercaptoethanol) and incubated at 100 °C for 10 min. DNaseI was added to a final concentration of 100 µg / mL, followed by incubation at 37 °C for 30 min. Subsequently, proteinase K was introduced to a final concentration of 2 mg / mL, and the mixture was incubated at 60 °C for 1 h. Tricine SDS-PAGE was employed for the separation and analysis of the extracted LPS. A 20 % acrylamide separation gel was prepared by mixing acrylamide / bisacrylamide (37.5:1, 40 %), 3 M Tris-HCl (pH 8.5), SDS (20 %), glycerol (50 %), ammonium persulfate (APS, 10 %), and N,N,N',N'- Tetramethylethylenediamine (TEMED). For the 4 % acrylamide stacking gel, acrylamide / bisacrylamide (37.5:1, 40 %), 3 M Tris-HCl (pH 8.5), and SDS (20 %), were combined with APS (10 %) and TEMED. LPS samples were diluted 1:5 with sample buffer and loaded onto the gel. Electrophoresis was carried out at 100 mA per gel for 1 hour in migration buffer (100 mM Tris-HCl (pH 8.3), 100 mM Tricine, and 0.1 % SDS). The gel was subsequently stained using Emerald strain according to manufacturer’s instructions (P20495, ThermoFisher). Capture Assay Unless otherwise specified, E. coli strains were cultured in Luria Broth (LB, 6271000, LLG Labware). Where specified, LB supplemented with 0.5 mM EDTA, or Lactose Broth (LacB, 70142, Merck) were used. The capture assay was performed in two setups. The first setup involved preparing magnetic beads by coupling them with biotinylated nanobody constructs, followed by adding these prepared beads to the diluted or spiked bacteria. The second setup entailed incubating the biotinylated constructs with the bacteria, diluting the solution, and then capturing the mixture with magnetic beads. For procedure 1, cultured strains were diluted to approximately 500 CFU in 1mL PBS unless stated otherwise, and then mixed with the prepared beads. To prepare the beads, 5 µg of magnetic streptavidin beads (Dynabeads™ MyOne™ Streptavidin C1)) were washed three times in PBSTB (PBS, 0.05 % Tween20, 1 % BSA). The beads were then resuspended in 1 mL of PBS and coupled with biotinylated nanobody constructs added at a 1.5-fold molar excess to the maximal binding capacity of the magnetic beads. The reaction proceeded for 30 min at room temperature with end- to-end rotation. Unbound nanobody was removed by using a magnet, and unoccupied biotin binding sites were saturated by adding 1 mL of 5 mM Biotin in PBST for 5 min. Subsequently, the beads were washed three times in PBSTB and used directly for capture. In the case of procedure 2 of the capture assay, bacteria were diluted to an OD600nm of 0.01, and incubated with 65 nM of biotinylated nanobody constructs were added. According to literature, one E. coli cell has about 100’000 OmpA molecules per cell and at OD600nm of 1 a concentration of 8 x 108 cells / mL. Using the Avogadro constant, an E. coli culture of an OD600nm of 0.01 has an OmpA concentration of 1.32 nM. To reach an excess of about 50 fold nanobody over OmpA, 65 nM nanobody is required. The mixture was incubated for 1 h at room temperature while shaking. Cells were then diluted three times at a 1:10 ratio in PBSTB and combined with previously washed magnetic beads in PBSTB. The capture was performed for 20 min at room temperature with end- over-end rotation. In both cases, after capture, the beads were separated using a magnet, and the supernatant was removed. A 1 / 10 fraction of the supernatant was plated on LB agar for CFU counting, and the beads were resuspended in a small volume of PBS for plating and CFU counting as well. Capture efficiencies were determined by dividing the bead-bound fraction (BBF) by the sum of the BBF and unbound fraction (UBF) counted in the supernatant. Sequence coverage and specificity determination To study OmpA diversity in sequenced E. coli isolates, two databases were screened. First, 2,093 genomes derived from clinical isolates from the University Hospital Basel (USB) and the Institute of Medical Microbiology Zürich (IMM) were screened using blastn v2.13 (Altschul et al., J Mol Biol, vol.215, 1990) and the ompA nucleotide sequence of K-12 (NC_000913.3). Gene sequences were translated to protein sequences, aligned using prank v.170427 (Loytynoja, Methods in Mol Biol, vol.2231, 2021) and a tree was calculated using IQ-TREE v2.2.0.3 (Minh et al., Mol Biol Evol, vol. 37, no.5, 202) and midpoint-rooted. Sequences were further analyzed in R and the loop variants identified (defined as unique combinations of the extracellular OmpA loops 1 to 4). Four outliers due to missing data or extremely divergent sequences were excluded. The sequences of the complete region on the nucleotide level ranging from loop 1 to loop 4 of the clinical strains were then used to query the 661k database (Blackwell et al., PLos Biol, 2021 ibid)using cobs v0.2.0 (Bingmann et al., arXiv:19050962, 2019) and a kmer similarity range from 0.8 to 1 to capture more diversity present in OmpA in E. coli strains. Hits were analyzed adapting the R script by Blackwell et al. (PLos Biol, 2021 ibid). The identified E. coli genomes were then used to build a database and queried with blastn using the reference ompA gene sequence as before. Sequence data was analyzed in R to identify loop variants as before. The individual loop sequences were concatenated and clustered using hierarchical clustering based on the Levenshtein distance in R. To identify possible cross-reactions of the nanobodies with other protein sequences similar to the extracellular ompA loops found in E. coli, hmmer profiles (hmmer v3.3.2 (http: / / hmmer.org / )) were generated using the amino acid alignments for each loop and queried against all proteins of E. coli as well as all proteins of other bacterial species (RefSeq release 213, (O’Leary et al., Nucleic Acids Res, vol. 44, 2016)) with a sequence E value threshold of 0.001. Data Availability Mass spectrometry data and NGS-derived sequence databases will be available at ProteomeXchange.. Scripts of the bioinformatic analysis of ompA diversity are available at https: / / gitlab.uzh.ch / appliedmicrobiologyresearch / amr_publications / ompa_paper. Cited references: All scientific publications and patent documents cited in the present specification are incorporated by reference herein. SEQUENCES: In the event of discrepancies between the sequences shown in the present specification and those of the enclosed sequence protocol according to WIPO standard ST.26, the sequences shown herein shall prevail.

Claims

Claims 1. A composition comprising: a. a first binding agent comprising a first single-domain polypeptide capable of specific binding to a first epitope variant of OmpA (SEQ ID NO 054; OmpA-short), and b. a second binding agent comprising a second single-domain polypeptide capable of specific binding to a second epitope variant of OmpA (SEQ ID NO 055; OmpA- long) wherein said first and second binding agent are: i) linked to a detectable label; or ii) attachable, or attached, to a surface, wherein i. the first single domain polypeptide comprises a CDR1, CDR2 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 014 [GTGFTFSKSPMS] and SEQ ID NO 015 [GSGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is selected from SEQ ID NO 017 [RRLGKTTYDY] and SEQ ID NO 018 [RRLGKRTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to one of SEQ ID NO 014, 015, combined with SEQ ID NO 016 and with one of SEQ ID NO 017 and 018, and the single domain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01), and / or ii. the second single domain polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 043 [GSIFNFNPMG], SEQ ID NO 044 [GSIFNFNLMG], SEQ ID NO 045 [GGIFNFNIMG], and SEQ ID NO 046 [GSIFNFNIMG] and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to either SEQ ID NO 043, 044, 045 or 046, and to 047, and the single domain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39).

2. The composition according to claim 1, wherein a. the first single domain polypeptide comprises a sequence selected from: i. SEQ ID NO 001 (Nb01) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSSTYY SDSVRGRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVTVS ;ii. SEQ ID NO 002 (Nb02) SQMQFVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGQGTPVT VS iii. SEQ ID NO 003 (Nb03) SQGQLVESGGGLVPPGGSLRLSCAVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYSDSVRGRFTISRDNAKNTVYLQMNNVKPEDTAVYYCGHRRLGKTTYDYRGKGTRVT VS iv. SEQ ID NO 004 (Nb04) SQRQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNDVQPEDSAVYYCGHRRLGKRTYDYRGKGTPVT VS v. SEQ ID NO 005 (Nb05) SQRQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVT VS vi. SEQ ID NO 006 (Nb06) SQVQLVESGGGLVQPGGSLRLSCVVSGTGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYSDSVRGRFTISRDNAKNTVYLEMNNVKPEDTAVYYCGHRRLGKTTYDYRGQGTRVT VS vii. SEQ ID NO 008 (Nb08) SQMQLVESGGGLVQPGGSLRLSCVVSGSGFTFSKSPMSWARQAPGKEREWVSAIFADSS TYYADSVKGRFTISRDNAKNTVYLQMNSVKPEDTAVYYCGYRRLGKTTYDYRGQGTPVT VS or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 001, 002, 003, 004, 005, 006 or 008, and having the same biological activity as SEQ ID NO 001 (Nb01), and / or b. the second single domain polypeptide comprises a sequence selected from: viii. SEQ ID NO 039 (Nb39) SVKGRFTISRDSAKNTLYLQMTDVKPGDAAVYICHANYRIGRNDLPVWGKGTPVTVS] ix. SEQ ID NO 038 (Nb38)SQRQLVESGGGTVQTGGSLRLSCVPNGSIFNFNLMGWYRQSSGQQRELVATLTRDGSEN YAEFVKGRFTISRDSGKNTMYLQMTDVKPSDTAVYICHANYRIGRNDLPVWGKGTRVTV S x. SEQ ID NO 040 (Nb40) SQRQLVESGGGLVQPGGSLRLSCVPNGSIFNFNIMGWYRQNAGNQRELVATMTRDGSAS YSDSVKGRFTISRDVDKNTIYLQLDSVKPEDTAVYICHANYRIGRNDLPVWGRGTRVTV S xi. SEQ ID NO 042 (Nb42) SQLQLVESGGGLVQPGGSLRLSCVPNGGIFNFNIMGWYRQNAGNQRELVATMTRDGSAS YSDSVKGRFTISRDVDKNTIYLQMDSVEPEDTAVYICHANYRIGRNDLPVWGQGTPVTV S or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 039, 038, 040, or 042, and having the same biological activity as SEQ ID NO 039 (Nb39).

3. A binding agent comprising a polypeptide, said polypeptide comprising a CDR1, CDR2 and CDR3 sequence, wherein CDR1 is selected from SEQ ID NO 014 [GTGFTFSKSPMS] and SEQ ID NO 015 [GSGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is selected from SEQ ID NO 017 [RRLGKTTYDY] and SEQ ID NO 018 [RRLGKRTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to one of SEQ ID NO 014, 015, combined with SEQ ID NO 016 and with one of SEQ ID NO 017 and 018, and the single domain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01).

4. The binding agent according to claim 3, wherein CDR1 is SEQ ID NO 014 [GTGFTFSKSPMS], CDR2 is SEQ ID NO 016 [AIFADSSTY], and CDR3 is SEQ ID NO 017 [RRLGKTTYDY], or wherein the sequence of CDR1, CDR2 and CDR3 show a total of one or two conversions relative to SEQ ID NO 014, 016 and 017, and the single domain polypeptide has the same biological activity as SEQ ID NO 001 (full sequence of Nb01).

5. The binding agent according to claim 3 or 4, wherein the first single-domain polypeptide comprises a sequence selected from: i. SEQ ID NO 001 (Nb01); ii. SEQ ID NO 002 (Nb02); iii. SEQ ID NO 003 (Nb03); iv. SEQ ID NO 004 (Nb04); v. SEQ ID NO 005 (Nb05); vi. SEQ ID NO 006 (Nb06);vii. SEQ ID NO 008 (Nb08); or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO SEQ ID NO 001, 002, 003, 004, 005, 006 or 008, and having the same biological activity as SEQ ID NO 001 (Nb01).

6. A binding agent comprising a polypeptide capable of selectively binding to OmpA of a clinical isolate of E. coli, wherein the polypeptide comprises a CDR1 and CDR3 sequence and wherein CDR1 is selected from SEQ ID NO 043 [GSIFNFNPMG], SEQ ID NO 044 [GSIFNFNLMG], SEQ ID NO 045 [GGIFNFNIMG], and SEQ ID NO 046 [GSIFNFNIMG] and CDR3 is SEQ ID NO 047 [NYRIGRNDLPV] or wherein the sequence of CDR1 and CDR3 show a total of one or two conversions relative to either SEQ ID NO 043, 044, 045 or 046, and to 047, and the single domain polypeptide has the same biological activity as SEQ ID NO 039 (full sequence of Nb39).

7. The binding agent according to claim 6, wherein the polypeptide comprises a sequence selected from: i.SEQ ID NO 039 (Nb39); ii.SEQ ID NO 038 (Nb38); iii.SEQ ID NO 040 (Nb40); iv.SEQ ID NO 042 (Nb42); or a sequence at least 90%, particularly ≥95%, identical to SEQ ID NO 039, 038, 040 or 042, and having the same biological activity as SEQ ID NO 039 (Nb39).

8. The composition according to any one of claims 1 to 4, or the binding agent according to any one of claims 3 to 7, wherein the binding agent is linked to a fluorescent dye.

9. The composition according to any one of the preceding claims 1 to 4, or 8, further comprising a chelator of divalent cations.

10. The composition according to any one of the preceding claims 1 to 4, 8 to 9, further comprising an anticoagulant agent, particularly an anticoagulant agent selected from citrate phosphate dextrose, sodium polyanethole sulfonate (SPS) and heparin.

11. The composition according to any one of the preceding claims 1 to 4, or 8 to 10, comprising additionally a third binding agent comprising a third single-domain polypeptide capable of specific binding to OmpF.

12. The composition according to any one of claims 1 to 4, or the binding agent according to any one of claims 3 to 7, wherein the biological activity is the same if the polypeptide thus qualified binds to its target, OmpA-short or OmpA-long, respectively, with the same specificity as the reference antibody.

13. The composition or the binding agent according to claim 12, wherein the biological activity is assayed via a method comprising the steps: a. labelling the polypeptide directly with AlexaFluor 647 (AF647) via a cysteine introduced at the C-terminus; b. observing binding of the polypeptide to its target, OmpA-short or OmpA-long, respectively, by fluorescence microscopy.

14. A method for analysis of a sample, said method comprising a. contacting the sample with a composition according to any one of claims 1 to 4, or 8 to 12, wherein the first binding agent and the second binding agent are attached to a detectable label, and b. detecting the presence of E. coli in said sample, particularly by determining the distribution of the label within the sample.

15. A method for isolating E. coli bacteria from a sample, said method comprising a. contacting a sample with a composition or binding agent that comprises a polypeptide capable of selectively binding to OmpA as described in any one of the preceding claims, wherein the binding agent comprises an attachment moiety capable of selectively binding to a surface, and b. contacting the sample with a surface to which the attachment moiety can bind; c. removing the sample and isolating E. coli bacteria from the surface.

Images