WO2022081090A1

WO2022081090A1 - System and method for allowing large molecules to enter the periplasm of e. coli

Info

Publication number: WO2022081090A1
Application number: PCT/SG2021/050614
Authority: WO
Inventors: Majid Eshaghi; Swaine Lin CHEN
Original assignee: National University Of Singapore
Priority date: 2020-10-12
Filing date: 2021-10-12
Publication date: 2022-04-21

Abstract

An isolated genetically engineered gram-negative bacteria cell having an inwardly leaky outer membrane, wherein the cell has been modified in at least one way, the modification(s) comprising: i) encoding and expressing a pilin domain; ii) knock out or overexpress an outer membrane protein regulator (ompR); iii) activate sigmaE through deletion of Anti-sigma-E factor rseA; iv) overexpress, mutate, or delete baeS and/or baeR; or v) comprise a mutation of cpxA, cpxR, and/or degP, or any combination thereof, wherein said genetically engineered bacteria cell allows large molecules, such as but not limited to intact folded proteins to pass through (or bypass) the outer membrane and enter the periplasmic space.

Description

SYSTEM AND METHOD FOR ALLOWING LARGE MOLECULES TO ENTER THE PERIPLASM OF E. COL!

FIELD OF THE INVENTION

The present invention relates to genetically engineered gram-negative bacteria that allow large molecules, such as but not limited to intact folded proteins, to pass through (or bypass) the outer membrane and enter the periplasmic space. Once in the periplasmic space, one possible practical application is to capture the protein using a bait or ligand protein that exists in the periplasm. More particularly, the genetically engineered gram-negative bacteria has been modified in at least one way, the modification(s) comprising: i) encoding and expressing a pilin domain; or ii) knock out or overexpress an outer membrane protein regulator { mpR) or iii) activate sigmaE through deletion of Anti-sigma-E factor rseA; or iv) overexpress, mutate, or delete baeS and/or baeR, or v) comprise a mutation of cpxA, cpxR, and/or degP, or any combination of i) to v), wherein said genetically engineered bacteria cell allows large molecules, such as but not limited to intact folded proteins, to pass through (or bypass) the outer membrane and enter the periplasmic space. The invention further relates to screening methods and nanobodies isolated using such methods.

BACKGROUND OF THE INVENTION

Escherichia coli is a workhorse model system for understanding bacterial physiology and regulation in addition to its importance as a human pathogen with increasing rates of antibiotic resistance. E. coli is a gram-negative bacterium with two membranes (inner and outer membranes); the intervening space is termed the periplasm and has been compared to the endoplasmic reticulum of eukaryotic cells. The outer membrane prevents many molecules, including antibiotics, from entering the cell.

The screening of very large protein libraries has been accomplished by a variety of techniques that rely on the display of proteins on the surface of viruses or cells (Ladner et al. 1993, U.S. Pat. No. 5,223,409) as well as other technologies. The underlying premise of the display technologies is that proteins engineered to be anchored on the external surface of biological particles (i.e., cells or viruses) are directly accessible for binding to ligands without the need for lysing the cells. Viruses or cells displaying proteins with affinity for a ligand can be isolated in a variety of ways including sequential adsorption/desorption from an immobilized ligand, by magnetic separations, or by flow cytometry.

The most widely used display technology for protein library screening applications is phage display. Phage display is a well-established and powerful technique for the discovery of proteins that bind to specific ligands and for the engineering of binding affinity and specificity. One of the most significant applications of phage display technology has been the isolation of high affinity antibodies (Maynard and Georgiou, Nat. Biotechnol., 20: 597-601 , 2002). However, several spectacular successes notwithstanding, the screening of phage- displayed libraries can be complicated by a number of factors. First, phage display imposes minimal selection for proper expression in bacteria by virtue of the low expression levels of the antibody fragment-gene III fusion necessary to allow phage assembly and yet sustain cell growth (Krebber et a/., Gene, 178: 71-74, 1996; Krebber et a/., J. Immunol. Methods, 201 : 35- 55, 1997). In addition, this is a fully in vitro technology and requires toggling between phage and E. coli stages.

Protein libraries have also been displayed on the surface of bacteria, fungi, or cells of other organisms. Cell-displayed libraries are typically screened by flow cytometry (Georgiou et al., Nat Biotechnol. 1997, 15(1): 29-34; Daugherty et al., J Immunol Methods. 2000, 243(1- 2): 211-27). However, just as in phage display, the protein has to be engineered for expression on the outer cell surface. This imposes several potential limitations. For example, the requirement for display of the protein on the surface of a cell imposes biological constraints that limit the diversity of the proteins and protein mutants that can be screened. Also, complex proteins consisting of several polypeptide chains cannot be readily displayed on the surface of bacteria, filamentous phage, or yeast.

There is a need in the art for technology which is an improvement on the above limitations and/or provides a different means for the screening for protein-protein interactions.

SUMMARY OF THE INVENTION

The present invention provides a method to achieve a remarkable phenotype. The method is the expression of some toxic proteins, which enables the phenotype of large proteins (or other molecules) being able to enter from outside the cell into the periplasm of E. coli. Once such large molecules enter the periplasm, we can capture them by having a “bait” protein expressed in the periplasm. The ability to capture these large molecules leads to a potentially useful technology application. We can perform in vivo screening for protein-protein interactions, which is immediately useful for selecting custom single domain antibodies (termed nanobodies) to arbitrary antigens. While other antibody production platforms exist, our system has a unique combination of features not available with any other existing platform. Specifically, the phenotype of enabling entry of large molecules into the periplasm allows combining direct assays of in vivo binding interactions (in the periplasm) with high throughput. In vivo conditions can alter association constants and dramatically increase local concentrations due to decreased diffusion. We hypothesize that these effects would provide additional unique benefits in increasing the success rate of our technology for discovery of functional antibodies in downstream applications over other in vitro antibody discovery platforms, thereby reducing overall cost.

In a first aspect, the present invention provides an isolated genetically engineered gram-negative bacteria cell having an inwardly leaky outer membrane, wherein the cell has been modified (including by transformation by at least one polynucleotide molecule or insertion, deletion, or mutation of any polynucleotide molecule within the cell, including the chromosome and endogenous or introduced plasmids); the modification comprising one or more of the following: i) encode and express a pilin domain; ii) knock out or overexpress an outer membrane protein regulator (ompF?); iii) activate sigmaE through deletion of Anti-sigma-E factor rseA; iv) overexpress, mutate, or delete baeS and/or baeR, or v) comprise a mutation of cpxA, cpxR, and/or degP, or any combination thereof, wherein said genetically engineered bacteria cell allows large molecules, such as intact folded proteins to pass from outside the cell into the periplasmic space.

In some embodiments, the pilin domain is a FimH pilin domain without the FimH lectin domain and without the FimC chaperone.

In some embodiments, resA is inactivated and baeS is overexpressed in the engineered cell. Such modifications enhance the bacterium inwardly leaky phenotype.

In some embodiments, the polynucleotide molecule comprises one or more genes which: encodes a pilin domain fused to a binding protein, wherein the binding protein is a protein such as a nanobody, fibronectin domain, or mammalian immunoglobulin domain.

In some embodiments, the fused binding protein is engineered to be expressed within the periplasmic space and binds to a target molecule that enters the periplasm from outside the cell.

In some embodiments, the bacteria is Escherichia coli.

In some embodiments, the at least one polynucleotide molecule comprising one or more genes is expressed from a plasmid.

In some embodiments, the pilin domain gene encodes a protein comprising an amino acid sequence set forth in SEQ ID NO: 1. In some embodiments, the amino acid sequence set forth in SEQ ID NO: 1 is encoded by a pilin domain gene comprises a polynucleotide sequence having at least 70% sequence identity, at least 80% sequence identity, at least 90% sequence identity or 100% sequence identity to the polynucleotide sequence set forth in SEQ ID NO: 2.

In some embodiments, the polynucleotide sequence is the codon optimized polynucleotide sequence set forth in SEQ ID NO: 3.

In a second aspect, the present invention provides a recombinant vector comprising a polynucleotide encoding a pilin domain, operably linked to a promoter, wherein an expressed pilin domain protein allows large molecules, such as intact folded proteins, to pass through the outer membrane into the periplasmic space.

In some embodiments, the vector further comprises a polynucleotide encoding a binding protein fused to the pilin domain, wherein an expressed binding protein is located to the periplasm.

In a third aspect, the present invention provides a method of screening for binding proteins in the periplasm, comprising: i) mixing or culturing the isolated genetically engineered gram-negative bacteria cell of any aspect of the invention with labelled target molecules; and ii) screening for binding of target molecules with said fused binding proteins.

In some embodiments, specific target binding within the periplasm is screened using fluorescence-activated cell sorting (FACS).

In some embodiments, a library of binding proteins, such as a nanobody library, is screened.

In a fourth aspect, the present invention provides a method of screening for antibiotic sensitivity, comprising: i) mixing or culturing the isolated genetically engineered gram-negative bacteria cell of any aspect of the invention in which at least one polynucleotide molecule comprises one or more genes which encodes a pilin domain fused to a binding protein, wherein the binding protein is a protein such as a nanobody, fibronectin domain, or mammalian immunoglobulin domain with antibiotic molecules; and ii) screening for sensitivity to said antibiotic.

In a fifth aspect, the present invention provides a method of in vivo screening for proteinprotein interaction, comprising: i) culturing the isolated genetically engineered gram-negative bacteria cell of any aspect of the invention in which at least one polynucleotide molecule comprises one or more genes which encodes a pilin domain fused to a binding protein, wherein the binding protein is a protein such as a nanobody, fibronectin domain, or mammalian immunoglobulin domain in a culture medium; ii) adding at least one dose of substance for testing to the culture medium; and iii) observing protein-protein interaction after a desired period of time.

In a sixth aspect, the present invention provides a kit for screening in vivo protein- protein interaction, comprising: at least one isolated genetically engineered gram-negative bacteria cell in which at least one polynucleotide molecule comprises one or more genes which encodes a pilin domain fused to a binding protein, wherein the binding protein is a protein such as a nanobody, fibronectin domain, or mammalian immunoglobulin domain; and a culture medium.

Several new nanobodies were isolated using the genetically engineered gram-negative bacteria cell and method of the invention. The isolated nanobodies bind to GFP, mCherry, IL- 6 or MHC class I polypeptide-related sequence B (MIC-B).

In a seventh aspect, the present invention provides an isolated nanobody comprising an amino acid sequence selected from the group comprising: i) the amino acid sequence set forth in SEQ ID NO: 5 which binds to GFP; ii) the amino acid sequence set forth in SEQ ID NO: 7 which binds to mCherry; iii) the amino acid sequence set forth in SEQ ID NO: 9 which binds to mCherry; iv) the amino acid sequence set forth in SEQ ID NO: 11 which binds to mCherry; v) the amino acid sequence set forth in SEQ ID NO: 13 which binds to IL-6; vi) the amino acid sequence set forth in SEQ ID NO: 15 which binds to IL-6; vii) the amino acid sequence set forth in SEQ ID NO: 17 which binds to IL-6; and viii) the amino acid sequence set forth in SEQ ID NO: 19 which binds to MHC class I polypeptide-related sequence B (MIC-B).

In some embodiments, the invention provides an isolated nanobody polynucleotide. In particular, the isolated polynucleotide molecule may comprise at least one nucleic acid sequence having at least 80%, at least 85%, at least 90%, or at least 95% sequence identity, due to redundancy of the genetic code, to a polynucleotide sequence selected from the group comprising: i) the polynucleotide sequence set forth in SEQ ID NO: 4 which encodes the amino acid sequence set forth in SEQ ID NO: 5; ii) the polynucleotide sequence set forth in SEQ ID NO: 6 which encodes the amino acid sequence set forth in SEQ ID NO: 7; iii) the polynucleotide sequence set forth in SEQ ID NO: 8 which encodes the amino acid sequence set forth in SEQ ID NO: 9; iv) the polynucleotide acid sequence set forth in SEQ ID NO: 10 which encodes the amino acid sequence set forth in SEQ ID NO: 11 ; v) the polynucleotide sequence set forth in SEQ ID NO: 12 which encodes the amino acid sequence set forth in SEQ ID NO: 13; vi) the polynucleotide sequence set forth in SEQ ID NO: 14 which encodes the amino acid sequence set forth in SEQ ID NO: 15; vii) the polynucleotide sequence set forth in SEQ ID NO: 16 which encodes the amino acid sequence set forth in SEQ ID NO: 17; and viii) the polynucleotide sequence set forth in SEQ ID NO: 18 which encodes the amino acid sequence set forth in SEQ ID NO: 19.

The isolated nucleic acid molecule according to the invention may be cloned into an expression vector, which may in turn be transformed into a host cell for the production of an antibody according to any aspect of the present invention. In particular, the host cell may be 293 cells, CHO cells, or recombinant plant cells.

BRIEF DESCRIPTION OF FIGURES

Figure 1 shows the expression of a nanobody-pilin fusion leads to specific accumulation of GFP in the periplasm, (a) Genetic constructs showing the domain structure of wt FimH, the nanobody-pilin fusion, and the individual domains, as well as the promoters driving their expression on plasmids, (b) FACS plots of MG 1655 cells carrying empty vector (left) and a plasmid expressing the nanobody-pilin fusion (right). Log GFP fluorescence is plotted on the X-axis. % of cells in each quadrant is indicated, (c) Quantification of GFP positive cells for different constructs in MG1655 as indicated on the X-axis, (d) Quantification of fluorescent cells when MG1655 expressing the nanobody-pilin fusion is mixed with the fluorescent protein indicated on the X-axis. Error bars represent standard deviation for n=3 biological replicates.

Figure 2 shows the fractionation of MG1655 cells expressing the nanobody-pilin fusion protein. Whole cells (lane 1), cytoplasm (lane 2), periplasm (lane 3), and membrane (lane 4) were assayed by Western blotting using a-FLAG antibodies (detecting the nanobody-pilin fusion). Fractionation controls were AP (phosphate, cytoplasmic), OmpA (outer membrane), and GroEL (cytoplasmic). The nanobody-pilin fusion protein is found in the periplasm.

Figure 3 shows the expression of the FimH pilin subunit leads to vancomycin sensitivity. MG1655 carrying various plasmids, all grown under induction conditions (25°C, LB + 0.2% arabinose, overnight), were plated on LB (left bars) or LB + 50 pg/mL vancomycin plates. Colony forming units are plotted on the y-axis. The plasmid carried by the strain is indicated by the bar color, as indicated by the colored labels at the top right. EV, empty vector (pBAD33); Pilin, pBAD33 carrying a gene encoding the FimH pilin domain; Nanobody-Pilin, pBAD33 carrying a gene encoding the Enhancer-pilin fusion. Error bars indicate the standard deviation for n=3 biological replicates.

Figure 4 shows that the permeability is not specific to GFP. (A) MG1655 expressing the LaM-4- Pilin fusion protein was mixed with purified mCherry or GFP, as indicated on the x- axis. The percent of fluorescent cells detected by FACS is plotted on the y-axis. Error bars represent standard deviation, n=3 biological replicates. (B) MG 1655 expressing the LaM-4- Pilin fusion protein was mixed simultaneously with purified mCherry and fluorescein-labeled vancomycin, then subjected to FACS analysis. Red fluorescence (mCherry) is plotted on the x-axis, green fluorescence (fluorescein-vancomycin) on the y-axis. Black vertical and horizontal lines separate the bulk of the histograms for red and green fluorescence, respectively. The numbers at the corners of the plot indicate the percentage of cells that fall into that quadrant of the graph.

Figure 5 shows that mutants in tol-pal and imp, which are known to leak periplasmic contents into the extracellular medium, do not have the same “inward leakiness” phenotype for large proteins or other large molecules. The percent of cells that accumulate GFP (assessed by FACS analysis) when mixed with purified GFP is indicated. Mutants in tol-pal and imp are indicated by different colors. Cells are transformed with an empty vector (taken to be wild type), a vector expressing a GFP-binding nanobody only, and a vector expressing the fusion protein of the GFP-binding nanobody with the pilin domain. Regardless of genotype, GFP accumulation depends on expression of the nanobody-pilin fusion protein.

Figure 6 shows the application of our permeability phenotype to measure specific Fos- Jun interactions, (a) Genetic constructs of fusion proteins used, (b) Quantification of GFP signal (indicative of Fos-Jun binding) for different combinations of Fos and Jun alleles, as indicated on the X-axis, (c) Sequence of the Fos and Jun leucine zipper domains indicating the residues mutated in (b).

Figure 7 shows (A) Histograms of GFP fluorescence for successive rounds of enrichment. For each round, the library of cells expressing the mutagenized nanobody library fused to the FimH pilin domain were mixed with GFP then analyzed by FACS. Green fluorescence is plotted on the x-axis, number of cells on the y-axis (arbitrary scale). Different colors indicate different rounds of the enrichment, as indicated by the label at the left. The black arrow for “3rd sort” indicates the cell population that was chosen for sequencing and validation. (B) All 10 clones that were sequenced from the “3rd sort” had an identical nanobody sequence, denoted NbG1. This sequence is aligned with the original GFP-binding Enhancer sequence. (C) Validation of GFP-binding by NbG1 by immunoprecipitation. NbG1 and LaM-4 were both fused to a FLAG epitope tag. GFP was fused to a 6xHis tag. Crude cell extracts were made from cells expressing each of these tagged proteins. The GFP-His extract was mixed in equal proportions with the NbG1 and LaM-4 extracts, then the mixtures were purified by nickel bead affinity chromatography. Proteins attached to the beads after washing were run on SDS-PAGE then blotted with anti-His or anti-FLAG antibodies. The first three lanes are crude extract controls. The right two lanes are the beads from the affinity purification of the mixed extracts. Only NbG1 , and not LaM-4, copurifies with His-GFP (compare lane 4 and 5, a-FLAG blot).

Figure 8 shows the discovery of a set of novel nanobodies that bind mCherry. (a) For each round, a library of cells expressing the mutagenized nanobody library was mixed with mCherry then analyzed by FACS. Sequencing of 10 clones from the second round of enrichment showed three different sequences, denoted as Nb_mCh2, Nb_mCh7, and Nb_mCh8. The predicted protein sequences are shown in an alignment with the published LaM-4 mCherry-binding nanobody, (b) Validation of mCherry binding using immunoprecipitation. The genes encoding Nb_mCh2, Nb_mCh7, and Nb_mCh8 were cloned and fused to a FLAG epitope tag. mCherry was cloned in a fusion with a 6xHis tag. Crude cell extracts were prepared from cells encoding each of the nanobodies and from another clone expressing the his-tagged mCherry. These extracts were mixed in a 1 :1 ratio, then affinity purified using a nickel bead column (to pull out the his-tagged mCherry). Proteins attached to the bead were then run on SDS-PAGE and blotted with anti-His (to confirm mCherry pulldown) and anti-FLAG antibodies (to check for pulldown of the nanobody by mCherry). The first lane (mCh) is a pure mCherry protein control. The next 4 lanes are crude extracts from cells expressing the indicated nanobody, without mixing with mCherry or any pulldown. The last four lanes (mCh Lam-4; mCh Nb_mCh7; mCh Nb_mCh3; mCh Nb_mCh8) are the washed beads after pulldown from the mixed mCherry and indicated nanobody crude cell extracts.

Figure 9 shows GFP accumulation is increased when rseA and baeS are mutated. Cells were transformed with the pTrc99A plasmid carrying a gene encoding the GFP-binding Enhancer nanobody fused to the pilin. Upon induction and mixing with purified GFP, cells were washed and analysed by FACS. The percentage of cells accumulating GFP increases from < 10% to -90% when the described mutations in rseA and baeS are introduced into MG 1655.

Figure 10 shows the immunoprecipitation assay for one of the IL-6 nanobodies discovered by the described System and Method. The left two lanes in each blot are input controls. The Nanobody is tagged with FLAG. The “Pulldown IgG” is a mouse monoclonal a- FLAG (which will pull down the FLAG-tagged Nanobody) (+) or a nonspecific mouse IgG (NS). The primary antibody for western is indicated at the bottom; a-IL-6 is a commercial rabbit polyclonal antibody. * indicates the heavy and light chains of IgG, which would react with the anti-mouse secondary antibody in the right panel.

Figure 11 shows far western using the custom IL-6 nanobody to detect purified IL-6. IL-6 was run on an SDS-PAGE gel, transferred to a PVDF membrane, and blotted with the same IL-6 nanobody shown in Figure 10. This nanobody is FLAG tagged; the signal seen is from a traditional western using an anti-FLAG primary antibody, resulting in a “far western” assay.

Figure 12 shows the size exclusion chromatography of three different IL-6 nanobodies. The arrow indicates the peak expected to represent the monomer, based on retention time.

DETAILED DESCRIPTION OF THE INVENTION

Bibliographic references mentioned in the present specification are for convenience listed in the form of a list of references and added at the end of the Examples. The whole content of such bibliographic references is herein incorporated by reference.

Definitions

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as is commonly understood by one of skill in the art to which the invention belongs. Certain terms employed in the specification, examples and appended claims are collected here for convenience.

It must be noted that as used herein and in the appended claims, the singular forms "a," "an," and "the" include plural reference unless the context clearly dictates otherwise. Thus, for example, a reference to "a target sequence" includes a plurality of such target sequences, and a reference to "an enzyme" is a reference to one or more enzymes and equivalents thereof known to those skilled in the art, and so forth.

The term "isolated" is herein defined as a biological component (such as a nucleic acid, peptide or protein) that has been substantially separated, produced apart from, or purified away from other biological components in the cell of the organism in which the component naturally occurs, i.e., other chromosomal and extra-chromosomal DNA and RNA, and proteins. Nucleic acids, peptides and proteins which have been isolated thus include nucleic acids and proteins purified by standard purification methods. The term also embraces nucleic acids, peptides and proteins prepared by recombinant expression in a host cell as well as chemically synthesized nucleic acids.

The phrases "nucleic acid" or "nucleic acid sequence," as used herein, refer to an oligonucleotide, nucleotide, polynucleotide, or any fragment thereof, to DNA or RNA of genomic or synthetic origin which may be single-stranded or double-stranded and may represent the sense or the antisense strand, to peptide nucleic acid (PNA), or to any DNA-like or RNA-like material.

As used herein, the term "oligonucleotide”, refers to a nucleic acid sequence of at least about 6 nucleotides to 60 nucleotides, preferably about 15 to 30 nucleotides, and most preferably about 20 to 25 nucleotides, which can be used in PCR amplification or in a hybridization assay or microarray. As used herein, the term "oligonucleotide" is substantially equivalent to the terms "amplimers," "primers," "oligomers," and "probes," as these terms are commonly defined in the art.

As used herein, a nanobody is a single-domain antibody (sdAb) and is an antibody fragment consisting of a single monomeric variable antibody domain. Similar to a whole antibody, it is able to bind selectively to a specific antigen.

As used herein, the term “comprising” or “including” is to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more features, integers, steps or components, or groups thereof. However, in context with the present disclosure, the term “comprising” or “including” also includes “consisting of”. The variations of the word “comprising”, such as “comprise” and “comprises”, and “including”, such as “include” and “includes”, have correspondingly varied meanings.

EXAMPLES

A person skilled in the art will appreciate that the present invention may be practiced without undue experimentation according to the method given herein. The methods, techniques and chemicals are as described in the references given or from protocols in standard biotechnology and molecular biology text books such as Green and Sambrook, Molecular Cloning: A Laboratory Manual, Cold Spring Harbour Laboratory, New York (2012).

EXAMPLE 1

METHODS

Materials and Methods

Optimization of the host E. coli strain MG1655

Two modifications were introduced into the host E. coli strain (MG1655) to make the “inward leaky” strain. This leaky strain is referred to as E. coli MG1655 rseA::kan baeS::po70. First was a knockout of the rseA gene. We used the Red recombinase system (described in (Murphy KC, Campellone KG. BMC Mol Biol. 2003 Dec 13;4:11. doi: 10.1186/1471-2199-4- 11)) to replace the rseA gene in E. coli MG1655 with a kanamycin cassette from pKD4 (Murphy KC, Campellone KG. BMC Mol Biol. 2003 Dec 13;4:11. doi: 10.1186/1471-2199-4-11). The deleted sequence spans coordinates 2708804-2709354 in MG 1655 (based on the Genbank NC_000913.3 reference sequence).

The second modification was the insertion of a strong promoter upstream of the baeS gene to overexpress it. We used a 2-step positive-negative selection system (described in (Khetrapal V, et al., Nucleic Acids Res. 2015 Jul 27;43(13):e83. doi: 10.1093/nar/gkv248) to insert a o70 promoter (designated pcr70, also named as “sigma70 promoter” in the sequence listing) directly upstream of the ATG start codon for the baeS gene (i.e. between coordinates 2162875 and 2162876 in MG1655). The inserted sequence was:

TTGACGGCTA GCTCAGTCCT AGGTACAGTG CTAGCTCATA ATTTTGTTTA ACTTTAAGAA GGAGATATA (SEQ ID NO: 27)

These modifications lead to a substantial increase in the percent of nanobodyexpressing cells that accumulate GFP (using the Enhancer nanobody as a control).

EXAMPLE 2

A GFP-binding phenotype dependent on expression of a nanobody-pilin fusion

The idea to assemble pili in which the FimH lectin was replaced with another immunoglobulin domain was previously attempted but unsuccessful (Munera D, Hultgren S, Fernandez LA. Mol Microbiol. 2007 Apr;64(2):333-46. doi: 10.1111/j.1365- 2958.2007.05657.x). In attempting to revisit this system, we made a nanobody-pilin fusion using the GFP-binding Enhancer nanobody (Figure 1A). We verified that this nanobody-pilin fusion did not assemble into pili (data not shown). However, purified GFP added extracellularly still associated with E. coli expressing the nanobody-pilin fusion. We developed a FACS-based assay to measure this binding (Figure 1 B-C). To prove that pilus assembly was not happening, we showed the GFP signal still occurred in a strain deleted for fimD, encoding the Type 1 pilus usher (Figure 1C). GFP binding required both the FimH pilin domain and the Enhancer nanobody (Figure 1C). The GFP binding was specific, as CFP and mCherry (which are not bound by this nanobody) (Kubala MH, et al., Protein Sci. 2010 Dec;19(12):2389-401. doi: 10.1002/pro.519) were not bound (Figure 1 D).

EXAMPLE 3

GFP binding is due to entry of GFP into the periplasm

We used cell fractionation to localize the proteins. The Enhancer domain was localized to the periplasm (Figure 2) - implying that GFP must gain access to the periplasm. Indicative of a “leaky” phenotype, cells expressing the nanobody-pilin fusion protein or the FimH pilin alone were more sensitive to vancomycin than cells expressing the Enhancer nanobody or carrying the empty expression vector (Figure 3).

EXAMPLE 4

Large molecules besides GFP can leak into the periplasm

To address the generality of this permeability, we tested for entry of other molecules. mCherry does not accumulate when the GFP-binding Enhancer nanobody was fused to the pilin subunit (Figure 1 D) However, fusion of an mCherry-binding LaM-4 nanobody (Fridy PC, et al., Nat Methods. 2014 Dec; 11 (12): 1253-60. doi: 10.1038/nmeth.3170) resulted in specific accumulation of mCherry and not GFP (Figure 4A). Fluorescein-labeled vancomycin is another large molecule that normally does not enter the periplasm. Expression of the mCherry- binding fusion protein enabled accumulation of vancomycin. Addition of extracellular mCherry and fluorescein-vancomycin showed that the same cells accumulate both mCherry and vancomycin, suggesting that the entry of both is through the same permeability defect (Figure 4B).

Two previously identified “leaky” strains are mutants in tol-pal and in the imp gene (Lloubes R, et al., Res Microbiol. 2001 Jul-Aug; 152(6):523-9. doi: 10.1016/s0923- 2508(01)01226-8; Sampson BA, Misra R, Benson SA. Genetics. 1989 Jul;122(3):491-501. doi: 10.1093/genetics/122.3.491). Expression of the nanobody-pilin fusion protein in the periplasm was still required for GFP to associate with the cells (Figure 5).

EXAMPLE 5

The permeability phenotype can be generalized to other protein-protein interactions

We took advantage of Fos and Jun, two proteins which are known to bind specifically to each other through leucine zippers (Lloubes R, et al., The Tol-Pal proteins of the Escherichia coli cell envelope: an energized system required for outer membrane integrity? Res Microbiol. 2001 Jul-Aug; 152(6):523-9. doi: 10.1016/s0923-2508(01)01226-8; Landschulz WH, Johnson PF, McKnight SL. Science. 1988 Jun 24;240(4860): 1759-64. doi: 10.1126/science.3289117). We fused the Jun leucine zipper to the FimH pilin (Pilin-Jun) and the Fos leucine zipper to GFP (Fos-GFP) (Figure 7A). Expression of Pilin-Jun in the periplasm of E. coli followed by incubation with purified Fos-GFP resulted in fluorescence accumulation. The binding strength of different mutants in Fos and Jun has been previously characterized. We saw similar results in our assay; an L4V mutation in Jun has only a minor effect on Fos binding (Waraho D, DeLisa MP. Proc Natl Acad Sci U S A. 2009 Mar 10;106(10):3692-7. doi: 10.1073/pnas.0704048106.), and this mutation in Pilin-Jun still accumulates GFP. In contrast, an L2V mutation in Fos abolishes interaction with both wt and L4V Jun (Waraho D, DeLisa MP. Proc Natl Acad Sci U S A. 2009 Mar 10; 106(10):3692-7. doi: 10.1073/pnas.0704048106.); we also saw no accumulation of L2V Fos-GFP in any Pilin-Jun expressing cells (Figure 6).

EXAMPLE 6

The permeability phenotype can be used to discover new nanobodies that bind to GFP and to mCherry

We adapted our FACS-based assay to screen for new GFP- and mCherry-binding nanobodies as a proof-of-concept application. We have access to two nanobody libraries: a camelized human VH3 library designed for phage display (Creative Biolabs, Shirley, NY, USA) and a fully synthetic nanobody library designed for yeast display (generous gift of Andrew Kruse, Harvard Medical School) (McMahon C, et al., Nat Struct Mol Biol. 2018 Mar;25(3):289- 296. doi: 10.1038/s41594-018-0028-6). We fused the synthetic library to the fimH pilin, so that each cell in the library should express a different nanobody-pilin fusion. We then added GFP (or mCherry) extracellularly and screened by FACS as in Example 2. After three rounds of enrichment (two rounds for mCherry), we saw a population of cells with higher fluorescence (Figure 7A and data not shown). Sequencing of 10 clones showed these all had the same nanobody sequence for GFP, and three different sequences for mCherry. Of note, the new GFP nanobody sequence (termed NbG1 ; polynucleotide sequence SEQ ID NO: 4 and amino acid sequence SEQ ID NO: 5) was distinct from the Enhancer nanobody sequence, and the three new mCherry nanobody sequences were all distinct from the known mCherry-binding Lam-4 nanobody sequence (Figure 7B, 8A). We cloned the new nanobody sequences, tagged them with 6xHis residues, and purified them. We then verified that they specifically bound GFP (or mCherry) by immunoprecipitation (Figure 7C, 8B). The first mCherry nanobody has the polynucleotide sequence set forth in SEQ ID NO: 6, and an amino acid sequence as set forth in SEQ ID NO: 7. The second mCherry nanobody has the polynucleotide sequence set forth in SEQ ID NO: 8, and an amino acid sequence as set forth in SEQ ID NO: 9. The third mCherry nanobody has the polynucleotide sequence set forth in SEQ ID NO: 10, and an amino acid sequence as set forth in SEQ ID NO: 11.

EXAMPLE 7

Expanded application of the described system and method on additional proteins

The library was induced in the engineered leaky E. coli strain MG1655 rseA::kan baeS::pcr70 at 25°C. The bacterial cells were grown overnight at 25°C in LB supplemented with 50 pg/ml chloramphenicol and kanamycin. The overnight culture was diluted to OD 0.1 and continued to grow until it reached to OD 0.3. Subsequently, the culture was induced with 20% arabinose at final concentration of 0.2% for 6 to 8 hrs with shaking at 160 rpm at 25°C.

10⁸ to 10⁹ cells from above culture were washed twice with PBS and resuspended into 150 ul of TN Buffer (10mM Tris pH 8, 300mM NaCI) before adding 50ul of following purified proteins independently: GFP or mCherry proteins or IL-6 or MHC class I polypeptide-related sequence A (MICA) or MHC class I polypeptide-related sequence B (MICB) coupled with AF488 or AF546. The mixture was incubated O/N at 4°C with rotation. Subsequently, the cells were washed thrice with PBS before subjecting to sorting using S3e Cell Sorter from BIORAD.

EXAMPLE 8

Improved induction conditions

The following changes were made to the induction conditions described in Example 7:

1. After transformation of the nanobody library into host cells, 2% glucose and 2% fucose were supplemented during recovery (after electroporation).

2. The induction of the nanobody-expressing cells was shortened from overnight to 6-8 hours.

3. The media used to recover cells after FACS was supplemented with 2% glucose and 2% fucose.

The use of glucose and fucose (both of which help to repress the arabinose promoter controlling nanobody expression; note that fucose also cannot be metabolized by the host E. coli strain) helps to reduce some false positive hits in the screening due to more rapid growth of non-nanobody expressing cells (for example, clones carrying truncated genes or genes with nonsense mutations). The shorter induction time also reduces the period where nanobody expression reduces growth, helping to address the same issue.

Improved FACS conditions

The following heuristics are applied to FACS sorting to enrich for binding clones. These optimizations provide a better balance of maintaining multiple binding clones while still achieving reasonable enrichment in each round (sorting too stringently might provide foldenrichment in each round but risks losing some (presumably lower affinity) binding clones). Each run begins with an initial 5 minute analysis run to set the gating based on the overall fluorescence signal distribution. If there is a clear gap between two populations (assigned putatively as the negative and positive binding populations), then the gating is set in that gap to exclude the negative (lower fluorescence) population. If the populations are not fully separated, then the gating is set to include most of the putative positive population (assuming a Gaussian distribution for both populations), leading to inclusion of parts of the putative negative population. The estimated enrichment of binding clones using these heuristics ranges from 100* (when the populations are well separated) to 5-1 Ox (when the populations are not well separated).

EXAMPLE 9

Isolation and initial characterization of multiple nanobodies to a single human protein

Three distinct nanobodies to the human IL-6 protein were isolated over the course of two applications of the described system and method in Examples 1 and 8. These were subjected to two binding validation assays: immunoprecipitation and (far) western blot. These were also subjected to characterization of potential aggregation using size exclusion chromatography.

All 3 custom nanobodies demonstrated detectable immunoprecipitation of purified IL-6 (data for one is shown in Figure 10). One of the nanobodies was further tested for immunoprecipitation of IL-6 produced by 5637 cells (a human bladder epithelial cell line); this nanobody was also able to immunoprecipitate IL-6 from these cell lysates. This same nanobody was tested as the primary antibody in a far western blot to detect purified IL-6 and IL-6 from 5637 cell lysates; this nanobody was able to mediate detection of IL-6 in both instances (Figure 11).

All 3 custom nanobodies were characterized with size exclusion chromatography. Two of them were largely (estimated >90%) monomeric, while one showed aggregates that would generally be unfavourable for other applications (Figure 12).

The first IL-6 nanobody has the polynucleotide sequence set forth in SEQ ID NO: 12, and an amino acid sequence as set forth in SEQ ID NO: 13. The second IL-6 nanobody has the polynucleotide sequence set forth in SEQ ID NO: 14, and an amino acid sequence as set forth in SEQ ID NO: 15. IL-6 nanobody has the polynucleotide sequence set forth in SEQ ID NO: 16, and an amino acid sequence as set forth in SEQ ID NO: 17.

A nanobody to the human MIC-B protein was also isolated and found to have binding activity. The MIC-B nanobody has the polynucleotide sequence set forth in SEQ ID NO: 18, and an amino acid sequence as set forth in SEQ ID NO: 19. The results for creating custom nanobodies to different protein antigens is summarized in Table 1.

Table 1

The described system and method was used for a total of 12 selections, each consisting of 2-3 rounds. The average time for one round of enrichment (antigen binding, sorting, and regrowth) is less than 1 week.

Summary

There are 3 key features of the present invention:

1. Specific genetic features (including a novel fusion protein)

This enables the entire system - we can allow large molecules (i.e. intact, folded proteins) to cross the outer membrane of E. coli and enter the periplasm. This is the key observation that enables all the custom antibody screening in an in vivo compartment.

2. Ability to screen for binding proteins in an in vivo compartment

To the best of our knowledge this is unique to our system - every other in vitro system for screening antibodies, even if it uses live cells, does the binding on the surface, i.e. in the extracellular space. This potentially could allow selection of binding proteins (such as nanobodies) with different binding properties than other existing antibody systems.

3. Allows very large molecules (i.e. intact, folded proteins) to cross the outer membrane of E. coli. This is unique to our system. Besides the method described here for screening for nanobodies, this feature could also have other applications such as for antibiotic screening or synthetic biology applications. REFERENCES

1. Daugherty PS, et al., Flow cytometric screening of cell-based libraries. J Immunol Methods. 2000, 243(1-2): 211-27.

2. Fridy PC, et al., A robust pipeline for rapid production of versatile nanobody repertoires. Nat Methods. 2014 Dec; 11 (12): 1253-60. doi: 10.1038/nmeth.3170.

3. Georgiou G, et al., Display of heterologous proteins on the surface of microorganisms: from the screening of combinatorial libraries to live recombinant vaccines. Nat Biotechnol. 1997, 15(1): 29-34.

4. Khetrapal V, et al., A set of powerful negative selection systems for unmodified Enterobacteriaceae. Nucleic Acids Res. 2015 Jul 27;43(13):e83. doi: 10.1093/nar/gkv248.

5. Krebber et al., Inclusion of an upstream transcriptional terminator in phage display vectors abolishes background expression of toxic fusions with coat protein g3p. Gene, 178: 71-74, 1996.

6. Krebber et al., Reliable cloning of functional antibody variable domains from hybridomas and spleen cell repertoires employing a reengineered phage display system. J. Immunol. Methods, 201 : 35-55, 1997.

7. Kubala MH, et al., Structural and thermodynamic analysis of the GFP:GFP-nanobody complex. Protein Sci. 2010 Dec;19(12):2389-401. doi: 10.1002/pro.519.

8. Ladner et al. 1993, U.S. Pat. No. 5,223,409.

9. Landschulz WH, Johnson PF, McKnight SL. The leucine zipper: a hypothetical structure common to a new class of DNA binding proteins. Science. 1988 Jun 24;240(4860): 1759-64. doi: 10.1126/science.3289117.

10. Lloubes R, et al., The Tol-Pal proteins of the Escherichia coli cell envelope: an energized system required for outer membrane integrity? Res Microbiol. 2001 Jul-Aug; 152(6):523-9. doi: 10.1016/s0923-2508(01)01226-8.

11. Maynard and Georgiou, Protection against anthrax toxin by recombinant antibody fragments correlates with antigen affinity. Nat. Biotechnol. 20: 597-601 , 2002.

12. Munera D, Hultgren S, Fernandez LA. Recognition of the N-terminal lectin domain of FimH adhesin by the usher FimD is required for type 1 pilus biogenesis. Mol Microbiol. 2007 Apr;64(2):333-46. doi: 10.1111/j.1365-2958.2007.05657.x.

13. Murphy KC, Campellone KG. Lambda Red-mediated recombinogenic engineering of enterohemorrhagic and enteropathogenic E. coli. BMC Mol Biol. 2003 Dec 13;4:11 . doi: 10.1186/1471-2199-4-11.

14. McMahon C, et al., Yeast surface display platform for rapid discovery of conformationally selective nanobodies. Nat Struct Mol Biol. 2018 Mar;25(3):289-296. doi: 10.1038/s41594-018- 0028-6. 15. Sampson BA, Misra R, Benson SA. Identification and characterization of a new gene of Escherichia coli K-12 involved in outer membrane permeability. Genetics. 1989 Jul;122(3):491-501. doi: 10.1093/genetics/122.3.491.

16. Waraho D, DeLisa MP. Versatile selection technology for intracellular protein-protein interactions mediated by a unique bacterial hitchhiker transport mechanism. Proc Natl Acad

Sci U S A. 2009 Mar 10;106(10):3692-7. doi: 10.1073/pnas.0704048106.

Claims

1. An isolated genetically engineered gram-negative bacteria cell having an inwardly leaky outer membrane, wherein the cell has been modified (including by transformed by at least one polynucleotide molecule or insertion, deletion, or mutation of any polynucleotide molecule within the cell, including the chromosome and endogenous or introduced plasmids); the modification comprising one or more of the following: i) encode and express a pilin domain; ii) knock out or overexpress an outer membrane protein regulator (ompF?); iii) activate sigmaE through deletion of Anti-sigma-E factor rseA; iv) overexpress, mutate, or delete baeS and/or baeR, or v) comprise a mutation of cpxA, cpxR, and/or degP, or any combination thereof, wherein said genetically engineered bacteria cell allows large molecules, such as intact folded proteins to pass from outside the cell into the periplasmic space.

2. The isolated genetically engineered gram-negative bacteria cell of claim 1 or 2, wherein the pilin domain is a FimH pilin domain without the FimH lectin domain and without the FimC chaperone.

3. The isolated genetically engineered gram-negative bacteria cell of claim 1 , wherein resA is inactivated and baeS is overexpressed.

4. The isolated genetically engineered gram-negative bacteria cell of any one of claims 1 to 3, wherein the polynucleotide molecule comprises one or more genes which: encodes a pilin domain fused to a binding protein, wherein the binding protein is a protein such as a nanobody, fibronectin domain, or mammalian immunoglobulin domain.

5. The isolated genetically engineered gram-negative bacteria cell of claim 4, wherein the fused binding protein is engineered to be expressed within the periplasmic space and binds to a target molecule that enters the periplasm from outside the cell.

6. The isolated genetically engineered gram-negative bacteria cell of any one of claims 1 to 5, wherein said bacteria is Escherichia coli.

7. The isolated genetically engineered gram-negative bacteria cell of any one of claims 1 to

6, wherein said at least one polynucleotide molecule comprising one or more genes is expressed from a plasmid.

8. The isolated genetically engineered gram-negative bacteria cell of any one of claims 1 to

7, wherein said pilin domain gene encodes a protein comprising an amino acid sequence set forth in SEQ ID NO: 1.

9. The isolated genetically engineered gram-negative bacteria cell of claim 8, wherein said pilin domain gene comprises a polynucleotide sequence having at least 70% sequence identity, at least 80% sequence identity, at least 90% sequence identity or 100% sequence identity to the sequence set forth in SEQ ID NO: 2, such as the codon optimized polynucleotide sequence set forth in SEQ ID NO: 3.

10. A recombinant vector comprising a polynucleotide encoding a pilin domain, operably linked to a promoter, wherein an expressed protein allows large molecules, such as intact folded proteins, to pass through the outer membrane into the periplasmic space.

11. The recombinant vector of claim 10, further comprising a polynucleotide encoding a binding protein fused to the pilin domain, wherein an expressed protein is located to the periplasm.

12. A method of screening for binding proteins in the periplasm, comprising i) mixing or culturing the isolated genetically engineered gram-negative bacteria cell of any one of claims 4 to 9 with labelled target molecules; and ii) screening for binding of target molecules with said fused binding proteins.

13. The method of claim 12, wherein specific target binding within the periplasm is screened using fluorescence-activated cell sorting (FACS).

14. The method of claim 12 or 13, wherein a library of binding proteins, such as a nanobody library, is screened.

15. A method of screening for antibiotic sensitivity, comprising i) mixing or culturing the isolated genetically engineered gram-negative bacteria cell of any one of claims 1 to 3 with antibiotic molecules; and ii) screening for sensitivity to said antibiotic.

16. A method of in vivo screening for protein-protein interaction, comprising: culturing the isolated genetically engineered gram-negative bacteria cell of any one of claims 4 to 9 in a culture medium; adding at least one dose of substance for testing to the culture medium; and observing protein-protein interaction after a desired period of time.

17. An isolated nanobody comprising an amino acid sequence selected from the group comprising: i) the amino acid sequence set forth in SEQ ID NO: 5 which binds to GFP; ii) the amino acid sequence set forth in SEQ ID NO: 7 which binds to mCherry; iii) the amino acid sequence set forth in SEQ ID NO: 9 which binds to mCherry; iv) the amino acid sequence set forth in SEQ ID NO: 11 which binds to mCherry; v) the amino acid sequence set forth in SEQ ID NO: 13 which binds to IL-6; vi) the amino acid sequence set forth in SEQ ID NO: 15 which binds to IL-6; vii) the amino acid sequence set forth in SEQ ID NO: 17 which binds to IL-6; and viii) the amino acid sequence set forth in SEQ ID NO: 19 which binds to MHC class I polypeptide-related sequence B (MIC-B).

18. The isolated nanobody of claim 17, wherein the nanobody is encoded by a polynucleotide sequence having at least 80% sequence identity, due to redundancy of the genetic code, to a polynucleotide sequence selected from the group comprising: i) the polynucleotide sequence set forth in SEQ ID NO: 4 which encodes the amino acid sequence set forth in SEQ ID NO: 5; ii) the polynucleotide sequence set forth in SEQ ID NO: 6 which encodes the amino acid sequence set forth in SEQ ID NO: 7; iii) the polynucleotide sequence set forth in SEQ ID NO: 8 which encodes the amino acid sequence set forth in SEQ ID NO: 9; iv) the polynucleotide sequence set forth in SEQ ID NO: 10 which encodes the amino acid sequence set forth in SEQ ID NO: 11 ; v) the polynucleotide sequence set forth in SEQ ID NO: 12 which encodes the amino acid sequence set forth in SEQ ID NO: 13; vi) the polynucleotide sequence set forth in SEQ ID NO: 14 which encodes the amino acid sequence set forth in SEQ ID NO: 15; vii) the polynucleotide sequence set forth in SEQ ID NO: 16 which encodes the amino acid sequence set forth in SEQ ID NO: 17; and viii) the polynucleotide sequence set forth in SEQ ID NO: 18 which encodes the amino acid sequence set forth in SEQ ID NO: 19.

19. A kit for screening in vivo protein-protein interaction, comprising: at least one isolated genetically engineered gram-negative bacteria cell of any one of claims 4 to 9; and a culture medium.

22