HK40090619A - Yeast-based immunotherapy against clostridium difficile infection - Google Patents

Description

Yeast-based immunotherapy for Clostridium difficile infection

This application is a divisional application of application number 201680072004.8, filed on October 13, 2016, entitled "Yeast-based immunotherapy for Clostridium difficile infection".

Statement on Federal Government Sponsorship of Research and Development

This invention was completed with government support under grant numbers DK084509 and AI109776 granted by the National Institutes of Health. The government holds certain rights to this invention.

sequence list

The sequence list in electronic (ASCII text file) format is submitted with this application and is incorporated herein by reference. The ASCII file is named "2016_1343A_ST25.txt"; it was created on October 13, 2016; and its size is 407KB.

Background Technology

Clostridium difficile is the most common cause of hospital-acquired antibiotic-associated diarrhea and the pathogen of pseudomembranous colitis[1]. It is estimated that more than 500,000 cases of Clostridium difficile-associated disease (CDI) occur annually in the United States, with an annual mortality rate of approximately 3–17%, depending on the strain. With the emergence of highly virulent and antibiotic-resistant strains, the mortality rate of CDI patients has increased rapidly[2].

CDI is primarily caused by two Clostridium difficile exotoxins, TcdA and TcdB (since TcdA-TcdB strains are non-toxic) [21,22]. These two toxins are structurally similar and exhibit similar modes of action on host cells. Both toxins target host Rho GTPases, leading to their inactivation and cytoskeleton disruption. The related roles of the two toxins in the pathogenesis of CDI are not well understood, but it is clear that either toxin alone can induce CDI in animals [22,23].

Treatment options for patients with CDI are limited, and relapse rates are high (20–35% of patients). Current standard treatment for CDI with antibiotics leads to microbiome disruption and a relapse rate approaching 35% [3,13]. Despite attempts at other interventions (e.g., probiotics, toxin-absorbing polymers, and toxin-like vaccines), prevention and treatment strategies have not kept pace with the increasing incidence and severity of this infection. The risk of further CDI flare-ups may exceed 50% in relapsed patients [14], and a subset of patients experience multiple relapses. Relapses of CDI can be caused by the same strain or newly colonized strains [15–18].

Newer immunotherapy-based therapies have shown slight efficacy in clinical trials, including intravenous immunoglobulin (IVIG) for severe CDI[4-8] and human monoclonal antibodies for recurrent CDI[9]. Fidaxomicin, a narrow-spectrum macrocyclic antibiotic, has shown similar efficacy to oral vancomycin for CDI, but is more effective in reducing recurrence rates[10]. Fecal transplantation is effective for refractory and recurrent CDI, but is difficult to standardize and is risk-associated[11,12].

CDI is a frustrating disease that is difficult to treat and can affect patients for months or even years, resulting in significant morbidity and mortality[19]. Therefore, there is a need for new therapies for CDI and means to prevent primary and recurrent CDI in subjects at risk of developing CDI.

Summary of the Invention

This article presents antibody-based fusion protein binders that selectively bind to Clostridium difficile virulence factors TcdA and TcdB, and genetically engineered Saccharomyces strains to express and secrete these Clostridium difficile toxin binders. Both the yeast and the binders have shown practical applicability in treating and preventing primary and recurrent CDI in subjects. Oral administration of the Saccharomyces strain, which secretes the binders in the host gut, alleviated ongoing CDI and prevented recurrence.

Therefore, the present invention relates to Clostridium difficile toxin binders, yeast strains, including but not limited to Saccharomyces boulardii engineered to produce binders, methods for preparing engineered yeast strains, and methods for treating and preventing primary and recurrent CDI using binders and engineered yeast strains, as well as other important characteristics.

binder

The binders of the present invention include simple _VHH peptide monomers and linking groups of _VHH peptide monomers (including 2, 3, 4 or more monomers) as well as more complex binders including _VHH peptide monomers linked to the Fc domain of an antibody and _VHH peptide monomers linked to partial or complete IgG antibodies.

In a first embodiment, the present invention relates to a conjugate comprising a _VHH peptide monomer and a linking group comprising 2, 3, 4 or more monomers of the _VHH peptide monomer, wherein each of the monomers binds to TcdA and/or TcdB, preferably with specificity. Therefore, the present invention covers _VHH peptide conjugates comprising at least one _VHH peptide monomer, wherein each _VHH peptide monomer has binding specificity to an epitope of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). In some aspects, these conjugates comprise 2, 3, 4 or more linked _VHH peptide monomers. _VHH peptide monomers include, but are not limited to, _VHH peptide monomers 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7).

In several aspects of this embodiment, in which two or more monomers are linked, the monomers may be linked by flexible peptide linkers, typically containing 10-20 amino acids. Suitable linkers include, but are not limited to, linker-1 (SEQ ID NO: 9), linker-2 (SEQ ID NO: 11), and linker-3 (SEQ ID NO: 13).

In some aspects of this embodiment, the binder specifically binds to TcdA and/or TcdB. In some aspects of this embodiment, the binder exhibits TcdA and/or TcdB neutralizing activity.

In a specific aspect of this embodiment, the binder comprises four linked _VHH peptide monomers, wherein two monomers are specific for binding to the epitope of TcdA, and two monomers are specific for binding to the epitope of TcdB. The epitopes of TcdA may be the same or different. The epitopes of TcdB may be the same or different.

In a specific aspect of this embodiment, the binder comprises the amino acid sequence shown in SEQ ID NO: 19 or a sequence variant thereof having at least 95% sequence identity with it, wherein said sequence variant retains the binding specificity of TcdA and/or TcdB, or the sequence variant retains toxin-neutralizing activity, or both. In some cases, the variant amino acid of the sequence variant is located in the frame region of the _VHH peptide monomer.

In a second embodiment, the present invention relates to a conjugate comprising a _VHH peptide monomer linked to an IgG antibody, wherein the conjugate binds TcdA and/or TcdB. In these IgG-based conjugates, the variable regions of the light and heavy chains of the IgG antibody are replaced by one, two, three, four, or more _VHH peptide monomers.

In some aspects of this embodiment, these binders comprise 2, 3, 4 or more linked _VHH peptide monomers connected to the amino termini of the IgG light and heavy chains in place of the variable regions. _VHH peptide monomers include, but are not limited to, _VHH peptide monomers 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7).

In several aspects of this embodiment, in which two or more monomers are linked, the monomers may be linked by flexible peptide linkers, typically containing 10-20 amino acids. Suitable linkers include, but are not limited to, linker-1 (SEQ ID NO: 9), linker-2 (SEQ ID NO: 11), and linker-3 (SEQ ID NO: 13).

In a first sub-in embodiment, the present invention relates to a tetraspecific octamer binder comprising an IgG antibody, two sets of linked first and second _VHH peptide monomers, and two sets of linked third and fourth _VHH peptide monomers, wherein the IgG antibody comprises two arms, each arm comprising a light chain lacking a variable region and a heavy chain lacking a variable region, and each chain having an amino terminus, wherein for each arm of the antibody, one set of linked first and second _VHH peptide monomers is linked to the amino terminus of the light chain, and one set of linked third and fourth _VHH peptide monomers is linked to the amino terminus of the heavy chain, and wherein the _VHH peptide monomers have binding specificity to epitopes of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). This binder is referred to as "tetraspecific" because it recognizes four different toxin epitopes. Because it carries eight _VHH peptide monomers (two copies of the first monomer, two copies of the second monomer, two copies of the third monomer, and two copies of the fourth monomer), it is referred to as an "octamer".

In this sub-example, the first, second, third, and fourth _VHH peptide monomers each have binding specificity to different epitopes.

In some aspects of this sub-example, the two _VHH peptide monomers have binding specificity to the epitope of TcdA, and the two _VHH peptide monomers have binding specificity to the epitope of TcdB.

In some aspects of this sub-example, the _VHH peptide monomer independently exhibits binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

In a specific aspect of this sub-example, the light (kappa) chain of the binder comprises the amino acid sequence shown in SEQ ID NO: 46 (AA6/E3kappa) or a sequence variant having at least 95% sequence identity therewith, and the heavy chain of the binder comprises the amino acid sequence shown in SEQ ID NO: 44 (AH3/5D heavy) or a sequence variant having at least 95% sequence identity therewith. Since this binder is an IgG-based binder, it will be apparent to those skilled in the art that the two heavy chain polypeptides and two light chain polypeptides having said amino acid sequences will assemble via disulfide bonds to provide a complete binder. The sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin-neutralizing activity or both. The variant amino acids of the sequence variant may be located within the frame region of the _VHH peptide monomer.

In a second sub-in embodiment, the present invention relates to a bispecific or tetraspecific tetrameric binder comprising an IgG antibody and first, second, third, and fourth _VHH peptide monomers, wherein the IgG antibody comprises two arms, each arm comprising a heavy chain lacking a variable region and a light chain lacking a variable region, and each chain having an amino terminus, wherein for the first arm of the antibody, a first _VHH peptide monomer is attached to the amino terminus of the light chain, and a second _VHH peptide monomer is attached to the amino terminus of the heavy chain, wherein for the second arm of the antibody, a third _VHH peptide monomer is attached to the amino terminus of the light chain, and a fourth _VHH peptide monomer is attached to the amino terminus of the heavy chain, and wherein the _VHH peptide monomers have binding specificity to epitopes of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). When the binder is “tetraspecific,” it recognizes four different toxin epitopes; when it is “bispecific,” it recognizes two different toxin epitopes. Because they carry four _VHH peptide monomers, the binders are "tetrameric" (when bispecific, the first and third monomers have the same sequence and bind the same epitope, and the second and fourth monomers have the same sequence and bind the same epitope; when tetraspecific, each monomer has a different sequence and binds a different epitope).

When the binder is bispecific, the first and second monomers have binding specificity for different epitopes, the first and third monomers have the same amino acid sequence, and the second and fourth monomers can have the same amino acid sequence. One of the _VHH peptide monomers can have binding specificity for the TcdA epitope, and one of the _VHH peptide monomers can have binding specificity for the TcdB epitope.

When the binding agent is quadruple specific, each _VHH peptide monomer has binding specificity for a different epitope. Two _VHH peptide monomers can bind specifically to the TcdA epitope, and two _VHH peptide monomers can bind specifically to the TcdB epitope.

In some aspects of this sub-example, each _VHH peptide monomer has binding specificity to the epitope of TcdA.

In some aspects of this sub-example, each _VHH peptide monomer has binding specificity to the epitope of TcdB.

In some aspects of this sub-example, the VHH peptide monomer independently exhibits binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

In a specific aspect of this sub-example, the light (kappa) chain of the binder comprises the amino acid sequence shown in SEQ ID NO: 40 (AA6 kappa) or a sequence variant having at least 95% sequence identity therewith, and the heavy chain of the binder comprises the amino acid sequence shown in SEQ ID NO: 36 (AH3 heavy) or a sequence variant having at least 95% sequence identity therewith. Since this binder is an IgG-based binder, it will be apparent to those skilled in the art that the two heavy chain polypeptides and two light chain polypeptides having said amino acid sequences will assemble via disulfide bonds to provide a complete binder. The sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin-neutralizing activity or both. The variant amino acids of the sequence variant may be located within the frame region of the _VHH peptide monomer.

In another specific aspect of this sub-example, the light (kappa) chain of the binder comprises the amino acid sequence shown in SEQ ID NO: 42 (E3kappa) or a sequence variant having at least 95% sequence identity therewith, and the heavy chain of the binder comprises the amino acid sequence shown in SEQ ID NO: 38 (5D heavy) or a sequence variant having at least 95% sequence identity therewith. Since this binder is an IgG-based binder, it will be apparent to those skilled in the art that the two heavy chain polypeptides and two light chain polypeptides having said amino acid sequences will assemble via disulfide bonds to provide a complete binder. The sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin-neutralizing activity or both. The variant amino acids of the sequence variant may be located within the frame region of the _VHH peptide monomer.

In some aspects of this embodiment and its sub-embodiments, the binder specifically binds to TcdA and/or TcdB. In some aspects of this embodiment, the binder exhibits TcdA and/or TcdB neutralizing activity.

In a third embodiment, the present invention relates to a binder comprising _VHH peptide monomers linked to the Fc domain of an antibody, wherein the binder binds TcdA and/or TcdB. In these Fc domain-based binders, one, two, three, four, or more _VHH peptide monomers are linked to the hinge, _CH2 , and _CH3 regions of each arm of the Fc domain of the antibody heavy chain. Thus, the peptide monomers replace the Fab region of the antibody.

In some aspects of this embodiment, these binders comprise 2, 3, 4 or more linked _VHH peptide monomers at the amino terminus of the arm connected to the Fc domain. _VHH peptide monomers include, but are not limited to, _VHH peptide monomers 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7).

In this embodiment where two or more monomers are linked, the monomers may be linked by flexible peptide linkers, typically comprising 10-20 amino acids. Suitable linkers include, but are not limited to, linker-1 (SEQ ID NO: 9), linker-2 (SEQ ID NO: 11), and linker-3 (SEQ ID NO: 13).

In a first sub-in embodiment, the present invention relates to a tetraspecific octamer binder comprising an antibody Fc domain and two sets of linked first, second, third, and fourth _VHH peptide monomers, wherein the antibody Fc domain comprises two arms, each arm comprising a hinge of an antibody heavy chain, _CH2 and _CH3 regions, and each arm having an amino terminus, wherein for each arm of the Fc domain, a set of linked first, second, third, and fourth _VHH peptide monomers is linked to the amino terminus of the arm, and wherein the _VHH peptide monomers have binding specificity to epitopes of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). This binder is referred to as "tetraspecific" because it recognizes four different toxin epitopes. Because it carries eight _VHH peptide monomers (two copies of the first monomer, two copies of the second monomer, two copies of the third monomer, and two copies of the fourth monomer), it is referred to as an "octamer."

In some aspects of this sub-example, the first, second, third and fourth _VHH peptide monomers each have binding specificity for different epitopes.

In some aspects of this sub-example, the two _VHH peptide monomers have binding specificity to the epitope of TcdA, and the two _VHH peptide monomers have binding specificity to the epitope of TcdB.

In some aspects of this sub-example, the _VHH peptide monomer independently exhibits binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

In a specific aspect of this sub-example, the binder comprises the amino acid sequence shown in SEQ ID NO: 22 (ABAB-Fc) or a sequence variant having at least 95% sequence identity with it, wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity, or both. Since this binder is based on the Fc domain, it will be clear to those skilled in the art that two identical polypeptides having the stated amino acid sequence act as arms of the binder, and the arms will assemble via disulfide bonds to provide the complete binder. The variant amino acid of the sequence variant may be located within the frame region of the _VHH peptide monomer.

In a second sub-in embodiment, the present invention relates to a bispecific tetrameric binder comprising an antibody Fc domain and two sets of linked first and second _VHH peptide monomers, wherein the antibody Fc domain comprises two arms, each arm comprising a hinge of the antibody heavy chain, _CH2 and _CH3 regions, and each arm having an amino terminus, wherein for each arm of the Fc domain, a set of linked first and second _VHH peptide monomers is linked to the amino terminus of the arm, and wherein the _VHH peptide monomers have binding specificity to epitopes of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). This binder is referred to as “bispecific” because it recognizes two different toxin epitopes. Because it carries four _VHH peptide monomers (two copies of the first monomer and two copies of the second monomer), it is referred to as a “tetramer.”

In some aspects of this sub-example, the first and second _VHH peptide monomers have binding specificity to the same or different epitopes.

In some aspects of this sub-example, the _VHH peptide monomer independently exhibits binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

In a specific aspect of this sub-example, the binder comprises the amino acid sequence shown in SEQ ID NO: 32 (AH3/5D-Fc) or a sequence variant having at least 95% sequence identity with it, wherein the sequence variant retains TcdA and/or TcdB binding specificity, or retains toxin neutralizing activity, or both. Since this binder is based on the Fc domain, it will be clear to those skilled in the art that two identical polypeptides having the stated amino acid sequence act as arms of the binder, and the arms will assemble via disulfide bonds to provide the complete binder. The variant amino acids of the sequence variant may be located within the frame region of the _VHH peptide monomer.

In another specific aspect of this sub-example, the binder comprises the amino acid sequence shown in SEQ ID NO: 34 (AA6/E3-Fc) or a sequence variant having at least 95% sequence identity with it, wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity, or both. Since this binder is based on the Fc domain, it will be clear to those skilled in the art that two identical polypeptides having the stated amino acid sequence act as arms of the binder, and the arms will assemble via disulfide bonds to provide the complete binder. The variant amino acid of the sequence variant may be located within the frame region of the _VHH peptide monomer.

In some aspects of this embodiment and its sub-embodiments, the binder specifically binds to TcdA and/or TcdB. In some aspects of this embodiment, the binder exhibits TcdA and/or TcdB neutralizing activity.

This invention includes humanized variants of each binder provided in the various embodiments and aspects defined herein. Similarly, this invention includes epitope-binding fragments of each binder provided in the various embodiments and aspects defined herein.

Polynucleotides, expression vectors and host cells

This invention includes polynucleotides comprising a nucleotide sequence encoding each of the various embodiments and aspects provided herein, and their complementary strands. The invention also includes expression vectors (e.g., bacteria and yeast) comprising said polynucleotides and host cells (e.g., bacteria, yeast, mammals, insects) comprising said expression vectors. The invention further includes a method for producing the binding agents defined herein, comprising culturing host cells under conditions that promote expression of the binding agent encoded by the expression vector and recovering the binding agent from the cell culture.

Engineered strains of Saccharomyces boulardii

In a fourth embodiment, the present invention relates to *Saccharomyces* yeast strains, such as *Saccharomyces cerevisiae* and *Saccharomyces boulardii*, engineered to produce one or more binding agents as defined herein. In a preferred aspect, the engineered strain of *Saccharomyces* yeast secretes the binding agent.

The identity of a yeast strain is limited to its ability to be engineered to produce and preferably secrete one or more of the binding agents of this invention. In a preferred aspect of the invention, the yeast strain engineered to produce one or more binding agents is *Saccharomyces cerevisiae* or *Bacillus blakeana*. Therefore, the invention includes engineered strains of *Saccharomyces cerevisiae* that produce one or more binding agents as defined herein, and engineered strains of *Saccharomyces cerevisiae* that secrete one or more binding agents as defined herein. The invention also includes engineered strains of *Bacillus blakeana* that produce one or more binding agents as defined herein, and engineered strains of *Bacillus blakeana* that secrete one or more binding agents as defined herein.

In this example embodiment, the present invention relates to an engineered strain of *Saccharomyces cerevisiae* that produces a binding agent, wherein the binding agent is a _VHH peptide monomer or a linker group comprising 2, 3, 4 or more monomers of the _VHH peptide monomer, each monomer preferably specifically binding to TcdA and/or TcdB. Therefore, the present invention includes an engineered strain of *Saccharomyces cerevisiae* that produces a _VHH peptide binding agent comprising at least one _VHH peptide monomer, wherein each VHH peptide monomer has binding specificity to an epitope of *Clostridium difficile* toxin _A (TcdA) or toxin B (TcdB). In some aspects, these binding agents comprise 2, 3 _{, 4 or more linked VHH peptide monomers. VHH peptide monomers} include, but are not limited to, _VHH peptide monomers 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7).

In another example of this embodiment, the invention relates to an engineered strain of *Saccharomyces cerevisiae* that produces a binding agent comprising a _VHH peptide monomer linked to an IgG antibody, wherein the binding agent binds to TcdA and/or TcdB as defined herein. In these IgG-based binding agents, the variable regions of the light and heavy chains of the IgG antibody are replaced by one, two, three, four, or more _VHH peptide monomers.

In another example of this embodiment, the invention relates to an engineered strain of *Saccharomyces cerevisiae* that produces a binding agent comprising a _VHH peptide monomer linked to an antibody Fc domain, wherein the binding agent binds TcdA and/or TcdB as defined herein. In these Fc domain-based binding agents, one, two, three, four, or more _VHH peptide monomers are linked to the hinge region, _CH2 , and _CH3 regions of each arm of the Fc domain of the antibody heavy chain. Thus, the peptide monomers replace the Fab region of the antibody.

In another example of this embodiment, the present invention relates to an engineered strain of *Saccharomyces cerevisiae* that produces a four-specific tetrameric binder, wherein the binder comprises linked first, second, third, and fourth _VHH peptide monomers, and wherein the _VHH peptide monomers independently have binding specificity to epitopes of *Clostridium difficile* toxin A (TcdA) or toxin B (TcdB). In some aspects, the first, second, third, and fourth _VHH peptide monomers each have binding specificity to different epitopes. In some aspects, two _VHH peptide monomers have binding specificity to the epitope of TcdA, and two _VHH peptide monomers have binding specificity to the epitope of TcdB. In some aspects, the _VHH peptide monomers independently have binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

In a preferred example of this embodiment, the invention relates to an engineered yeast strain, wherein the binder is ABAB, wherein the first and third monomers are specific for binding to the TcdA epitope and the first and third monomers are _VHH peptide monomers AH3 (SEQ ID NO: 7) and AA6 (SEQ ID NO: 5), respectively, and wherein the second and fourth monomers are specific for binding to the TcdB epitope and the second and fourth monomers are _VHH peptide monomers 5D (SEQ ID NO: 1) and E3 (SEQ ID NO: 3), respectively. In some aspects, the ABAB binder comprises the amino acid sequence shown in SEQ ID NO: 19 or a sequence variant having at least 95% sequence identity therewith, wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin-neutralizing activity or both. In some aspects, the ABAB binder further comprises an N-terminal secretion signal selected from AT secretion signals (MRFPSIFTAVLFAASSALA (SEQ ID NO: 99)) and IVS secretion signals (MLLQAFLFLLAGFAAKISA (SEQ ID NO: 103)).

In some respects, the ABAB binder is expressed by a plasmid within yeast, wherein the ABAB binder comprises the amino acid sequence shown in SEQ ID NO: 107 or a sequence variant having at least 95% sequence identity with it, and wherein said sequence variant retains TcdA and/or TcdB binding specificity, or said sequence variant retains toxin-neutralizing activity or both. The plasmid may be, but is not limited to, pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88).

In some respects, an ABAB binding agent coding sequence is integrated into the chromosome of a yeast strain, wherein the ABAB binding agent comprises the amino acid sequence shown in SEQ ID NO: 109 or a sequence variant having at least 95% sequence identity with it, and wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity or both.

This embodiment includes engineered strains of *Saccharomyces* yeast that produce therapeutic proteins having binding specificity to a unique epitope of *Clostridium difficile* toxin A (TcdA) or toxin B (TcdB), or both. Preferably, the engineered strain of *Saccharomyces* yeast is *Saccharomyces cerevisiae* or *Bulachaeta*. A therapeutic protein is any protein that can bring about improvement or cure in a subject's medical condition, or that can inhibit or prevent the development of a medical condition in a subject. Suitable therapeutic proteins include, but are not limited to, proteins capable of (a) replacing defective or abnormal proteins; (b) enhancing existing pathways; (c) providing novel functions or activities; (d) interfering with molecules or organisms; and (e) delivering other compounds or proteins, such as radionuclides, cytotoxic drugs, or effector proteins. Therapeutic proteins also include antibodies and antibody-based drugs, Fc fusion proteins, anticoagulants, blood factors, bone morphogenetic proteins, engineered protein scaffolds, enzymes, growth factors, hormones, interferons, interleukins, and thrombolytic agents. Therapeutic proteins further include bispecific monoclonal antibodies (mAbs) and multispecific fusion proteins, mAbs conjugated to small molecule drugs, and proteins with optimized pharmacokinetic properties.

Methods for preparing engineered strains of Saccharomyces boulardii

The present invention also relates to a method for preparing yeast strains engineered to produce one or more binding agents as defined herein.

Therefore, the present invention includes a method for preparing yeast strains engineered to produce one or more binding agents as defined herein, comprising (a) transforming the yeast strains with an expression vector encoding the binding agent, and (b) screening the yeast strains of (a) to produce the binding agent. In one aspect, the expression vector is the plasmid pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88).

Therefore, the present invention includes a method for preparing engineered yeast strains that produce one or more binding agents as defined herein, comprising (a) integrating a polynucleotide sequence encoding the binding agent into the genome of a yeast strain, and (b) screening the yeast strains of (a) to produce the binding agent. In some aspects, chromosome integration is performed by:

(a) An integration cassette was generated by amplifying a polynucleotide sequence encoding an ABAB binding agent from plasmid pCEV-G4-Km-TEF-AT-yABAB hAA6T83N-tagless (SEQ ID NO: 90) using primers containing (i) nucleic acid sequences homologous to a selected yeast chromosome integration site and (ii) nucleic acid sequences homologous to the 5' and 3' regions of the ABAB binding agent encoding sequence of the plasmid.

(b) Under conditions that promote spontaneous integration of the integration cassette into the double-strand break site, the integration cassette generated in (a) was transformed into yeast with pCRI-Sb-δ1 (SEQ ID NO: 91) or pCRI-Sb-δ2 (SEQ ID NO: 92) to induce double-strand breaks in the corresponding yeast chromosome δ sites.

(c) Screening the transforming yeast from (b) to produce ABAB binding agents.

In some aspects of these methods, the yeast strains engineered to produce the binders are auxotrophic strains of the yeast genus, such as the ura3- strain of yeast. The ura3- strain of yeast can be used under the ura3 selection.

In some aspects of these methods, the yeast strains engineered to produce the binder are Saccharomyces cerevisiae or Saccharomyces blakeana.

In some aspects of these methods, screening is performed using immunoassays such as ELISA.

pharmaceutical preparations

This invention includes pharmaceutical formulations comprising one or more binders as defined herein and pharmaceutically acceptable carriers or diluents. This invention also includes pharmaceutical formulations comprising one or more engineered strains of yeasts as defined herein and pharmaceutically acceptable carriers or diluents. In some aspects, the yeasts are *Saccharomyces cerevisiae* or *Bacillus brasiliensis*.

Treatment and prevention methods

In a sixth embodiment, the present invention relates to a method for treating or preventing disease symptoms induced by Clostridium difficile in a subject, comprising administering to a subject suffering from or developing Clostridium difficile infection a therapeutically effective amount of one or more binders as defined herein and/or one or more engineered strains of yeast. In a preferred aspect, the yeast is Saccharomyces cerevisiae or Saccharomyces blakeana.

In some aspects of this embodiment, the disease symptom induced by Clostridium difficile is diarrhea.

In a seventh embodiment, the present invention relates to a method for neutralizing Clostridium difficile toxins TcdA and/or TcdB in a subject infected with Clostridium difficile, comprising administering to the subject suffering from Clostridium difficile infection a therapeutically effective amount of one or more binders as defined herein and/or one or more engineered strains of yeast. In a preferred aspect, the yeast is Saccharomyces cerevisiae or Saccharomyces boulardii.

In an eighth embodiment, the present invention relates to a method for treating or preventing Clostridium difficile infection in a subject, comprising administering to a subject suffering from or developing Clostridium difficile infection a therapeutically effective amount of one or more binders as defined herein and/or one or more engineered strains of yeast. In a preferred aspect, the yeast is Saccharomyces cerevisiae or Saccharomyces boulardii. In some aspects of the eighth embodiment, the method further comprises administering to the subject a therapeutically effective amount of antibiotic.

In a ninth embodiment, the present invention relates to a method for maintaining normal intestinal function in a subject suffering from Clostridium difficile infection, comprising administering to a subject suffering from or developing Clostridium difficile infection a therapeutically effective amount of one or more binders as defined herein and/or one or more engineered strains of yeast. In a preferred aspect, the yeast is Saccharomyces cerevisiae or Saccharomyces boulardii. In some aspects of the ninth embodiment, the method further comprises administering to the subject a therapeutically effective amount of antibiotics.

In some aspects of this method, the binder is in a pharmaceutical formulation comprising the binder and a pharmaceutically acceptable carrier or diluent.

In some aspects of this method, the therapeutically effective amount of the binder is between 10 μg/kg and 100 mg/kg of the agent per subject's body weight.

In some aspects of this method, the drug is administered to the subject orally, parenterally, or rectally.

In some aspects of this method, engineered strains of yeast are contained in a pharmaceutical preparation comprising the engineered strain and a pharmaceutically acceptable carrier or diluent. In a preferred aspect, the yeast is *Saccharomyces cerevisiae* or *Bacillus spp.*

In some aspects of this method, the therapeutically effective dose of the engineered yeast strain is from 10 μg/kg to 100 mg/kg of engineered strain per subject's body weight. In a preferred aspect, the yeast strain is *Saccharomyces cerevisiae* or *Saccharomyces blakeana*.

In some aspects of this method, an engineered strain of yeast is administered to the subject orally, nasally, or rectally. In a preferred aspect, the yeast is *Saccharomyces cerevisiae* or *Saccharomyces boulardii*.

The features and technical advantages of the invention have been outlined quite extensively above to facilitate a better understanding of the following detailed description. Additional features and advantages of the invention will be described herein, forming the subject matter of the claims. Those skilled in the art will understand that any concepts and specific embodiments disclosed herein can readily be used as the basis for modifications or the design of other structures for achieving the same purpose as the invention. Those skilled in the art will also recognize that such equivalent constructions do not depart from the spirit and scope of the invention as set forth in the appended claims. The novel features (whether in their organization or manner of operation) and other objects and advantages considered to be characteristic of the invention will be better understood when the following description is considered in conjunction with the accompanying drawings. However, it is to be clearly understood that any descriptions, drawings, examples, etc., provided are for illustrative and descriptive purposes only and are not intended to limit the scope of the invention.

Attached Figure Description

Figure 1. Explanation of the strategy used to prepare the binder of the present invention.

Figure 2. Illustration of Clostridium difficile toxins TcdA and TcdB, showing the glucosyltransferase domain (GT), cysteine protease domain (CPD), translocation domain (TD), and receptor-binding domain (RBD) of each toxin. _VHHs that recognize and bind to different toxin domains are shown. Underlined _VHHs exhibit toxin-neutralizing activity.

Figures 3A-3F. Monomeric or dimeric _VHH exhibits potent neutralizing activity. At nM concentrations, _VHH blocks cell rounding induced by TcdA (Figure 3A) or TcdB (Figure 3B). (Figure 3C) Illustration of the two heterodimers targeting TcdA or TcdB. The His ₍₆₎ tag at the N-terminus facilitates purification; a flexible spacer region (FS) separates the two _VHHs . (Figure 3D) The 5D/E3 dimer shows at least a 10-fold increase in neutralizing activity compared to a simple mixture of the two _VHHs . The heterodimer provides complete protection to mice from lethal IP attacks by TcdB (Figure 3E) or TcdA (Figure 3F).

Figure 4. Illustration of ABAB. His-tag and E-tag are epitope tags used for purification and detection, respectively. FS: flexible linker; ABP: albumin-binding peptide.

Figures 5A-5B. ABAB is highly effective in protecting mice from Clostridium difficile spores (Figure 5A) and toxins (Figure 5B). MK HuMabs: A mixture of Merck anti-TcdA (actoxumab) and anti-TcdB (bezlotoxumab) human monoclonal antibodies currently undergoing clinical trials.

Figures 6A-6B. Antitoxin sera against these two toxins protect mice from CDI. Mice were intraperitoneally injected with 50 μl of alpaca antiserum against TcdA (“anti-A”), TcdB (“anti-B”), or TcdA+TcdB (“anti-A+anti-B”) or 100 μl of presera or PBS (“CTR”) for ⁴ hours prior to inoculation with Clostridium difficile spores (UK1 strain, 10⁶ spores/mouse). Examples of mouse survival (Figure 6A; anti-A+anti-B vs. PBS, p = 0.006) and weight loss (Figure 6B) are illustrated (*, p < 0.05 between anti-A+anti-B and control groups).

Figure 7. Illustration of ABAB and ABAB-IgG molecules.

Figures 8A-8B. ELISA analysis of the binding of ABAB-IgG to TcdA (Figure 8A) and TcdB (Figure 8B) compared to the binding of each V _H H to each toxin.

Figures 9A-9B. Sandwich ELISA analysis combining four specific antibodies, IgG-ABAB and TcdA and TcdB. Figure 9A shows the serially diluted ABAB-IgG added to an ELISA plate coated with TcdA (TxA) followed by TcdB (TxB). Figure 9B shows the serially diluted ABAB-IgG added to an ELISA plate coated with TcdB (TxB) followed by TcdA (TxA).

Figures 10A-10B. Neutralizing activity of ABAB-IgG against TcdA (Figure 10A) and TcdB (Figure 10B).

Figure 11. This figure shows a comparison of the in vivo neutralizing activity of ABAB-IgG against Clostridium difficile infection and Merck antibodies against TcdA and TcdB (actoxumab and bezlotoxumab) in mice.

Figure 12. Study design for the effect of preventive ABAB-IgG on Clostridium difficile infection.

Figures 13A-13B. Bispecific sandwich ELISA. (Figure 13A) Illustration of the toxin and antibody established in the ELISA. (Figure 13B) O.D. readings for different TcdA concentrations; 125 ng/ml of TcdA was selected for subsequence ELISA.

Figures 14A-14B. Activity of ABAB secreted by Sc-ABAB. (Figure 14A) Neutralization of secreted ABAB in Saccharomyces cerevisiae culture supernatant. Sc: Saccharomyces cerevisiae (BY4741); Sc-ABAB: Saccharomyces cerevisiae (BY4741)-pD1214-FAKS-ABAB; r-ABAB: recombinant ABAB. ABAB in the ScABAB supernatant provides adequate protection of cells from poisoning. ELISA O.D. readings from the supernatant of a single Sc-ABAB clone (Figure 14B).

Figures 15A-15B. ABAB secretion levels with various secretion signals. (Figure 15A) ABAB secretion was measured by ELISA and based on O.D. 600 normalized cell density in *Saccharomyces cerevisiae*. Statistical significance was determined by Kruskal-Wallis test followed by Dunn's multiple comparison test. *p<0.05**p<0.01 (Figure 15B) ABAB secretion was measured by ELISA and based on O.D. 600 normalized cell density in *Saccharomyces blakeana*. Statistical significance was determined by Mann-Whitney test. ****p<0.0001.

Figure 16. Illustration of homologous recombination targeting of chromosome-coding genes in *Bacillus blazei*.

Figures 17A-17D. *Saccharomyces boulardii* URA3Δ/Δ expressing ABAB. (Figure 17A) Comparison of YPD growth with and without vancomycin (1 mg/ml). (Figure 17B) Stability of ABAB in *Saccharomyces boulardii* culture supernatant determined by ELISA after 2 hours of culture. (Figure 17C) Neutralizing activity of ABAB from *Saccharomyces boulardii* URA3Δ/Δ culture supernatant expressing ABAB. (Figure 17D) Detection of ABAB in *Saccharomyces boulardii* URA3Δ/Δ culture supernatant by Western blotting. Enrichment: ABAB contains a c-Myc tag at the C-terminus and was further concentrated using an α-c-Myc tag antibody.

Figures 18A-18C. Protective effect of *Saccharomyces boulardii* expressing ABAB in the prevention of CDI in mice. (Figure 18A) Survival rate, (Figure 18B) Weight loss, (Figure 18C) Recorded and presented diarrheal events throughout the infection process. Significance was determined by Fisher's exact test with two tails and 95% confidence intervals*; for Figure 18A, p-value was 0.0108, and for Figures 18B and 18C, conventional two-way ANOVA (without repeated measures) was used, followed by Dunnett's multiple comparison test, *P ≤ 0.05. "Sb:EP" is *Saccharomyces boulardii* with an empty plasmid; "Sb:ABAB" is *Saccharomyces boulardii* expressing ABAB.

Figures 19A-19C. Protection of *Saccharomyces boulardii* expressing ABAB in CDI-treated mice. (Figure 19A) Survival rate, (Figure 19B) Weight loss, (Figure 19C) Recorded and presented diarrheal events throughout the infection process. Significance was determined by Fisher's exact test with two tails and 95% confidence intervals*; for Figure 19A, the p-value was 0.0256; for Figures 19B and 19C, conventional two-way ANOVA (without repeated measures) was used, followed by Dunnett's multiple comparison test. For Figures 19B and 19C, **P ≤ 0.01 ****P ≤ 0.0001. "Sb:EP" is *Saccharomyces boulardii* with an empty plasmid; "Sb:ABAB" is *Saccharomyces boulardii* expressing ABAB.

Figures 20A-20C. Protective effect of Saccharomyces boulardii expressing ABAB in CDI relapse mice. (Figure 20A) Survival rate, (Figure 20B) Weight loss, (Figure 20C) Recorded and presented diarrheal events throughout the infection process. Significance was determined by Fisher's exact test with two tails and 95% confidence intervals*; for Figure 20A, the p-value was 0.017; for Figures 20B and 20C, conventional two-way ANOVA (without repeated measures) was used, followed by Dunnett's multiple comparison test. For Figures 20B and 20C, *P≤0.05**P≤0.01****P≤0.0001. "Sb:EP" is Saccharomyces boulardii with an empty plasmid; "Sb:ABAB" is Saccharomyces boulardii expressing ABAB.

Figure 21. Illustration of δ-site targeted chromosome integration using CRISPR. δ-site sequences were obtained by blasting the draft genome of MYA796 using Ty1-H3 (Genbank accession number M18706). Compiled sequences were used to identify common prototypical spacer adjacent motif (PAM) sites and the prototypical spacer. Two PAM site sequences were selected based on optimal coverage of multiple sites upstream and downstream of the prototypical spacer and PAM sites and common homologous sequences for easy integration of the ABAB expression cassette. PAM site “I” is provided in SEQ ID NO: 93; PAM site “II” is provided in SEQ ID NO: 94. Homologous recombination sequences used in the primers used to generate the ABAB expression cassette by PCR are underlined.

Figures 22A-22B. ABAB secretion in *Saccharomyces boulardii* using CRISPR-based δ-site chromosome integration. (Figure 22A) ABAB secretion was measured by ELISA. ITG: ABAB integration cassette. Low: CRISPR plasmid to ITG ratio of 2; High: CRISPR plasmid to ITG ratio of 0.25. (Figure 22B) Comparison of ABAB secretion levels. M-/- ^Cir0 : pKC, M-/- ^Cir+ : ABAB, M-/- ^Cir0 : ABAB is plasmid-based. ^ChIns : Chromosomal integration of a single site targeting the ABAB cassette via conventional homologous recombination. ^CRISPR 1-2: ABAB cassette integration at site I. ^CRISPR 3-4: ABAB cassette integration at site II.

Figures 23A-23C. Protective effect of *Bula spp.* expressing ABAB in CDI mice. (Figure 23A) Survival rate, (Figure 23B) Weight loss, (Figure 23C) Recorded and presented diarrheal events throughout the infection process. Significance was determined by Fisher's exact test with two tails and 95% confidence intervals*; for Figure 23A, the p-value was 0.0325; for Figures 23B and 23C, conventional two-way ANOVA (without repeated measures) was used, followed by Dunnett's multiple comparison test. For Figures 23B and 23C, *P≤0.05**P≤0.01.

Detailed Implementation

I. Definition

Unless otherwise stated, technical terms are used according to their usual usage. Definitions of commonly used terms in molecular biology can be found, for example, Benjamin Lewin, Genes VII (ISBN 019879276X), Oxford University Press, 2000; Kendrew et al. (eds); The Encyclopedia of Molecular Biology (ISBN 0632021829), Blackwell Publishers, 1994; and Robert A. Meyers (eds), Molecular Biology and Biotechnology: a Comprehensive Desk Reference (ISBN 0471186341), Wiley, John & Sons, Inc., 1995; and other similar technical references.

As used herein, “a” or “one” can mean one or more. As used herein, when used with the word “including”, the word “a” or “one” can mean one or more. As used herein, “another” can mean at least a second or more. Furthermore, unless the context requires otherwise, singular terms include plurals, and plural terms include singulars.

As used herein, “about” refers to a numerical value, including, for example, integers, fractions, and percentages, whether explicitly stated or not. The term “about” generally refers to a range of numerical values (e.g., +/- 5% to 10% of the value) that a person skilled in the art would consider equivalent to the value (e.g., having the same function or result). In some cases, the term “about” may include a numerical value rounded to the nearest significant figure.

II. This invention

Clostridium difficile-associated disease (CDI) is primarily caused by two large exotoxins: toxin A (TcdA) and toxin B (TcdB), both produced by the bacteria. These toxins are structurally similar large single-chain proteins (TcdA is approximately 300 kDa; TcdB is approximately 270 kDa) that exhibit similar modes of action on host cells. Both toxins target the host Rho GTPase, leading to enzyme inactivation, followed by cytoskeleton disintegration and apoptosis. In intestinal epithelial cells, TcdA catalyzes the glucosylation of Rho GTPase, resulting in the reorganization of the actin cytoskeleton, accompanied by morphological changes such as complete cell rounding and disruption of intestinal barrier function. The toxins can induce CDI independently in animals, and TcdA-TcdB strains are non-toxic.

Numerous independent studies have demonstrated that antibodies neutralizing antitoxins confer protection against CDI [24-33]. Since TcdA and TcdB are fundamental virulence factors in Clostridium difficile, neutralizing antibodies against these two toxins prevent infection with toxin-producing Clostridium difficile in animal models [30-33]. In humans, high serum antitoxin antibody levels are associated with reduced disease severity and relapse rates [9,25,29].

Therefore, there are prophylactic reasons for systemic and oral administration of antitoxin antibodies. However, monoclonal antibodies targeting a single epitope are often of low affinity, and the use of such antibodies carries the risk of inducing mutations within the toxin epitope, resulting in additional toxin strains. Therefore, neutralizing antitoxins targeting multiple, key, and conserved toxin epitopes is highly desirable.

This invention is based on existing knowledge regarding anti-TcdA and anti-TcdB antibodies for the treatment and prevention of CDI and CDI symptoms. This document provides antibody-based fusion protein conjugates derived from human and camelid immunoglobulins, optionally expressed by a probiotic yeast strain in the subject. These conjugates recognize and specifically bind to Clostridium difficile TcdA and/or TcdB. Some of these conjugates exhibit toxin-neutralizing activity. These yeast-based immunotherapeutic agents can be used to treat or prevent primary and recurrent CDI, as well as symptoms of primary and recurrent CDI. In a preferred aspect, the yeast strain is Saccharomyces cerevisiae or Saccharomyces boulardii.

As discussed in detail below, camelids (dromedary camels, Bactrian camels, wild Bactrian camels, llamas, alpacas, llamas, and maroon llamas) produce a class of functional immunoglobulins lacking light chains, and are therefore heavy-chain-only antibodies (HCAbs)[34] with binding properties comparable to those achieved by conventional IgG[35]. The _VH domain of HCAbs is called _VHH , which is similar to the conventional human _VH domain but has unique sequence and structural features[36]. DNA encoding this domain can be readily cloned and expressed in microorganisms to produce soluble protein monomers that retain the antigen-binding properties of the parental HCAb. These _VHH peptide monomer binders are small (~15 kDa), easy to produce, and generally more stable than conventional antibody fragments[37-39]. _VHHs are being explored for the treatment of intestinal diseases because they are relatively resistant to proteases and can be further engineered to enhance these properties[40]. They can also be produced as fusion proteins with human antibodies such as IgG and fragments of human antibodies such as Fc domains.

This invention utilizes the advantageous characteristics of HCAbs in preparing conjugates for the treatment and prevention of CDI. As disclosed herein, _VHH peptide monomers are screened for TcdA and TcdB epitope recognition and binding. Those monomers exhibiting epitope binding and toxin-neutralizing activity are linked to produce the conjugates of this invention. Conjugates include simple _VHH peptide monomers and linker groups of _VHH peptide monomers (including 2, 3, 4 or more monomers) as well as more complex conjugates comprising _VHH peptide monomers linked to the Fc domain of antibodies and _VHH peptide monomers linked to IgG antibodies (see Figure 1).

In addition, Saccharomyces boulardii, a generally GRAS (Generally Recognized As Safe) biological agent by the FDA, is generally available over-the-counter for promoting gut health and improving gastrointestinal disorders caused by diarrheal diseases. This yeast strain has been studied in several randomized, double-blind, placebo-controlled clinical trials to determine its safety and efficacy against intestinal diseases, including CDI [42-46]. Treatment with Saccharomyces boulardii significantly reduced CDI relapses [44-46], with significantly less Saccharomyces boulardii in the stool of patients with relapses than in those without [43]. Immunomodulatory effects of Saccharomyces boulardii have been described to provide protection against Clostridium difficile toxin-induced inflammation [47-49]. Furthermore, Saccharomyces boulardii may help maintain a normal microbiome [50]; a recent clinical trial (NCT01473368) found that treatment with Saccharomyces boulardii prevented some antibiotic-induced microbiome changes while reducing antibiotic-associated diarrhea.

Saccharomyces cerevisiae (commonly known as "brewer's yeast"), which is genetically related to Saccharomyces bulacanthus, has been successfully used to express _VHH in high yields[51]. Saccharomyces bulacanthus is physiologically different from Saccharomyces cerevisiae, although genomic analysis shows that the two genomes are very similar at the nucleotide level[52,53]. Therefore, molecular genetic tools previously developed for Saccharomyces cerevisiae are now being used with Saccharomyces bulacanthus[54-56], making this probiotic a candidate for CDI therapeutics.

Several additional metabolic characteristics make *Saccharomyces boulardii* an ideal candidate for oral therapeutic use. Unlike *Saccharomyces cerevisiae*, *Saccharomyces boulardii* grows well at 37°C and is more resistant to acidic conditions[57], making the strain particularly well-suited for better survival and persistence in the human gut after oral administration. Furthermore, an experimental rodent oral colonization model with yeasts has been well characterized[58]; using this model, protection against oral challenges with intestinal pathogens such as *Salmonella typhimurium*[58,59] and *Salmonella enteritidis*[60] has been reported in conventionally treated mice with *Saccharomyces boulardii*, as well as protection against CDI challenges in pretreated germ-free animals[58,61]. Genetically engineered *Saccharomyces boulardii* to secrete _VHH binding agents capable of neutralizing both *Clostridium difficile* TcdA and TcdB can significantly enhance the probiotic's ability to treat ongoing and recurrent CDI.

Given the unique characteristics of *Saccharomyces boulardii*, *Saccharomyces boulardii* strains express the binding agents (which are produced and tested) as defined herein. As illustrated in the examples, these yeast-based immunotherapeutic agents can be used to treat or prevent primary and recurrent CDI, as well as the symptoms of primary and recurrent CDI.

_VHH monomer & _VHH heterodimer

As originally reported in WO 16/127104, the inventors have established an efficient platform for screening _VHH monomers targeting specific domains of Clostridium difficile toxins. Using highly immunogenic, non-toxic whole toxins for immunization and bioactive chimeric toxins (with normal functional domains) for screening, groups of _VHH monomers binding to different domains of TcdA or TcdB were prepared. Most of these _VHH monomers exhibited potent neutralizing activity, and their binding to specific domains of TcdA and TcdB was determined (Figure 2).

Several _VHH monomers bind highly conserved TcdA/TcdB epitopes. For example, the E3 _VHH monomer binds to the Rho GTPase binding site and blocks glucosylation; the AH3 _VHH monomer binds to the GT domain of the toxin; and the 7F _VHH monomer binds to the cysteine protease cleavage site and blocks GT domain cleavage and release. Some _VHH monomers exhibit potent toxin-neutralizing activity, capable of blocking toxin cytotoxic activity at nM concentrations (underlined monomers in Figure 2; see also Figures 3A and 3B). Table 1 lists the amino acid and nucleic acid sequences of some of these _VHH peptide monomers, including wild-type and codon-optimized versions, referencing the sequence listing. While both optimized and unoptimized versions can be used to produce various binding agents of the present invention, codon-optimized versions are preferred for expression in mammalian cells.

This invention includes each _VHH peptide monomer and its sequence variants mentioned in Table 1, having at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity over the entire length of the peptide sequence and retaining the toxicant-binding and/or neutralizing activity of the wild-type peptide. This invention also includes the polynucleotide sequence encoding each _VHH peptide monomer and its sequence variants in Table 1, as well as its complementary strand.

Table 1

To enhance the binding activity of peptide monomers, _VHH peptide homodimer and heterodimer binders were developed, in which two _VHH peptide monomers are linked (Figure 3C). Homodimer binders consist of two identical monomers that bind to the same epitope on two different toxins. Heterodimer binders contain two different monomers that bind to two different epitopes on the same toxin or different epitopes on two different toxins. Compared to an equimolar mixture of single _VHH peptide monomers including heterodimers, _VHH heterodimers were found to have significantly enhanced neutralizing activity (Figure 3D). In fact, heterodimers 5D/E3 and AH3/AA6 were found to completely protect mice from lethal systemic TcdB or TcdA attacks, respectively, while mixtures of 5D and E3 or AA6 alone provided only partial protection (Figures 3E and 3F).

Short, flexible linkers of 10 to 20 amino acids are used to link _VHH monomers in homodimers and heterodimers. Suitable linkers include those provided in Table 2. Table 2 also includes codon-optimized versions of three linkers. While both optimized and unoptimized versions can be used to produce various binding agents of the present invention, codon-optimized versions are preferred for expression in mammalian cells.

Table 2

name Optimized codons? SEQ ID NO of amino acid sequence SEQ ID NO of nucleic acid sequence Connector-1 yes 9 10 Connector-2 yes 11 12 Connector-3 yes 13 14 Connector-1 no 56 57 Connector-2 no 58 59 Connector-3 no 60 61

Those skilled in the art will understand that minute alterations can be made to the sequence of flexible linkers without departing from the properties of the peptide. Therefore, flexible linker sequence variants based on which they are based can be used, exhibiting at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity and preserving properties across the entire length of the peptide sequence.

This invention includes _VHH peptide homodimer binders comprising any monomer pairs listed in Table 1 linked by flexible linkers as defined above. This invention also includes _VHH peptide heterodimer binders comprising any combination of two monomers listed in Table 1 linked by flexible linkers as defined above. Exemplary heterodimers are provided in Table 3.

Table 3

name SEQ ID NO of amino acid sequence SEQ ID NO of nucleic acid sequence AH3-5D 15 16 AA6-E3 17 18 5D-E3 62 63 AH3-AA6 64 65

The invention also includes sequence variants of VHH peptide homodimers and heterodimers having at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity across the entire length of the protein sequence and retaining the toxicant-binding and/or neutralizing activity of the wild-type protein. The invention also includes polynucleotide sequences encoding each _VHH peptide homodimer and heterodimer, its sequence variants, and _their complementary strands.

This invention also includes _VHH peptide isotrimeric and heterotrimeric binders, wherein the three monomers are linked using flexible linkers as defined in Table 2 above. Any combination of monomers from Table 1 can be used, including trimers containing three copies of the same monomer, trimers containing two copies of one monomer and three copies of another monomer, and trimers containing three different monomers. This invention includes sequence variants of _VHH peptide isotrimeric and heterotrimeric molecules that have at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity over the entire length of the protein sequence and retain the toxin-binding and/or neutralizing activity of the wild-type protein. This invention further includes a polynucleotide sequence encoding each _VHH peptide isotrimeric and heterotrimeric molecule and its sequence variant, as well as its complementary strand.

ABAB

The success of the peptide monomers and heterodimers allowed the inventors to develop a conjugate containing four linked _VHH peptide monomers. This was a goal of the research, as early studies suggested that the most useful agents for treating and preventing CDI would be single antibodies that could simultaneously neutralize both TcdA and TcdB, as this is essential for delivering complete protection against most pathogenic Clostridium difficile strains. By generating a four-specificity conjugate that recognizes and binds to two epitopes on each toxin, the binding and neutralizing activity of the protein can be enhanced. Thus, a four-domain (four-specificity) _VHH conjugate was developed.

Tetraspecific tetramer binders can be prepared from any combination of monomers in Table 1, wherein the monomers are linked using flexible linkers in Table 2. These binders include those having four copies of the same monomer, those having three copies of the same monomer, those having two copies of the same monomer, those having four unique monomers, and variants thereof. The invention includes sequence variants of the tetramer that have at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity over the entire length of the protein sequence and retain the toxin-binding and/or neutralizing activity of the wild-type protein. The invention further includes a polynucleotide sequence encoding each tetramer and its sequence variants, as well as its complementary strand.

ABBA is the specific binder of this invention, which comprises four linked V _H H monomers AH3-E3-E3-AA6. Thus, ABBA has two identical monomers (E3) and two additional different monomers (AH3 and AA6) (see Table 1).

ABAB is another specific binder of the present invention, comprising four linked _VHH monomers, each monomer having binding specificity for a different epitope of TcdA or TcdB. ABAB is thus a tetraspecific tetrameric binder composed of four distinct neutralizing _VHH monomers, two targeting TcdA and two targeting TcdB. This structural feature allows ABAB to simultaneously bind to two distinct neutralizing epitopes on each toxin. As described below, ABAB exhibits affinity/affinity and neutralizing activity that are more than three orders of magnitude higher than those of human monoclonal antibodies (HuMabs) currently undergoing clinical trials for the treatment of CDI.

ABAB binders were prepared by linking _VHH monomers AH3, 5D, AA6, and E3 (Table 1) using flexible linkers (Table 2). These binders target conserved, non-overlapping epitopes and exhibit excellent toxin-neutralizing activity. In the ABAB design (Figure 4), since AH3 and AA6 bind to GT and TD respectively (Figure 2), the _VHH peptide monomers AH3 and AA6 are separated by 5D placed between them (Figure 2), resulting in them being spatially far apart. This design allows AH3 and AA6 to bind TcdA simultaneously.

The complete amino acid sequence of ABAB is provided in SEQ ID NO: 19; the nucleic acid sequence encoding the protein is provided in SEQ ID NO: 20. Therefore, the present invention includes the ABAB binding agent provided in SEQ ID NO: 19, and sequence variants of the ABAB binding agent having at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity over the entire length of the protein sequence and retaining the toxin-binding and/or neutralizing activity of the wild-type protein. Sequence variants include those wherein the variant is a humanized variant and/or wherein the amino acids are optimized for use in yeast production and secretion.

The present invention further includes a polynucleotide sequence encoding an ABAB binding agent (e.g., SEQ ID NO: 20) and its sequence variants, as well as the complementary strand of the polynucleotide sequence.

Modified versions of the ABAB binders covered by this invention include those having one or more of the following modifications: (i) a His ₍₆₎ tag at the amino terminus of the protein (HHHHHH; SEQ ID NO: 66) to aid purification; (ii) an E tag at the carboxyl terminus of the protein (GAPVPYPDPLEPR; SEQ ID NO: 67) to aid detection; (iii) an albumin-binding peptide (ABP) (DICLPRWGCLWD; SEQ ID NO: 21) at the carboxyl terminus of the construct to increase the serum half-life of the protein, since the _VHH monomer has a half-life of 2-3 hours and the inclusion of ABP can increase the serum half-life to 10 hours (see Figure 4); and a D7 tag at the carboxyl terminus of the protein (SSAPTKAKRRVVQREKT; SEQ ID NO: 112). This invention includes versions of ABAB binders having 1, 2, 3, or 4 of these tags and peptides. An exemplary ABAB binder containing His and D7 tags comprises the amino acid sequence shown in SEQ ID NO: 113 (the coding sequence is shown in SEQ ID NO: 114).

When yeast strains are engineered to produce ABAB, proteins can also be modified to include a secretion signal at the amino terminus of the protein. The secretion signal can be, but is not limited to, one of the sequences shown in Table 4.

Table 4 - Secretion sequences of proteins in yeast

Exemplary modified ABAB binders that include an amino-terminal secretion signal include AT-ABAB and IVS-ABAB.

Exemplary modified ABAB binders expressed from plasmids in yeast or bacteria include the ABAB binder shown in SEQ ID NO: 107, which is encoded by the polynucleotide sequence shown in SEQ ID NO: 108.

Exemplary modified ABAB binding agents expressed in yeast after chromosome integration include the ABAB binding agent shown in SEQ ID NO: 109, which is encoded by the polynucleotide sequence shown in SEQ ID NO: 110.

Each binder of the present invention specifically binds to TcdA and/or TcdB. In certain aspects of the present invention, the binders exhibit TcdA and/or TcdB neutralizing activity.

For clarity, it may be noted that, as used herein, terms such as “monospecific,” “bispecific,” “trispecific,” and “tetraspecific” refer to specific binders that bind to different epitopes, such as 1, 2, 3, and 4, respectively. Similarly, terms such as “monomer,” “dimer,” “trimer,” and “tetramer” refer to specific binders that have individual _VHH peptide monomers (1, 2, 3, and 4) that bind to the epitopes. Therefore, a monospecific dimeric binder will exhibit two _VHH peptide monomers binding to the same epitope (e.g., a homodimer), and a bispecific dimeric binder will have two _VHH peptide monomers binding to two different epitopes (e.g., a heterodimer). A tetraspecific octamer binder has eight _VHH peptide monomers recognizing four different epitopes.

V _H H-FC

Chimeric Fc fusion proteins are known to have the potential to increase protein half-life in vivo. This strategy has been applied to several FDA-approved drugs, such as etanercept. Evidence from principle studies suggests that single-chain antibodies can be correctly assembled and expressed in B cells of transgenic mice carrying a mini-Ig construct encoding unicamal _vHH and the Fc domain of human IgG. EG2-Fc, a chimeric anti-EGFR/EGFRvIII _vHH , has shown excellent tumor accumulation in vivo and possesses pharmacokinetic properties that can improve glioblastoma targeting.

This invention includes a conjugate comprising a _VHH peptide monomer linked to an antibody Fc domain (VHH-Fc), wherein the conjugate binds TcdA and/or TcdB. In these Fc domain-based conjugates, one, two, three, four, or more _VHH peptide monomers are linked to the hinge region, _CH2 , and _CH3 region of the Fc domain of the antibody heavy chain. Thus, the peptide monomers replace the Fab region of the antibody.

_VHH peptide monomers can be any of those provided in Table 1 above, and include 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7) _VHH peptide monomers. When two or more monomers are linked, the monomers can be linked by a flexible peptide linker, typically containing 10-20 amino acids. Suitable linkers include those provided in Table 2, such as linker-1 (SEQ ID NO: 9), linker-2 (SEQ ID NO: 11), and linker-3 (SEQ ID NO: 13).

Although _VHH -Fc typically consists of two identical chains that self-assemble within the cell after generation, this invention also includes _VHH -Fc binders comprising two different Fc chains. In this case, the sequence of the individual _VHH monomer may differ between the two Fc chains, or the Fc chains themselves may differ in sequence, or both the _VHH monomer and the Fc chain may differ in sequence.

One type of _VHH -Fc conjugate is an octamer conjugate comprising an antibody Fc domain and first, second, third, and fourth _VHH peptide monomers, wherein the _VHH peptide monomers are specifically binding to epitopes of TcdA or toxin B TcdB. The first, second, third, and fourth _VHH peptide monomers are linked together and attached to the amino termini of two antibody Fc domains, and the antibody Fc domains contain hinge regions, _CH2 , and _CH3 regions of the antibody heavy chain. Because this conjugate has four _VHH peptide monomers, it can be monospecific (where all monomers bind to the same epitope), bispecific (where monomers bind to two different epitopes), trispecific (where monomers bind to three different epitopes), or tetraspecific (where monomers bind to four different epitopes).

A specific example of a tetraspecific _VHH -Fc binder is the ABAB-Fc binder, a tetraspecific octamer binder comprising an antibody Fc domain and two sets of linked first, second, third, and fourth _VHH peptide monomers. The antibody Fc domain comprises two arms, each including a hinge region of the antibody heavy chain, a _CH2 region, and a _CH3 region, and each arm has an amino terminus. For each arm of the Fc domain, a set of linked first, second, third, and fourth _VHH peptide monomers is attached to the amino terminus of the arm, and the _VHH peptide monomers have binding specificity for either TcdA or TcdB epitopes (see Figure 1). This binder is called "tetraspecific" because it recognizes four distinct toxin epitopes. Because it carries eight _VHH peptide monomers (two copies of the first monomer, two copies of the second monomer, two copies of the third monomer, and two copies of the fourth monomer), it is called an "octamer." ABAB-Fc has been found to exhibit specific binding and neutralizing activity.

The ABAB-Fc binder was prepared by generating an expression vector encoding _VHH peptide monomers AH3/5D/AA6/E3 (linked in the order shown) linked to the human IgG1 Fc domain. The _VHH peptide monomers were isolated via flexible linkers as shown in Table 2. The nucleic acid sequences encoding each strand are provided in SEQ ID NO: 23. The amino acid sequences of each strand are provided in SEQ ID NO: 22. Upon self-assembly into paired strands after expression, a tetraspecific octamer binder is generated. This invention includes the ABAB-Fc binder of SEQ ID NO: 22, modified versions of the ABAB binder as defined above, and sequence variants having at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity over the entire length of the protein sequence and retaining the toxin-binding and/or neutralizing activity of the wild-type protein. The present invention further includes multinucleotide sequences encoding these sequence variants and their complementary strands.

Monospecific _VHH -Fc binders (AH3-Fc, 5D-Fc, E3-Fc, AA6-Fc) and bispecific _VHH -Fc binders (e.g., AH3/5D-Fc and AA6/E3-Fc) are also prepared using this Fc fusion system. For monospecific binders, a single _VHH peptide monomer is linked to the human IgG1 Fc domain. During expression and assembly, the paired chains produce either a monospecific dimer binder (when the chains are the same) or a bispecific dimer binder (when the chains are different). For bispecific binders, two linked _VHH peptide monomers ( _VHH homodimers or heterodimers) are linked to the human IgG1 Fc domain. During expression and assembly, the paired chains produce either a bispecific tetramer binder (when the chains are the same) or a tetraspecific tetramer binder (when the chains are different). Table 5 provides the sequences of some of these binders.

Table 5

Specific pairings with one monomer include: 5D-Fc+5D-Fc; E3-Fc+E3-Fc; AA6-Fc+AA6-Fc; AH3-Fc+AH3-Fc; 5D-Fc+E3-Fc; 5D-Fc+AA6-Fc; 5D-Fc+AH3-Fc; E3-Fc+AA6-Fc; E3-Fc+AH3-Fc and AA6-Fc+AH3-Fc. Specific pairings with two monomers include: AH3-5D-Fc+AH3-5D-Fc; AA6-E3-Fc+AA6-E3-Fc and AH3-5D-Fc+AA6-E3-Fc.

A bispecific tetramer VHH-Fc conjugate is generated, comprising an antibody Fc domain and two sets of linked first and second _VHH peptide monomers. The antibody Fc domain comprises two arms, each including a hinge region, a _CH2 region, and a _CH3 region of the antibody heavy chain, and each arm has an amino terminus. For each arm of the Fc domain, a set of linked first and second _VHH peptide monomers is attached to the amino terminus of the arm, and the _VHH peptide monomers are specific for binding to epitopes of TcdA or TcdB. This conjugate is called "bispecific" because it recognizes two different toxin epitopes. It is called a "tetramer" because it carries four _VHH peptide monomers (two copies of the first monomer and two copies of the second monomer). The first and second _VHH peptide monomers can be specific for binding to the same or different epitopes. The _VHH peptide monomers can independently bind to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

Specific examples of bispecific tetramer _VHH -Fc binders include the amino acid sequence shown in SEQ ID NO: 32 (AH3/5D-Fc). The invention also includes sequence variants having at least 95% sequence identity, wherein the sequence variants retain toxin-neutralizing activity. The variant amino acids of the sequence variants may be located within the frame region of the _VHH peptide monomer.

Specific examples of bispecific tetramer _VHH -Fc binders include the amino acid sequence shown in SEQ ID NO: 34 (AA6/E3-Fc). The invention also includes sequence variants having at least 95% sequence identity, wherein the sequence variants retain toxin-neutralizing activity. The variant amino acids of the sequence variants may be located within the frame region of the _VHH peptide monomer.

The V _H H-Fc binder specifically binds to TcdA and/or TcdB. In certain aspects of the invention, the binder exhibits TcdA and/or TcdB neutralizing activity.

V _H H-IgG

The present invention also includes a binding agent comprising _VHH peptide monomers, said _VHH peptide monomers being linked to antibodies with additional individual Fc domains. _VHH -IgG binding agents comprising 1, 2, 3, 4, or more _VHH peptide monomers are linked to the light chain (kappa or lambda) and heavy chain of an IgG antibody lacking the antibody's variable region. Thus, the peptide monomers replace the variable region of the antibody.

_VHH peptide monomers can be any of those provided in Table 1 above, and include 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7) _VHH peptide monomers. When two or more monomers are linked, the monomers can be linked by a flexible peptide linker, typically comprising 10 to 20 amino acids. Suitable linkers include those provided in Table 2, such as linker-1 (SEQ ID NO: 9), linker-2 (SEQ ID NO: 11), and linker-3 (SEQ ID NO: 13).

_VHH -IgG conjugates comprise an octamer containing an IgG antibody and first, second, third, and fourth _VHH peptide monomers, wherein the _VHH peptide monomers are specifically binding to epitopes of TcdA or TcdB. The first and second _VHH peptide monomers are linked together and attached to the amino terminals of the two light chains of the antibody, where the light chains lack the antibody variable region. The third and fourth _VHH peptide monomers are linked together and attached to the amino terminals of the two heavy chains of the antibody, where the heavy chains lack the antibody variable region. Because this conjugate has four _VHH peptide monomers, it can be monospecific (where all monomers bind to the same epitope), bispecific (where monomers bind to two different epitopes), trispecific (where monomers bind to three different epitopes), or tetraspecific (where monomers bind to four different epitopes).

A specific example of a tetraspecific _VHH -IgG binder is the ABAB-IgG binder, which comprises an IgG antibody, two sets of linked first and second _VHH peptide monomers, and two sets of linked third and fourth _VHH peptide monomers in a tetraspecific octamer. The IgG antibody comprises two arms, each containing a light chain lacking a variable region and a heavy chain lacking a variable region, and each chain has an amino terminus. For each arm of the antibody, one set of linked first and second _VHH peptide monomers is linked to the amino terminus of the light chain, and one set of linked third and fourth _VHH peptide monomers is linked to the amino terminus of the heavy chain. The _VHH peptide monomers have binding specificity for epitopes of TcdA or TcdB (see Figure 1). This binder is called "tetraspecific" because it recognizes four different toxin epitopes. It is called an "octamer" because it carries eight _VHH peptide monomers (two copies of the first monomer, two copies of the second monomer, two copies of the third monomer, and two copies of the fourth monomer). In some respects, the first, second, third, and fourth _VHH peptide monomers may each have binding specificity for different epitopes. In some respects, both _VHH peptide monomers can bind specifically to epitopes of TcdA, and both _VHH peptide monomers can bind specifically to epitopes of TcdB. In some respects, the _VHH peptide monomers independently bind specifically to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

Specific examples of the tetraspecific octamer ABAB-IgG binder include a light (kappa) chain having the amino acid sequence shown in SEQ ID NO: 46 (AA6/E3 kappa) or a sequence variant having at least 95% sequence identity therewith, and a heavy chain having the amino acid sequence shown in SEQ ID NO: 44 (AH3/5D heavy) or a sequence variant having at least 95% sequence identity therewith. In this respect, the sequence variant retains toxin-neutralizing activity. The variant amino acids of the sequence variant may be located within the framework region of the _VHH peptide monomer. This binder is prepared by preparing two separate expression vectors, a first encoding the _VHH peptide monomer AH3/5D (linked in the stated order) linked to a human IgG1 antibody heavy chain lacking the variable region, and a second encoding the _VHH peptide monomer AA6/E3 (linked in the stated order) linked to a human IgG1 antibody light (kappa) chain lacking the variable region. The nucleotide sequence encoding the AA6/E3-IgG1 light (kappa) chain is provided in SEQ ID NO: 47. The nucleotide sequence encoding the AH3/5D-IgG1 heavy chain is provided in SEQ ID NO: 45. This invention includes sequence variants of ABAB-IgG having at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity over the entire length of the protein sequence and retaining the toxin-binding and/or neutralizing activity of the wild-type protein. This invention further includes polynucleotide sequences encoding these sequence variants and their complementary strands.

Bispecific or tetraspecific tetrameric IgG binders are included in this invention. Such binders comprise an IgG antibody and first, second, third, and fourth _VHH peptide monomers, wherein the IgG antibody comprises two arms, each arm comprising a heavy chain lacking a variable region and a light chain lacking a variable region, and each chain has an amino terminus, wherein for the first arm of the antibody, the first _VHH peptide monomer is attached to the amino terminus of the light chain, and the second _VHH peptide monomer is attached to the amino terminus of the heavy chain, wherein for the second arm of the antibody, the third _VHH peptide monomer is attached to the amino terminus of the light chain, and the fourth _VHH peptide monomer is attached to the amino terminus of the heavy chain, and wherein the _VHH peptide monomers have binding specificity to epitopes of TcdA or TcdB. When the binder is “tetraspecific,” it recognizes four different toxin epitopes; when it is “bispecific,” it recognizes two different toxin epitopes. Because the binder has four _VHH peptide monomers, the binder is a "tetramer" (when bispecific, the first and second monomers have the same sequence and bind the same epitope, and the third and fourth monomers have the same sequence and bind the same epitope; when tetraspecific, each monomer has a different sequence and binds a different epitope).

When the binder is bispecific, the first and third monomers have binding specificity to different epitopes, the first and second monomers have the same amino acid sequence, and the third and fourth monomers have the same amino acid sequence. In some respects, one of the _VHH peptide monomers has binding specificity to the TcdA epitope, and one of the _VHH peptide monomers has binding specificity to the TcdB epitope.

When the binding agent is four-specific, each _VHH peptide monomer has binding specificity for a different epitope. In some respects, two VHH peptide monomers have binding specificity for the TcdA epitope and two _VHH peptide monomers have binding specificity for the TcdB epitope.

In some respects, each _VHH peptide monomer exhibits binding specificity to the TcdA epitope. In other respects, each _VHH peptide monomer exhibits binding specificity to the TcdB epitope.

In some respects, the _VHH peptide monomer independently exhibits binding specificity for epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB.

Specific examples of bispecific tetrameric IgG binders include a light (kappa) chain having the amino acid sequence shown in SEQ ID NO: 40 (AA6 kappa) and a heavy chain having the amino acid sequence shown in SEQ ID NO: 36 (AH3 heavy). The invention also includes sequence variants having at least 95% sequence identity, wherein the sequence variants retain toxin-neutralizing activity. The variant amino acids of the sequence variants may be located within the frame region of the _VHH peptide monomer.

Another specific example of a bispecific tetrameric IgG binder includes a light (kappa) chain having the amino acid sequence shown in SEQ ID NO: 42 (E3 kappa) and a heavy chain having the amino acid sequence shown in SEQ ID NO: 38 (5D heavy). The invention also includes sequence variants having at least 95% sequence identity, wherein the sequence variants retain toxin-neutralizing activity. The variant amino acids of the sequence variants may be located within the frame region of the _VHH peptide monomer.

Table 6 provides sequences for generating bispecific and tetraspecific _VHH -IgG binding agents. Other suitable pairings include (i) 5D-IgG1-heavy chain + AA6-light (kappa or lambda) chain, and (ii) AH3-IgG1-heavy chain + E3-light (kappa or lambda) chain.

Table 6

However, this invention includes the IgG1 heavy chain linked to any one of AH3, 5D, AA6, and E3, and the IgG1 light (kappa or lambda) chain linked to any one of AH3, 5D, AA6, and E3. Furthermore, this document covers all possible combinations of the heavy chain and the light (kappa or lambda) chain.

Humanized binder

Due to their small framework size and high identity with the human _VH framework of family III, _VHH peptide monomers are expected to exhibit low immunogenicity when administered to humans. Although the systematic application of small monovalent _VHH monomers appears to elicit little neutralizing antibody response, protein immunogenicity generally increases with size and complexity. Two major obstacles to repeated and/or long-term in vivo use of _VHH monomers are their potentially short half-life and potential immunogenicity. To increase potency and circulating half-life, _VHH monomers can be fused with the human IgG and Fc domains discussed herein. To address potential immunogenicity, _VHH monomers can be humanized as needed without compromising their expression levels, affinity, solubility, and stability. These strategies should result in good expression, stability, and solubility of the humanized _VHH monomer ( _hVHH monomer) while maintaining the antigen specificity and affinity of the cyclic donor _VHH .

_hVHH monomers with the highest identity to the human _VH gene and the highest binding/neutralizing activity are selected and then transferred into _VHH multimers (e.g., ABAB), _VHH -Fc, and VHH-IgG constructs to produce fully humanized binding agents, such as fully humanized ABAB, ABAB-IgG, and ABAB-Fc binding agents. The protein sequences of these humanized binding agents can be substantially identical to those of human antibody variants, although some CDR segments responsible for the antibody's ability to bind to its target antigen are of non-human origin. Therefore, this strategy reduces the chance of potential immunogenicity in vivo, thereby increasing their safety and half-life in vivo.

Therefore, the binding agents of the present invention cover humanized versions of each binding agent defined herein, including hV _HH peptide monomers.

Epitope binding fragment

The conjugates of this invention comprise epitope-binding fragments of each _VHH -Fc and _VHH -IgG conjugate as defined herein. Since the _VHH -Fc and _VHH -IgG conjugates are structurally equivalent to human IgG antibodies, wherein the variable region is replaced by a _VHH monomer, the term "human antibody fragment" also applies to such conjugates. Fragments include, but are not limited to, Fab fragments, F(ab') ₂ fragments, single-chain Fv (scFv) antibodies, fragments generated from Fab expression libraries, and bispecific and trispecific antibodies.

The _VHH -Fc and _VHH -IgG binding agents of the present invention include fully humanized, humanized, and chimeric binding agents. The binding agents can be monoclonal or polyclonal. Furthermore, the binding agents can be recombinant binding agents.

While preferably derived from mammals such as humans, apes, mice, rats, rabbits, guinea pigs, horses, cattle, sheep, goats, pigs, dogs, or cats, the binder can be produced in any animal species. For example, the binder can be human or humanized, or any binder formulation suitable for administration to humans.

Polynucleotides, expression vectors, host cells, and preparation methods

The present invention includes polynucleotides containing nucleotide sequences encoding each of the binding agents provided herein, and their complementary strands.

The present invention also includes expression vectors comprising polynucleotides and host cells containing the expression vectors. Suitable expression vectors include, for example, pcDNA3.1 and pSec-His, and plasmids for converting yeast cells into producers and secretors of the binding agents of the present invention. Suitable host cells include, for example, Chinese hamster ovary cells (CHO cells), human embryonic kidney cells 293 (HEK 293 cells), yeast cells, and insect cells.

The present invention further includes a method for producing a binder as defined herein, comprising culturing host cells under conditions that promote expression of a binder encoded by an expression vector, and recovering the binder from the cell culture.

Engineered strains of yeast

Each of the binding agents of the present invention can also be produced by engineered strains of yeasts. Therefore, the present invention also relates to strains of yeasts of the genus *Saccharomyces* (e.g., *Saccharomyces cerevisiae* and *Bula millefolium*), engineered to produce one or more binding agents as defined herein, including but not limited to _VHH monomeric binding agents (see Table 1), _VHH homodimeric binding agents, _VHH heterodimeric binding agents (see Table 3), ABAB binding agents, _VHH -Fc binding agents (see Table 5), _VHH -IgG binding agents (see Table 6), and their epitope binding fragments. In a preferred aspect, engineered strains of yeasts secrete the binding agents.

The identity of a yeast strain is limited to its ability to be engineered to produce and preferably secrete one or more of the binding agents of the present invention. In a preferred aspect of the invention, the yeast strain engineered to produce one or more binding agents is *Saccharomyces cerevisiae* or *Saccharomyces blakeana*. Therefore, the present invention includes engineered strains of *Saccharomyces cerevisiae* that produce one or more binding agents as defined herein, and engineered strains of *Saccharomyces cerevisiae* that secrete one or more binding agents as defined herein. The present invention also includes engineered strains of *Saccharomyces blakeana* that produce one or more binding agents as defined herein, and engineered strains of *Saccharomyces blakeana* that secrete one or more binding agents as defined herein. Suitable yeast strains also include *Schizosaccharomyces pombe*, *Saccharomyces paradoxus*, and *Saccharomyces unisporus*.

Saccharomyces boulardii is an FDA-designated generally considered safe (GRAS) organism and is commonly available over-the-counter for promoting gut health and improving gastrointestinal disorders caused by diarrhea. Numerous randomized, double-blind, placebo-controlled clinical trials have been conducted on this yeast strain to investigate its safety and efficacy against intestinal disorders, including CDI [42-46]. A suitable strain of Saccharomyces boulardii is Saccharomyces boulardii strain MYA796 (ATCC, Manassas, VA).

Specific examples of engineered yeast strains of the present invention are engineered yeast strains of the genus *Saccharomyces* that produce conjugates comprising a single chain of _VHH peptides linked to _VHH peptide monomer groups comprising 2, 3, 4 or more monomers, each of which preferably specifically binds to TcdA and/or TcdB. Therefore, the present invention includes engineered yeast strains of the genus *Saccharomyces* that produce _VHH peptide conjugates comprising at least one _VHH peptide monomer, wherein each _VHH peptide monomer has binding specificity to an epitope of *Clostridium difficile* toxin A (TcdA) or toxin B (TcdB). In some aspects, these conjugates comprise 2, 3, 4 or more linked _VHH peptide monomers. _VHH peptide monomers include, but are not limited to _{, VHH} peptide monomers 5D (SEQ ID NO: 1), E3 (SEQ ID NO: 3), AA6 (SEQ ID NO: 5), and AH3 (SEQ ID NO: 7).

Another specific example of an engineered yeast strain of the present invention is an engineered yeast strain that produces a conjugate comprising a _VHH peptide monomer linked to an IgG antibody, wherein the conjugate binds TcdA and/or TcdB as defined herein. In these IgG-based conjugates, the variable regions of the light and heavy chains of the IgG antibody are replaced by 1, 2, 3, 4 or more _VHH peptide monomers.

Another specific example of the engineered yeast strain of the present invention is an engineered yeast strain that produces a conjugate comprising _VHH peptide monomers linked to the Fc domain of an antibody, wherein the conjugate binds TcdA and/or TcdB as defined herein. In these Fc domain-based conjugates, one, two, three, four or more _VHH peptide monomers are linked to the hinge region, _CH2 , and _CH3 region of each arm of the Fc domain of the antibody heavy chain. Thus, the peptide monomers replace the Fab region of the antibody.

Another specific example of an engineered yeast strain of the present invention is an engineered yeast strain that produces a tetraspecific tetrameric binder, wherein the binder comprises linked first, second, third, and fourth _VHH peptide monomers, and wherein the _VHH peptide monomers independently have binding specificity to epitopes of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). In some aspects, the first, second, third, and fourth _VHH peptide monomers each have binding specificity to different epitopes. In some aspects, two _VHH peptide monomers have binding specificity to epitopes of TcdA, and two _VHH peptide monomers have binding specificity to epitopes of TcdB. In some aspects, the _VHH peptide monomers independently have binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB. Suitable _VHH peptide monomers include AH3 monomer (SEQ ID NO: 7), AA6 monomer (SEQ ID NO: 5), 5D monomer (SEQ ID NO: 1), and E3 monomer (SEQ ID NO: 3). Other monomers include, but are not limited to, those listed in Table 1.

In a preferred example, the present invention relates to an engineered strain of yeast, wherein the binding agent is ABAB, wherein the first and third monomers are specific for binding to the epitope of TcdA, the first and third monomers being _VHH peptide monomers AH3 (SEQ ID NO: 7) and AA6 (SEQ ID NO: 5), respectively, and wherein the second and fourth monomers are specific for binding to the epitope of TcdB, the second and fourth monomers being _VHH peptide monomers 5D (SEQ ID NO: 1) and E3 (SEQ ID NO: 3), respectively.

ABAB binding agents may include the amino acid sequence shown in SEQ ID NO: 19 or a sequence variant having at least 95% sequence identity with it, wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity or retains both.

The ABAB binder may also include an N-terminal secretion signal selected from the secretion signals provided in Table 4. In a preferred aspect, the N-terminal secretion signal is an AT secretion signal (MRFPSIFTAVLFAASSALA (SEQ ID NO: 99)) or an IVS secretion signal (MLLQAFLFLLAGFAAKISA (SEQ ID NO: 103)).

ABAB binders can be expressed from plasmids within yeast. The plasmid can be, but is not limited to, pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88). The ABAB binder encoded by the plasmid can include the amino acid sequence shown in SEQ ID NO: 107 or a sequence variant having at least 95% sequence identity with it, wherein said sequence variant retains TcdA and/or TcdB binding specificity, or retains toxin-neutralizing activity, or both.

ABAB binders can also be expressed from coding sequences integrated into the yeast chromosome. ABAB binders expressed from coding sequences integrated into the yeast chromosome may include the amino acid sequence shown in SEQ ID NO: 109 or a sequence variant having at least 95% sequence identity with it, wherein said sequence variant retains TcdA and/or TcdB binding specificity, or retains toxin-neutralizing activity, or both.

This invention also relates to engineered strains of *Saccharomyces* yeast that produce therapeutic proteins having binding specificity to a unique epitope of *Clostridium difficile* toxin A (TcdA) or toxin B (TcdB), or both. Preferably, the engineered strain of *Saccharomyces* yeast is *Saccharomyces cerevisiae* or *Bacillus blakeana*. Therapeutic proteins are any proteins that can bring about improvement or cure in a subject's medical condition, or that can inhibit or prevent the development of a subject's medical condition. Suitable therapeutic proteins include, but are not limited to, those capable of (a) replacing defective or abnormal proteins; (b) enhancing existing pathways; (c) providing novel functions or activities; (d) interfering with molecules or organisms; and (e) delivering other compounds or proteins, such as radionuclides, cytotoxic drugs, or effector proteins. Therapeutic proteins also include antibodies and antibody-based drugs, Fc fusion proteins, anticoagulants, blood factors, bone morphogenetic proteins, engineered protein scaffolds, enzymes, growth factors, hormones, interferons, interleukins, and thrombolytic agents. Therapeutic proteins further include bispecific monoclonal antibodies (mAbs) and multispecific fusion proteins, mAbs conjugated to small molecule drugs, and proteins with optimized pharmacokinetic properties.

Methods for preparing engineered yeast strains

This invention also relates to methods for engineering strains of yeast to produce one or more binding agents as defined herein. There are no particular limitations on the means used to produce engineered yeast strains, and many known techniques exist for engineering yeast to produce homologous and heterologous proteins known to those skilled in the art. In some aspects of these methods, *Saccharomyces cerevisiae* or *Saccharomyces blakeana* are engineered to produce binding agents.

As an example, yeast strains can be engineered to produce one or more binding agents as defined herein for the production of binding agents by (a) transforming a yeast strain of the genus *Saccharomyces* with an expression vector encoding the binding agent, and (b) screening the resulting yeast. In one respect, the expression vector is the plasmid pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88). Although this plasmid encodes a specific ABAB binding agent, the coding region of such a binding agent can be replaced by the coding region of any binding agent as defined herein.

As a further example, yeast strains can be engineered to produce one or more binding agents as defined herein by (a) integrating a polynucleotide sequence encoding the binding agent into the genome of a yeast strain of the genus *Saccharomyces*, and (b) screening the yeast strain from (a). In some respects, chromosome integration is performed using CRISPR technology [85-88]. As an example, such a method may include the following steps: (a) amplifying a polynucleotide sequence encoding an ABAB binder from plasmid pCEV-G4-Km-TEF-AT-yABAB hAA6T83N-tagless (SEQ ID NO: 90) using primers comprising (i) nucleic acid sequences homologous to a selected yeast chromosome integration site and (ii) nucleic acid sequences homologous to the 5' and 3' regions of the ABAB binder encoding sequence of the plasmid, to generate an integration cassette; (b) transforming yeast with pCRI-Sb-δ1 (SEQ ID NO: 91) or pCRI-Sb-δ2 (SEQ ID NO: 92) under conditions that promote spontaneous integration of the integration cassette into a double-strand break site to induce a double-strand break within the corresponding yeast chromosome δ site; and (c) screening the transformed yeast in (b) to generate the ABAB binder.

Although plasmid pCEV-G4-Km-TEF-AT-yABAB hAA6T83N-tagless encodes a specific ABAB binder, the coding region for this binder can be replaced by the coding region for any binder as defined herein.

Suitable means for screening yeasts for generating binders will be apparent to those skilled in the art, and include, but are not limited to, immunoassays such as ELISA or Western blotting.

Treatment and prevention methods

The binder and engineered yeast strains of the present invention can be used in methods for treating or preventing disease symptoms induced by Clostridium difficile in subjects. These methods typically involve administering a therapeutically effective amount of one or more binders and/or one or more engineered yeast strains as defined herein to a subject suffering from or at risk of developing Clostridium difficile infection. In some aspects of this embodiment, the disease symptom induced by Clostridium difficile is diarrhea.

The binders and engineered yeast strains of the present invention can also be used to neutralize Clostridium difficile toxins TcdA and/or TcdB in subjects with Clostridium difficile infection. These methods typically involve administering a therapeutically effective amount of one or more binders and/or one or more engineered yeast strains as defined herein to a subject suffering from Clostridium difficile infection.

The binders and engineered yeast strains of the present invention can be further used in methods for treating Clostridium difficile infection in subjects. These methods generally involve administering a therapeutically effective amount of one or more binders and/or one or more engineered yeast strains as defined herein to a subject suffering from Clostridium difficile infection. As defined herein, these same methods can be used to treat CDI.

The binders and engineered yeast strains of the present invention can also be used in methods for maintaining normal intestinal function in subjects suffering from Clostridium difficile infection. These methods typically involve administering a therapeutically effective amount of one or more binders and/or one or more engineered yeast strains as defined herein to a subject suffering from or at risk of developing Clostridium difficile infection.

Binders and engineered strains of yeast can also be used for immunoprophylaxis to prevent immediate CDI threats. Additionally, passive immunoprophylaxis can be used to prevent both direct and long-term CDI threats. Each approach has its specific advantages and is suitable for specific high-risk populations. These approaches typically involve administering a therapeutically effective amount of one or more binders and/or one or more engineered strains of yeast as defined herein to subjects at risk of Clostridium difficile infection.

In a preferred aspect of the method of the present invention, the yeast is *Saccharomyces cerevisiae* or *Saccharomyces blakeana*.

Each method of the present invention may include administering one or more binders and/or engineered strains of one or more yeast species into one or more pharmaceutical formulations, said pharmaceutical formulations comprising binders and/or engineered strains of yeast species and pharmaceutically acceptable carriers or diluents. In a preferred aspect, the yeast species is *Saccharomyces cerevisiae* or *Bacillus spp.*

As used herein, the terms "treat," "treating," and "treatment" have their common and customary meanings and include one or more of the following: blocking, improving, or reducing the severity and/or frequency of symptoms of Clostridium difficile infection or Clostridium difficile-associated disease (CDI) in a subject; and/or partially or completely inhibiting biological activity and/or promoting immune clearance of Clostridium difficile TcdA and/or TcdB in a subject infected with Clostridium difficile; and/or inhibiting the growth, division, spread, or proliferation of Clostridium difficile cells or Clostridium difficile infection in a subject. Treatment means blocking, improving, reducing, or inhibiting by about 1% to about 100% relative to a subject in which the methods of the present invention have not been implemented. Preferably, the blocking, improving, reducing, or inhibiting is about 100%, 99%, 98%, 97%, 96%, 95%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5%, or 1% compared to a subject in which the methods of the present invention have not been implemented.

As used herein, the terms “prevent,” “preventing,” and “prevention” have their common and customary meanings and include one or more of the following: preventing, avoiding, mitigating, or blocking the colonization, development, or progression of *Clostridium difficile* in a subject; and/or partially or completely inhibiting the biological activity and/or toxicity of TcdA and/or TcdB in a subject infected with *Clostridium difficile*; and/or preventing, avoiding, mitigating, or blocking the growth, division, spread, or proliferation of bacterial cells or bacterial infection in a subject. Prevention means prevention of at least about 95% of the bacterial cells or infection compared to a subject who was not given prevention. Preferably, prevention is about 100%, about 99%, about 98%, about 97%, about 96%, or about 95%. The effects of prevention can last for days (e.g., 1, 2, 3, 4, 5, 6, or 7 days), weeks (e.g., 1, 2, 3, or 4 weeks), or months (e.g., 1, 2, 3, 4, 5, 6, or more months).

The treatment and prevention methods provided herein can be supplemented by administering a therapeutically effective amount of antibiotics to the subject. Preferably, the antibiotics will have antibacterial activity against Clostridium difficile.

pharmaceutical preparations

Although the binder and engineered yeast strains can be directly administered to subjects, the method of the present invention is preferably based on the administration of an engineered yeast strain comprising one or more binders and/or one or more engineered yeast strains, and a pharmaceutically acceptable carrier or diluent. Therefore, the present invention includes pharmaceutical formulations comprising one or more binders and/or one or more engineered yeast strains as defined herein, and a pharmaceutically acceptable carrier or diluent.

Pharmaceutically acceptable carriers and diluents are generally known and will vary depending on the specific binder or engineered yeast strain being applied and the mode of administration. Examples of commonly used carriers and diluents include, but are not limited to, saline, buffered saline, dextrose, water for injection, glycerol, ethanol and combinations thereof, stabilizers, solubilizers and surfactants, buffers and preservatives, tonics, fillers and lubricants. Formulations including binders and/or engineered yeast strains will generally be prepared and cultured without any non-human components, such as animal serum (e.g., bovine serum albumin).

Pharmaceutical formulations comprising one or more binders and/or one or more engineered strains of yeast known to those skilled in the art can be administered to subjects using modalities and techniques known to them. The characteristics of CDI disease make it more suitable for treatment and prevention using colonic delivery of therapeutic agents (i.e., targeted delivery of the binder to the lower gastrointestinal tract, such as the large intestine or colon). Other delivery modalities include, but are not limited to, oral, nasal, rectal, and intravenous or aerosol administration. Other modalities include, but are not limited to, intradermal, subcutaneous (s.c., s.q., sub-Q, Hypo), intramuscular (i.m.), intraperitoneal (i.p.), intra-arterial, intramedullary, intracardiac, intra-articular (joint), intrasynovial (joint fluid region), intracranial, intraspinal, and intrathecal (spinal fluid).

Depending on the route of administration, the dose can be administered all at once, such as in oral formulations in capsules or liquids, or slowly over a period of time, such as intramuscular or intravenous administration.

The amount of the binder administered to the subject, either alone or in the pharmaceutical formulation, is the effective amount for treating or preventing infection. Therefore, when implementing the method of the present invention, a therapeutically effective amount is administered to the subject. Typically, the binder is administered at a rate of about 1 μg/kg to about 1000 mg/kg of body weight per subject. Suitable ranges also include about 50 μg/kg to about 500 mg/kg, and about 10 μg/kg to about 100 mg/kg. However, the amount of binder administered to the subject will vary within a wide range depending on the site, source, extent, and severity of the infection, the age and condition of the subject to be treated, the route of administration, etc. The physician will ultimately determine the appropriate dosage to be used.

The amount of engineered yeast strain administered to the subject, either alone or in the pharmaceutical formulation, is the effective amount for treating or preventing infection. Therefore, when implementing the method of the present invention, a therapeutically effective amount is administered to the subject. Typically, an engineered yeast strain of about 1 μg/kg to about 1000 mg/kg of body weight is administered to each subject. Suitable ranges also include about 50 μg/kg to about 500 mg/kg, and about 10 μg/kg to about 100 mg/kg. However, the amount of engineered yeast strain administered to the subject will vary within a wide range depending on the site, source, extent, and severity of the infection, the age and condition of the subject to be treated, the route of administration, etc. The physician will ultimately determine the appropriate dose to be used.

The frequency of administration of binders, engineered strains of yeast, and pharmaceutical formulations including binders and/or engineered strains of yeast will vary depending on factors including the location of the bacterial infection, details of the infection to be treated or prevented, and the mode of administration. Each formulation may be administered independently 4, 3, 2 or once daily, every other day, every three days, every four days, every five days, every six days, once weekly, every eight days, every nine days, every ten days, biweekly, monthly, and bimonthly.

The duration of treatment or prevention will be based on the location and severity of the infection being treated or the relative risk of infection, and will be best determined by the attending physician. However, continued treatment is expected to last for days, weeks, or months.

In each embodiment and aspect of the invention, the subject is a human, a non-human primate, a bird, a horse, a cow, a goat, a sheep, a companion animal such as a dog, a cat, or a rodent or other mammal. Subjects to whom the methods of the invention may be applied include those with underlying diseases or conditions that make them more susceptible to Clostridium difficile infection.

The present invention also provides a kit comprising one or more containers containing one or more binders, one or more engineered strains of yeast, or one or more pharmaceutical preparations including binders and/or engineered strains of yeast. The kit may also include instructions for use. Further information relating to the kit may be notifications from government agencies regarding regulations for the manufacture, use, or sale of pharmaceutical or biological products, reflecting approvals for human administration by the manufacturing, using, or selling organization.

III. Example

V _H H monomer and heterodimer binder

An efficient platform was established for screening single-domain (monomers) and single-specific _VHH peptide monomers targeting specific domains of toxins TcdA and TcdB. A group of VHH monomers binding to different domains of TcdA or TcdB was prepared using highly immunogenic nontoxic whole toxins for immunization and bioactive chimeric toxins (with normal functional domains) for screening. Most of these _VHH monomers had potent neutralizing activity and their binding to specific domains was determined (Figure 2). As previously mentioned, the nontoxic whole toxin has a point mutation at its enzymatic glucosyltransferase domain [33]. _The bioactive chimeric toxin was generated by switching functional domains between TcdA and TcdB, which has been previously described [33].

Several _VHH monomers bind to highly conserved TcdA/TcdB epitopes. For example, _VHHE3 binds to the Rho GTPase binding site and blocks glucosylation; _VHHAH3 binds to the GT domain of the toxin; and _VHH7F binds to the cysteine protease cleavage site and blocks GT domain cleavage and release. Some _VHH monomers possess potent neutralizing activity, capable of blocking the cytotoxic activity of the toxin at nM concentrations (see Table 1; Figures 3A and 3B).

To enhance binding activity, two domains (dimers) were generated, resulting in bispecific _VHH heterodimers (Table 3; Figure 3C), allowing a single protein to target two distinct epitopes of the toxin. These bispecific _VHH heterodimers exhibited significantly enhanced neutralizing activity compared to an equimolar mixture of the same two _VHH monomers (Figure 3D). The heterodimers 5D/E3 and AH3/AA6 were found to completely protect mice from lethal systemic TcdB or TcdA attacks, respectively, while mixtures of 5D and E3 or AA6 alone provided only partial protection (Figures 3E and 3F).

A tetravalent trispecific _VHH binder (ABA) was generated by genetic fusion of _VHH with the highest neutralizing activity targeting conserved, non-overlapping epitopes (AH3/E3/E3/AA6)[41]. This rationally designed toxin binder achieves significantly enhanced binding affinity and neutralizing activity for each monomer and effective therapeutic efficacy against fulminant CDI. ABA can broadly neutralize toxins from 11 different clinical isolates of Clostridium difficile with TcdA ^{+ and} TcdB ⁺ , but cannot neutralize TcdB from two TcdA- ^and TcdB ⁺ strains. The amino acid sequence of ABA is shown in SEQ ID NO: 111.

The _VHH monomers, including heterodimers, are connected using flexible connectors selected from SEQ ID NO: 9-13 (Table 2).

ABAB binder

By linking the _VHH monomers AH3, 5D, E3, and AA6, namely ABBA (AH3/5D/E3/AA6) and ABAB (AH3/5D/AA6/E3), four-domain (tetramer) tetraspecific _VHH binders are generated. These tetraspecific tetramer binders target conserved non-overlapping epitopes and exhibit excellent toxin-neutralizing activity. In the ABAB design (Figure 4), the _VHH peptide monomers AH3 and AA6 are separated by placing the 5D monomer between them, since AH3 and AA6 bind GT and TD respectively (Figure 2), and they are spatially far apart. This design allows AH3 and AA6 to bind TcdA simultaneously.

In the construction of the ABAB binding agent, a flexible linker is placed between the V _H H monomers (see Figure 4). The complete nucleic acid sequence encoding ABAB is provided in SEQ ID NO: 20; the amino acid sequence of the protein is provided in SEQ ID NO: 19.

In some variants, a His ₍₆₎ tag is provided at the amino terminus of the protein to aid purification, an E tag is provided at the carboxyl terminus of the protein to aid detection, and/or an albumin-binding peptide (ABP, DICLPRWGCLWD; SEQ ID NO: 21) is placed at the carboxyl terminus of the construct to increase the serum half-life of the protein (see Figure 4).

ABAB was found to exhibit significantly enhanced binding affinity (Table 7) and neutralizing activity (Table 8) for both single monomers and ABA. In Table 8, Vero cells were exposed to 5 ng/ml of TcdA in the presence of serially diluted AA6, AH3, ABAB, or Merck anti-TcdA HuMab [9]. The minimum dose of antibody that protects cells from TcdA-induced cell rounding disruption is shown.

Table 7

Table 8

AA6 AH3 ABAB Merck Anti-TcdA HuMab 8nM 8nM 0.25nM >10nM

It was also found that ABAB competes with all four individual _VHH peptide monomers in competitive ELISA and can bind both TcdA and TcdB simultaneously, as determined by sandwich ELISA. Furthermore, ABAB exhibits broad reactivity, capable of neutralizing toxins from 13 different Clostridium difficile strains representing most currently circulating strains (Table 9).

Table 9

Because ABAB has shown high efficacy in binding and neutralizing both toxins, its efficacy in treating fulminant CDI was evaluated. In mice, a single injection of ABAB as low as 40 μg/kg one day after Clostridium difficile spore challenge reversed fulminant CDI. Three days after infection, none of the ABAB-treated mice died, while 50% of the control mice were morbid (Fig. 5A). The preventive mortality rate of ABAB after TcdA and TcdB systemic challenges was four orders of magnitude higher than that of Merck HuMabs (Fig. 5B) [9]. Thus, ABAB has remarkable in vivo efficacy against Clostridium difficile toxins and spore challenges.

Animal and human studies have shown that passively administered antitoxin antibodies provide protection against CDI. Preliminary studies here also show that the antitoxin polysera protects mice from primary CDI (Figures 6A and 6B) and recurrent/relapsed CDI. These findings and results in Figures 5A and 5B support this hypothesis and provide the basic principles for developing parenteral ABAB immunization strategies for CDI prevention. To achieve the goal of optimizing ABAB for systemic delivery, chimeric and humanized ABABs, namely _VHH -Fc and _VHH -IgG binders and humanized proteins _hVHH -Fc and _hVHH -IgG, were generated as shown in Figure 1, followed by evaluation of the in vivo neutralizing activity and protective effects of the major proteins in animal models. Details regarding the preparation and testing of the additional binders are provided in the following paragraphs.

ABAB-Fc

The ABAB-Fc binder was prepared by generating an expression vector encoding a _VHH peptide monomer AH3/5D/AA6/E3 (linked in the order stated) linked to a human IgG1 Fc domain. The _VHH peptide monomer was isolated via the flexible linkers listed in Table 2. The nucleic acid sequence encoding the protein is provided in SEQ ID NO: 23. ABAB-Fc was expressed and purified from the supernatant of stably transfected HEK293 cell line cultures using protein A beads, under conditions allowing disulfide bond formation and divalent molecule generation. The expression level was approximately 20 mg/L of culture supernatant. ABAB-Fc was fully functional in binding and neutralizing TcdA and TcdB (data not shown). The amino acid sequence of ABAB-Fc is provided in SEQ ID NO: 22.

Monospecific _VHH -Fc binders (AH3-Fc, 5D-Fc, E3-Fc, and AA6-Fc) and bispecific _VHH -Fc binders (AH3/5D-Fc and AA6/E3-Fc) were also prepared using this Fc-fusion system. The sequences of these additional binders are provided in Table 5 above.

ABAB-IgG

As shown in Figure 1, bispecific _VHH -IgG (AH3/5D-IgG and E3/AA6-IgG) can be generated by fusing monomers with the heavy and light (kappa or lambda) chains of human IgG, respectively. Tetramer-specific _VHH -IgG (ABAB-IgG) binders can be generated by fusing dimers with the heavy and light chains of human IgG, respectively. Co-transfection of heavy and light chain constructs yields AH3/5D-IgG, E3/AA6-IgG, and ABAB-IgG chimeric proteins. Separating the two _VHHs into heavy and light chains may improve the yield and stability of bispecific and tetraspecific _VHH chimeric proteins. This can determine whether _VHH -human IgG chimeric antibodies contribute to the stability and efficacy of ABAB in vivo. Similarly, further improvements in the in vivo half-life of ABAB-IgG can be tested in ABAB-IgG variants with enhanced binding affinity to the FcRn receptor.

Bispecific (AH3/5D-IgG1 and E3/AA6-IgG1) and tetraspecific (ABAB-IgG1) IgG1 binders were prepared by co-transfection of expression vectors encoding the heavy and light (kappa) chains of each binder. _VHH peptide monomers were separated using the flexible linkers shown in Table 2.

Bispecific tetrameric _VHH -IgG1 conjugates are generated by preparing two separate expression vectors: the first encodes a _VHH peptide monomer linked to the heavy chain (CH1-hinge-CH2-CH3) of a human IgG1 antibody lacking the heavy chain variable region, and the second encodes a _VHH peptide monomer linked to the light (kappa) chain (CK) of a human IgG1 antibody lacking the light chain variable region. These conjugates are bispecific and tetrameric because each light chain of the resulting conjugate is linked to the first _VHH monomer, and each heavy chain of the resulting conjugate is linked to the second _VHH monomer. The sequences of these additional conjugates are provided in Table 6 above. Suitable pairings include (i) AH3-IgG1-heavy chain + AA6-light (kappa or lambda) chain, (ii) 5D-IgG1-heavy chain + E3-light (kappa or lambda) chain, (iii) 5D-IgG1-heavy chain + AA6-light (kappa or lambda) chain, and (iv) AH3-IgG1-heavy chain + E3-light (kappa or lambda) chain.

Tetraspecific octamer ABAB-IgG binders were prepared. These binders are tetraspecific and octamer because each light (kappa or lambda) chain of the resulting binder is linked to two (first and second) linked _VHH monomers, and each heavy chain of the resulting binder is linked to two (third and fourth) linked _VHH monomers, wherein the first, second, third, and fourth monomers bind different epitopes.

A specific tetraspecific octamer ABAB-IgG (Figure 7) binder was generated by preparing two separate expression vectors. The first vector encodes a _VHH peptide monomer AH3/5D (linked in the order described above) linked to the heavy chain (CH1-hinge-CH2-CH3) of a human IgG1 antibody lacking the heavy chain variable region, and the second vector encodes a _VHH peptide monomer AA6/E3 (linked in the order described above) linked to the light chain (kappa) chain (CK) of a human IgG1 antibody lacking the light chain variable region. The nucleotide sequence encoding the AH3/5D-IgG1 heavy chain is provided in SEQ ID NO: 45; the amino acid sequence is provided in SEQ ID NO: 44. The nucleotide sequence encoding the AA6/E3-IgG1 kappa chain is provided in SEQ ID NO: 47; the amino acid sequence is provided in SEQ ID NO: 46.

Bispecific (AH3/5D-IgG1 and E3/AA6-IgG1) and tetraspecific (ABAB-IgG1) IgG1 binders were expressed and purified in the supernatant of stably transfected HEK293 cell line cultures under conditions allowing disulfide bond formation and divalent molecule production. SDS-PAGE showed that the purified ABAB-IgG1 had a purity of over 90% of the total molecular weight (light and heavy chains together) around 218 kDa on a non-reducing gel (data not shown). On a reduced gel, the heavy chain molecular weight was 68 kDa and the light chain molecular weight was 41 kDa.

The binding of ABAB-IgG1 to TcdA and TcdB was determined. Figures 8A-8B show a comparison of the binding of ABAB-IgG1 to both toxins with individual components (AH3, AA6, E3, and 5D). Figure 8A shows the results where plates were coated with 1 μg/ml TcdA (TxA). Serially diluted ABAB-IgG was added at concentrations of 0, 0.64, 3.2, 16, 80, 400, and 2,000 ng/ml. Plates were washed and Merck Ab (anti-TcdA), Fc-ABBA (ABAB-Fc), Habab (ABAB-IgG), and _VHH anti-TcdB monomers AA6 and AH3 were added at specified amounts (ng/ml). Appropriate labeled antibodies were used for detection. Figure 8B shows the results where plates were coated with 1 μg/ml TcdB (TxB). Serially diluted ABAB-IgG was added at concentrations of 0, 0.64, 3.2, 16, 80, and 400 ng/ml. Plates were washed and Merck Ab (anti-TcdA), Fc-abba (ABAB-Fc), Habab (ABAB-IgG), and V _H H anti-TcdB monomers E3 and 5D were added at specified amounts (ng/ml). Appropriate labeled antibodies were used for detection.

As expected, as determined by sandwich ELISA (Figures 9A-9B), the four specific antibodies can bind to both TcdA and TcdB simultaneously. In the first set of experiments, plates were plated with 1 μg/ml TcdA (TxA). Serially diluted ABAB-IgG (Habab) was added at concentrations of 0, 1.6, 8, 40, 200, and 1000 ng/ml. Plates were washed and the following amounts of TcdB were added: 1.6, 8, 40, 200, and 1000 ng/ml. Mouse anti-TxB antibody (500x) and goat anti-mouse IgG-HRP antibody (3000x) were used for detection. The results provided in Figure 9A show the detection of TxB by TxA plate-coating, indicating that IgG-ABAB binds to both TxA and B simultaneously. In the second set of experiments, plates were plated with 1 μg/ml TcdB (TxB). Serially diluted ABAB-IgG (Habab) was added at concentrations of 0, 1.6, 8, 40, 200, and 1000 ng/ml. Plates were washed and TcdA was added at concentrations of 1.6, 8, 40, 200, and 1000 ng/ml. Detection was performed using mouse anti-TxA antibody (500x) and goat anti-mouse IgG-HRP antibody (3000x). The results shown in Figure 9B, demonstrating TxA detection via TxB plating, further indicate that IgG-ABAB binds to both TxA and B.

The neutralizing activity of ABAB-IgG1 against the cytopathic effects of toxins on cultured cells was also investigated. Before adding 100 μL of Vero cell monolayer to the medium, TcdA (100 ng/mL, Fig. 10A) was mixed with serially diluted Merck anti-TcdA human monoclonal antibody ABAB-IgG1 (Hababa) and _VHH anti-TcdA monomers AA6 and AH3, and incubated at 37°C for 24 hours. The results shown in Fig. 10A indicate that ABAB-IgG1 was at least 1000-fold more effective than the Merck antibody in neutralizing TcdA. In a similar experiment, before adding 100 μL of Vero cell monolayer to the medium, TcdB (10 pg/mL, Fig. 10B) was mixed with serially diluted Merck anti-TcdB human monoclonal antibody ABAB-IgG1 (Hababa) and _VHH anti-TcdB monomers E3 and 5D, and incubated at 37°C for 24 hours. The results shown in Figure 10B indicate that ABAB-IgG1 is at least 1000 times more effective than Merck antibody in neutralizing TcdB.

The in vivo neutralizing activity of ABAB-IgG1 was investigated in a mouse CDI model, and the results are shown in Figure 11. Mice were administered a lethal dose of a mixture of TcdA and TcdB (25 ng of each toxin per mouse) followed by administration of ABAB-IgG (10, 30, or 100 μg/kg), a mixture of Merck antitoxin A and antitoxin B antibodies (10 mg/kg), or PBS 4 hours later. The results showed that the neutralizing activity of ABAB-IgG was significantly greater than that of Merck antibodies, even at lower concentrations.

Animal testing of ABAB-IgG

ABAB-IgG binders were tested in both prophylactic treatment and re-attack survival assays. Figure 12 shows the experimental design for both studies. Female C57 mice aged 6–8 weeks were used, and conditions included: PBS: 10 ml/kg, i.p., n = 14; ABAB-IgG: 200 μg/kg, i.p., n = 10; ABAB-IgG: 1 mg/kg, i.p., n = 10; ABAB-IgG: 5 mg/kg, i.p., n = 10.

Table 10 provides a summary of the results observed in mice treated with Clostridium difficile spores (UK1,027/BI/NAP1 prevalent strain). ABAB-IgG or PBS was administered one day prior to the administration of Clostridium difficile spores. It can be seen that ABAB-IgG showed dose-related prophylactic protection against CDI, with 5 mg/kg showing complete protection against all examined parameters, and 200 μg/kg was found to be more effective than 200 μg/kg of the bispecific _VHH fusion antibody ABA [41].

Table 10

Table 11 summarizes the results observed in mice after re-challenge to Clostridium difficile spores. ABAB-IgG or PBS was administered 15 days prior to the Clostridium difficile spore administration. It can be seen that a single dose of ABAB-IgG showed some protection against CDI induced by spore re-challenge, but the protective efficacy was much lower compared to the initial challenge. This is likely due to the decline in antibody levels over time and the production of antibodies in the PBS group after the initial challenge.

Table 11

The intestinal delivery of IgG-ABAB was also tested to protect mice from fulminant CDI. Injection of a single IgG-ABAB into the ceca of mice after laparotomy completely protected them from fulminant CDI with fatal outcomes, while 50% of control mice died (data not shown). Disease progression and severity were assessed daily using a clinical scoring system modified from previous publications [62], which included four criteria (activity level, posture, coat, and diarrhea), each graded on a scale of 0–4 and summed to produce a score with a maximum value of 16. A normal mouse scored 0, and a mouse found to be dead scored 16. Mice with scores equal to or greater than 11 were to be euthanized. Only one mouse in the IgG-ABAB treatment group developed transient diarrhea, while mice injected with PBS developed severe CDI disease symptoms (data not shown). Therefore, Ig-ABAB delivered manually into the mouse intestine by injection showed a potent therapeutic effect.

Expression, purification and evaluation of binders

Multiple selection criteria were used to select the binding agents produced in the experiments described in this paper. First, each construct defined herein was used for transient transfection of 293T cells to prepare small-scale recombinant proteins by protein A affinity chromatography. The yield of each construct was determined by quantitative ELISA. Second, the binding activity of the recombinant proteins was screened using ELISA and surface plasmon resonance (SPR) to select constructs that retained their original binding activity to the toxin. Third, the neutralizing activity of the proteins was assessed in in vitro assays (Figure 3).

Accumulated observations suggest that the multireactivity and/or autoreactivity of recombinant conjugates in vivo are potential concerns related to their in vivo safety and half-life. The application of selected ABAB conjugates as systemic binding agents for the prevention of primary acute CDI may require chimeric and humanized ABAB proteins to limit multireactivity and/or autoreactivity. Advances in proteomics have enabled the in vitro screening of recombinant antibodies for multireactivity and autoreactivity, providing a valuable tool for alternative therapeutic antibodies. Therefore, the potential multireactivity and autoreactivity of selected humanized conjugates with good yields, high binding affinity, and effective neutralizing activity can be further tested using autoantigen microarray assays and ProtoArray protein microarrays (Invitrogen).

The in vivo toxicity, serum half-life, and immunogenicity of the candidate ABAB-Fc and ABAB-IgG binders can be evaluated from the above in vitro assays.

Production of ABAB by Saccharomyces cerevisiae (Sc-ABAB)

Means for producing and delivering the binding agent in vivo to the gut of subjects with CDI or at risk of developing CDI have been developed. Because Saccharomyces cerevisiae is genetically similar to Saccharomyces blakeana [52,53] and genetic tools are readily available for Saccharomyces cerevisiae, Saccharomyces cerevisiae was first used for validation of ABAB secretion.

A novel bispecific sandwich ELISA method was first developed to assess ABAB secretion. This setup utilized purified TcdA and TcdB as binding antigens for bispecific ABAB and an α-TcdA antibody for detection (Figure 13A). For standardization, plates were spread with TcdB (1 μg/ml), and serially diluted ABBA ((AH3-E3-E3-AA6)) standards were added. Then, serially diluted rTcdA (1 μg/ml to 7.8 ng/ml) was added. TcdA capture was then measured by adding a monoclonal antibody against TcdA, followed by an HRP-conjugated second antibody. The results of the standard curve are shown in Figure 13B. Based on these results, a standard curve using 125 ng/ml rTcdA was selected to determine the ABAB secretion level in yeast culture supernatant and was used in all subsequent ELISAs.

A shuttle plasmid (pD1214-FAKS) containing replication origins from both *Escherichia coli* (pUC) and yeast (2-micron loop) and the yeast auxotroph selection marker URA3 (conferring the ability to synthesize uracil) was obtained from DNA 2.0 (Newark, CA). Sequences encoding ABAB (SEQ ID NO: 20) at the N-terminus and the His tag (SEQ ID NO: 66) and the D7 tag (SEQ ID NO: 112) at the C-terminus of ABAB, respectively, were inserted into the backbone of this plasmid, wherein transcription is controlled by a strongly constitutive yeast translation elongation factor promoter ( _PTEF ) and extracellular secretion is provided by fusion with the α-mating factor secretion signaling leader sequence (FAKS). The sequence of the resulting plasmid (pD1214-FAKS-His-hABAB-D7) is provided in SEQ ID NO: 68.

The plasmid pD1214-FAKS-His-hABAB-D7 was transformed into *Saccharomyces cerevisiae* strain BY4741 (MATahis3Δ1leu2Δ0Met15Δ0ura3Δ0), a laboratory strain with URA3 knockout S288C derivative. The yeast transformants were then cultured overnight at 30°C and 250 rpm in uracil-free YNB medium containing exfoliating mixture ( ₁ L sterile ddH₂O, 6.8 g YNB, 20 g glucose, 2 g exfoliating mixture) on a shaker to reach OD₁. Cells were then centrifuged and lysed by sonication in 1X SDS loading buffer. After sonication, total cell lysates were treated at 98°C for 5 min and then loaded onto SDS gels. An equal volume of yeast control cell lysates was loaded into each well, except that control cells did not survive in uracil-free YNB medium and were therefore cultured in uracil-compatible YNB.

Cells were isolated by centrifuging culture supernatants from 25 yeast transformants and 3 yeast control colonies. The cell-free supernatants were then diluted 1:3 with 2.5% milk in PBS containing 0.05% Tween 20 and incubated in a shaker at 250 rpm and 30°C for 24 hours, followed by screening by ELISA as described above. Figure 14B shows that all yeast transformants secreted ABAB in the culture supernatant compared to the culture supernatant from the yeast control colonies.

The biological activity of secreted ABAB in culture supernatant was assessed using a cell-based neutralization assay. In this assay, a sufficient amount of toxin A or toxin B, which caused 100% cell rounding within 4 hours, was added to cell-free culture supernatant from either the BY4741 control colony or the BY4741-ABAB colony. Recombinant ABAB was used as a positive control. The biological activity of secreted ABAB in culture supernatant was determined by the level of neutralization activity that prevented cell rounding. Full-length ABAB secreted from *Saccharomyces cerevisiae* indeed maintained its neutralizing activity compared to purified recombinant ABAB (Figure 14A). These combined results suggest the plausibility of *Saccharomyces cerevisiae* ABAB secretion.

Further experiments demonstrated that oral gavage administration of Sc-ABAB at a dose of ^10¹⁰ CFU to mice had no adverse effects, and the detachment of live Sc-ABAB from mice was identified by coating feces onto Sabouraud CAF agar (data not shown). The isolate recovered from mice using the above assay retained its ability to produce functional ABAB.

ABAB secretion optimization

ABAB secretion levels are closely associated with in vivo therapeutic efficacy. Therefore, the possibility of further optimizing ABAB secretion by replacing the existing FAKS secretion signal with a number of commercially available secretion signals was explored. Prior to extracellular export, secretory sequences facilitate the co-translation or post-translational translocation of heterologous proteins into the endoplasmic reticulum and Golgi compartment. Although α-mating factors are commonly used signaling sequences for heterologous protein secretion, typically producing high yields of secreted proteins in Saccharomyces cerevisiae [69,70], studies have shown that other secretory sequences from other proteins, such as inulinase or invertase, may be better suited for the secretion of certain heterologous proteins [71,72].

Within the same pD1214 plasmid backbone, 11 different commercially available secretion signals (Table 4; DNA 2.0, Newark, CA) were fused to the ABAB gene under the control of the TEF promoter. Plasmids encoding ABAB with optional secretion signals include the following plasmids, wherein the FAKS secretion signal was replaced by the new secretion signals described in Table 4, and wherein the his tag and D7 tag were removed:

Plasmid pD1214-AKS-hABAB (SEQ ID NO: 70)

Plasmid pD1214-AK-hABAB (SEQ ID NO: 71)

Plasmid pD1214-AT-hABAB (SEQ ID NO: 72)

Plasmid pD1214-AA-hABAB (SEQ ID NO: 73)

Plasmid pD1214-GA-hABAB (SEQ ID NO: 74)

Plasmid pD1214-IN-hABAB (SEQ ID NO: 75)

Plasmid pD1214-IVS-hABAB (SEQ ID NO: 76)

Plasmid pD1214-KP-hABAB (SEQ ID NO: 77)

Plasmid pD1214-LZ-hABAB (SEQ ID NO: 78)

Plasmid pD1214-SA-hABAB (SEQ ID NO: 79)

In addition, the his and D7 tags were removed from the original ABAB construct (pD1214-FAKS-His-hABAB-D7) to generate plasmid pD1214-FAKS-hABAB (SEQ ID NO: 69), and the incubation temperature was increased to 37°C to better suit in vivo and clinical testing conditions. All 11 plasmids were then transformed into BY4741, and five independent colonies from each selected plate were selected to produce culture supernatant. The amount of secreted ABAB was determined by the same ELISA as described above. Furthermore, the E/O value was used to provide fair comparisons across all groups. The E/O value was defined by normalizing the ELISA O.D. value to the culture O.D. value. The two optimal secretion signals for ABAB were found to be AT and IVS (Table 4; Figure 15A).

Due to the lack of auxotrophic mutant strains of *Saccharomyces boulardii*, a 2µm-based plasmid carrying the aphA1 gene encoding resistance to G418 (PCEV-G4-Km; SEQ ID NO: 80; a gift from Lars Nielsen & Claudia Vickers (Addgene plasmid #46819)) was used instead of the pD1214 plasmid to confirm ABAB secretion in *Saccharomyces boulardii*. The two optimal secretion signals (AT and IVS) of *Saccharomyces cerevisiae* were fused to the ABAB gene and inserted into the backbone of the pCEV-G4-Km plasmid to generate plasmids pCEV-G4-Km-TEF-AT-hABAB* (SEQ ID NO: 81) and pCEV-G4-Km-TEF-IVS-hABAB* (SEQ ID NO: 82). ELISA confirmed that both plasmids were used to transform *Saccharomyces boulardii* (strain MYA796) and that ABAB secretion from AT and IVS in *Saccharomyces boulardii* was comparable to that in *Saccharomyces cerevisiae* (Figure 15B). Prepare another construct (pCEV-G4-Km-TEF-AT-hABAB(SEQ ID NO: 83)) that is different from pCEV-G4-Km-TEF-AT-hABAB* because it contains a molecular cloning site between the AT and hABAB sequences.

ABAB secretion was then further optimized by yeast codon optimization (yABAB) at the nucleotide level in the construct with the AT secretion signal, resulting in plasmid pCEV-G4-Km-TEF-AT-yABAB (SEQ ID NO: 84). A 40-nucleotide sequence between the _PTEF and ABAB coding sequences was also found to be unnecessary for ABAB secretion and was removed, resulting in plasmid pCEV-G4-Km-TEF-X40-AT-yABAB (SEQ ID NO: 85). Further sequences containing two restriction cloning sites between the AT and ABAB sequences were found to negatively impact ABAB secretion, and therefore this sequence was also omitted (plasmid pCEV-G4-Km-TEF- ^AT- yABAB; SEQ ID NO: 115) for subsequent studies to maximize ABAB secretion.

Next, the secretion of individual monomers was measured, revealing that AA6 secretion was the least. To improve AA6 secretion and thereby further optimize ABAB secretion, a set of key amino acid residues was used. The T83N mutation was found to improve AA6 secretion. Furthermore, strains of *Saccharomyces boulardii* carrying the hAA6 sequence were found to secrete more AA6 than strains carrying the yeast-optimized yAA6 sequence. Therefore, a comparison was made between ABAB carrying the T83N mutation in AA6 (AT-yABAB T83N; plasmid pCEV-G4-Km-TEF-AT-yABAB AA6T83N; SEQ ID NO: 116) and ABAB in which the yAA6 sequence is replaced by the hAA6 T83N sequence (AT-yABAB hAA6 T83N; plasmid pCEV-G4-Km-TEF-AT-yABAB hAA6T83N, which has the sequence of SEQ ID NO: 90 but lacks the coding sequence for c-Myc) to determine which sequence exhibited better secretion. These constructs showed no significant differences from AT-yABAB hAA6T83N, leading to the conclusion that the final sequence was advanced. The nucleotide sequence encoding AT-yABAB hAA6 T83N is provided in plasmid pCEV-G4-Km-TEF-AT-yABABhAA6T83N-tagless (SEQ ID NO: 90). The amino acid sequence of AT-yABAB hAA6 T83N is provided in SEQ ID NO: 117.

Production of auxotrophic Bula brasiliensis strains

Expression plasmids encoding ABAB can be cloned into *Bacillus blakeana* strains. *Bacillus blakeana* strains are more tolerant of normal body temperature and acidic conditions than *Saccharomyces cerevisiae*, which could enhance the efficacy as a novel oral yeast-based therapeutic strategy. Two modifications can be made to wild-type *Bacillus blakeana* strains to maintain the in vivo stability of the expression plasmid conferred by the yeast URA3 metabolic selection marker: 1) a diploid auxotrophic mutant carrying a deletion can be constructed in both chromosomal alleles of URA3, and 2) endogenous 2-micron loops can be cured in *Bacillus blakeana* to prevent accidental recombination from interfering with ABAB expression.

The most straightforward and efficient method for constructing auxotrophic mutants in wild-type yeast strains involves targeting the gene encoded by the deletion chromosome via homologous recombination, which occurs at a very high frequency in yeasts. Complete deletion of the target gene is superior to selecting a spontaneous mutation that can be restored to the wild type. Therefore, gene deletion is preferred for haploid states in *Saccharomyces cerevisiae*, which is typically induced from wild-type diploids by sporulation through incubation in a poorly nutrient-rich growth medium at low temperatures (30°C). However, *Bula* sporulation is insufficient and haploid cells are difficult to form under normal sporulation conditions [64,65]. A two-step approach using deleted chromosome gene alleles (e.g., URA3) is used, where each deletion step can be selected. This process is schematically outlined in Figure 16.

All chromosome deletions were performed via lithium acetate-assisted genetic transformation using linear DNA deletion cassettes [73]. Lithium acetate-based transformation originated from the Saccharomyces cerevisiae protocol and was found to be compatible with Blastomyces, although Blastomyces was found to be difficult to transform [55,56]. The difference was approximately 100-fold. Transformation efficiency in Saccharomyces cerevisiae can be improved by adjusting glucose concentration and heat shock time [74]. Therefore, various glucose concentrations and heat shock times were introduced into Blastomyces transformation for optimization. The optimal conditions for Blastomyces testing were a pre-culture glucose concentration of 2% and a heat shock time of 20 minutes at 42°C, and these conditions were used for all transformation processes in all studies.

Using pCEV-G4-Km (SEQ ID NO: 80) and pCEV-G4-Ph (SEQ ID NO: 86) (a gift from Lars Nielsen & Claudia Vickers (Addgene plasmid #46820)) as templates, two deletion cassettes containing the genes aphA1 and ble, respectively, conferring resistance to fulvic acid in G418 and yeast, were generated by PCR. Both deletion cassettes were laterally attached to two loci with X over P1 (loxP) in the same orientation, allowing the removal of antibiotic resistance genes using Cre recombinase. Homologous sequences of 40 base pairs upstream of the URA3 promoter ( _PURA3 ) and downstream of the URA3 stop codon were integrated into PCR primers to generate two final deletion cassettes for site-specific gene deletion in *Bula blazei* (see Figure 16). The exact sequence and location of the URA3 gene on chromosome V of *Bula blazei* were mapped using URA3 gene annotations from sequences published online in the Yeast Genome Database (SGD). The selection of a crossover 1, which replaces the first URA3 allele with an aphA1 deletion cassette, was used to utilize resistance to G418 [66]; the selection of a second crossover, which replaces the second URA3 allele with a ble deletion cassette, was used to utilize resistance to fumonisin [75] (Fig. 16). The replacement of the two URA3 alleles with aphA1 and ble deletion cassettes was demonstrated by resistance to both antibiotics (data not shown) and by the lack of growth on minimal synthetic medium plates lacking uracil (data not shown). Growth on Sabouraud plates with chloramphenicol (100 ug/ml) also confirmed the yeast phenotype (data not shown). In addition, three sets of unique primers targeting the URA3, aphA1, or ble genes in the URA3 chromosomal region were designed, and PCR was performed using wild-type (WT), URA3Δ::aphA/URA3 (first crossover), and URA3Δ::aphA1/Δ::ble (second crossover) genomic DNA as templates. The expected PCR product sizes targeting the URA3, aphA1, or ble genes in the URA3 chromosomal region were 766 bp, 1183 bp, and 662 bp, respectively. Using these three unique primer sets, DNA electrophoresis of the PCR products from WT, first crossover, and second crossover clones confirmed the deletion of the URA3 allele and the integration of the aphA1 and ble deletion cassettes in the second crossover clone.

The second cross strain was then transformed with pPL5071_TEF1-Cre_URA3 (pPL5071; SEQ ID NO: 95)[76] to remove the aphA1 and ble deletion cassettes. The strain carried pPL5071 constitutively expressed Cre recombinase under P _TEF Cre recombinase and then targeted the loxp sequences flanking the aphA1 and ble deletion cassettes; this resulted in the excision of the aphA1 and ble deletion cassettes, leaving only one loxp site in the URA3 chromosomal region. The strain that underwent successful excision of the aphA1 and ble deletion cassettes could not grow in the presence of G418 or fumonisin; however, the loss of two URA3 alleles was still retained, so it could only grow on minimal synthetic medium plates in the presence of uracil and did not show growth on minimal synthetic medium plates without uracil supplementation.

pPL5071 removal was achieved by growing in YPD and selecting colonies that would later grow on a minimal synthetic medium containing uracil and the pyrimidine analog 5-fluoro-orotic acid (5-FOA)[77]. Strains possessing pPL5071 carry the URA3 gene, which synthesizes the toxic intermediate 5-fluorodeoxyuridine, a potent inhibitor of thymidylate synthase, which interrupts DNA synthesis and leads to cell death, allowing selection of strains that have lost pPL5071. The absence of pPL5071 was also confirmed by PCR and DNA electrophoresis of the PCR products using pPL5071-specific primers.

The 2 μm plasmid is a very stable 6.1 kb plasmid that is ubiquitous in yeast strains. This plasmid has no selective advantage over yeast host organisms and is significantly stable due to the presence of an efficient REP1-REP2-STB plasmid distribution system [68]. The *Bula* strain used also contained this plasmid, which was confirmed by PCR. To remove the 2 μm plasmid, the 2 μm plasmid was immobilized with pBIS-GALkFLP-URA3 (SEQ ID NO: 87) [67] and then removed with uracil and 5-FOA. The loss of the 2 μm plasmid was confirmed by PCR using primers specific to the origin of replication.

The auxotrophic strain of *Bula sacchariformis* produced by these operations was named *Bula sacchariformis* URA3Δ/Δ.

Auxotrophic Saccharomyces boulardii strains for in situ ABAB delivery

To construct a auxotrophic *Bacillus brasiliensis* strain for in situ ABAB delivery, the aphA1 cassette of plasmid pCEV-G4-Km-TEF-X40-AT-yABAB (SEQ ID NO: 85) was replaced with the URA3 cassette from the pD plasmid to generate plasmid pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88). This plasmid was then used to convert *Bacillus brasiliensis* URA3Δ/Δ. In a cytotoxicity assay, the resulting strain secreted fully functional ABAB compared to purified ABAB (Fig. 17C). Western blotting using an HRP-conjugated α-Llama antibody showed the corresponding ABAB band from *Bacillus brasiliensis* culture supernatant (Fig. 17D). The C-terminus of the ABAB contained a c-Myc tag and could be further pulled down by the α-c-Myc antibody (Fig. 17D).

For the empty plasmid (EP) control, the AT-yABAB sequence was later removed from pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88) to generate pCEV-URA3-TEF-cMyc (SEQ ID NO: 89). The *Saccharomyces boulardii* URA3Δ/Δ strain transformed with this plasmid produced a strain complementary to URA3 but not secreting ABAB. The *Saccharomyces boulardii* URA3Δ/Δ strain secreting ABAB also showed no growth inhibition when cultured in YPD containing vancomycin (1 mg/ml) (Fig. 17A). This indicates that *Saccharomyces boulardii* can be co-administered with vancomycin, commonly used to treat CDI patients, and secrete ABAB to treat ongoing CDI. Furthermore, the stability of ABAB purified from the culture supernatant collected from *Saccharomyces boulardii* at an O.D. of 10 for more than 2 hours suggests that the secreted ABAB may diffuse out of *Saccharomyces boulardii* without degradation.

Oral safety assessment of Saccharomyces boulardii in mice treated with antibiotics

Before assessing whether ABAB-expressing *Saccharomyces boulardii* URA3Δ/Δ could protect mice in a CDI model [20,33,62,78], a safety assessment was conducted to determine the safe dose of *Saccharomyces boulardii* in antibiotic-treated mice. In this safety assessment, mice were first given an antibiotic mixture in their daily drinking water for three days, then switched to regular water. The day before oral administration of *Saccharomyces boulardii*, mice were intraperitoneally injected with clindamycin. This completed antibiotic treatment of the mice, and then *Saccharomyces boulardii* was delivered orally to the mice for safety assessment, which included monitoring daily weight changes in these antibiotic-treated mice and the persistence of *Saccharomyces boulardii* in fecal samples. The mice showed no signs of disease and stable weight gain during the 6-day monitoring period, with oral administration of 10^¹⁰ cells consistent with the idea of *Saccharomyces boulardii* as a GRAS organism. However, for the subsequent CDI mouse study, only 10 ^⁹ *Saccharomyces boulardii* cells were administered due to the ease of particle resuspension and the small dose variation in mice, which may occur when high viscosity is present in the resuspension. Bula spp. also showed limited colonization in the gastrointestinal tract of mice treated with these antibiotics; Bula spp. was not detected on Sabouroud plates three days after the last gavage (data not shown).

Protective effect of ABAB-expressing Saccharomyces boulardii against primary CDI in mice

The protective effect of ABAB-expressing *Bacillus blakeana* was evaluated using an established primary mouse model of CDI. *Bacillus blakeana* expressing ABAB was delivered as a preventative or therapeutic agent against primary CDI in mice. Briefly, primary CDI was established in mice by supplementing their drinking water with an antibiotic mixture for three days, followed by intraperitoneal injection of clindamycin 24 hours prior to *Clostridium difficile* spore challenge. CDI was induced by gavage of ^{10⁵ *} Clostridium difficile* spores (UK1,027/BI/NAP1 prevalent strain). For prophylactic evaluation, mice were given oral doses of *Bacillus blakeana* starting the day after switching to regular drinking water, administered daily for 7 days. For therapeutic evaluation, mice were given oral doses of *Bacillus blakeana* at 6, 24, 48, and 72 hours post-spore challenge. Controls included PBS and *Bacillus blakeana* transformed with an empty plasmid. In both approaches, mice receiving *Bula bladderwrack* expressing ABAB were significantly protected from CDI-induced death (Figs. 18A and 19A; PBS; negative control; Sb:EP: *Bula bladderwrack* transformed with empty plasmid; Sb:BAB: *Bula bladderwrack* secreting ABAB). CDI mice typically suffered weight loss, with most of the weight loss occurring on days 2 through 3 due to diarrhea and gradually recovering. Mice receiving *Bula bladderwrack* expressing ABAB recovered weight rapidly (Figs. 18B and 19B), and the percentage of diarrheal events on day 2 post-challenge was significantly reduced (Figs. 18C and 19C).

Protective effect of ABAB-expressing Saccharomyces boulardii against recurrent chronic dysplasia (CDI) in mice.

To evaluate the protective effect of ABAB-expressing *Bacillus blakeana* against recurrent CDI in mice. To induce recurrent CDI, mice were given an antibiotic mixture in their daily drinking water for three days. After three days of antibiotic water, the mice were then switched back to drinking normal water. One day before oral administration of ^{10⁵ *} Clostridium difficile* spores (UK1, 027/BI/NAP1 epidemic strain), mice were intraperitoneally injected with clindamycin. Six hours after spore challenge, the normal water was replaced with water containing 0.5 mg/ml vancomycin for six days, and then switched back to normal water for the remainder of the study. Four days after vancomycin discontinuation, mice typically developed signs of CDI without further *Clostridium difficile* spore challenges. During the relapse model, *Bacillus blakeana* was orally administered once daily with vancomycin water for 12 days. This model was used to evaluate the protective efficacy of ABAB-expressing *Bacillus blakeana* against CDI relapse in mice. Survival, weight loss, and diarrhea events in these mice were monitored daily. Controls included PBS and *Bacillus blakeana* transformed with an empty plasmid. Mice receiving *Bula bladderwrack* expressing ABAB were significantly protected against relapse-induced CDI death (Fig. 20A; PBS; negative control; Sb:EP: *Bula bladderwrack* transformed with empty plasmid; Sb:BAB: *Bula bladderwrack* secreting ABAB). Similar to mice with primary CDI, mice with relapsed CDI typically experienced weight loss around day 4 to day 5 after vancomycin water withdrawal. Mice receiving *Bula bladderwrack* expressing ABAB were significantly protected from weight loss (Fig. 20B) and had a significantly reduced percentage of diarrhea relapse events (Fig. 20C).

Stability optimization through chromosome integration of ABAB boxes

Recent studies in *Saccharomyces cerevisiae* and *Bula sacchari* have demonstrated the use of CRISPR-Cas9-based systems for genome editing [79-81]. Furthermore, large deletions can be achieved by simultaneously targeting two guide sequences [82]. When there is no selection pressure, foreign genes are generally more stably maintained when integrated into the chromosome compared to introduction via plasmids. However, chromosomal integration typically requires multiple rounds to achieve high copy numbers. A recently published protocol overcomes this obstacle by simultaneously integrating large fragments at multiple copies of common sequences in the *Saccharomyces cerevisiae* genome, such as δ sites, through CRISPR-induced double staining disruption [83]. DNA double-strand breaks can be repaired by non-homologous end joining or homologous recombination; however, when endogenous homologous sequences are present, the host preferentially uses homologous sequences to repair DNA double-strand breaks via homologous recombination [83].

The δ site is a long terminal repeat (LTR) sequence belonging to Ty elements I and II, and is the most abundant LTR in Saccharomyces cerevisiae. There are five types of Ty elements (1-5), represented by type II transposons (retrotransposons), which are more common in Saccharomyces cerevisiae. It is estimated that there are about 51 retrotransposons (Ty1-5) and 251 δ sites in the Saccharomyces cerevisiae genome [84]. These δ sites are target sequences that attract the integration of ABAB expression cassettes into the Saccharomyces bulagen chromosome. However, much less is known about δ sites in Saccharomyces bulagen. Therefore, Ty1-H3 (Genbank accession number M18706) [84] was first used as a probe to investigate Ty1-2 elements in Saccharomyces bulagen strain MYA796 (ATCC, Manassas, VA) (from the NCBI genome draft) to identify possible Ty1-2 elements and their δ sites in the Saccharomyces bulagen genome. Surprisingly, no complete Ty1-2 elements were found in MYA796. A total of 57 δ sites were identified; these include 44 complete δ sites and 12 partial sites, as well as partial Ty elements containing one complete δ site identified in all 16 chromosomes (Table 12).

Table 12 - Number and distribution of δ loci on chromosome MYA796

/>

Due to the diploid state of *Bula capillaris*, the *Bula capillaris* genome contains approximately 114 δ-sites. To allow for simple multiplex δ-site targeting via CRISPR, all 57 δ-site sequences were compiled into multiple sequence alignments using MUSCLE to identify a large number of prototypical spacer adjacent motif (PAM) sites within the 57 δ-sequences. Two PAM sites were selected as upstream and downstream sequences based on the highest uniform number of δ-sequences within the prototypical spacers. The sequences of these PAM sites are shown in Figure 21, and the specific sequences are as follows:

PAM site I

PAM site II

In Pam site I and Pam site II, the sequence underlined by the dashed line corresponds to the upstream homologous sequence; the sequence underlined by the single line corresponds to the 20 bp prototype spacer; the sequence underlined by the double line corresponds to the PAM sequence; and the sequence underlined by the wavy line corresponds to the downstream homologous sequence.

These two PAM sites, along with their shared upstream and downstream homologous sequences within the δ site, allow for the simple chromosomal integration of the ABAB expression cassette into the *Blavus* genome. An ABAB integration cassette containing homologous recombination sequences was generated by PCR using primers comprising an upstream homologous sequence with the last three nucleotides removed from the 3' end and a downstream homologous sequence with the first two nucleotides removed from the 5' end, along with the corresponding annealing sequences required for PCR using plasmid pCEV-G4-Km-TEF-AT-yABAB hAA6T83N-tagless as a template (SEQ ID NO: 90).

PCR products from ABAB integration cassettes containing CRISPR plasmids, including the corresponding guide sequences (pCRI-Sb-δ1 (SEQ ID NO: 91) and pCRI-Sb-δ2 (SEQ ID NO: 92)), were co-transformed with *Saccharomyces boulardii* for ABAB integration into the chromosome, independently and sequentially targeting PAM sites I and II. The ratio of PCR product to CRISPR plasmid was found to be important for generating successful integrated clones (Fig. 22A; ^low ITG to ^high ITG). Furthermore, repeated transformations of the highest ABAB-secreting clone from ^{the high} ITG group with the same integration cassette and CRISPR plasmid did not further improve overall ABAB secretion from independent clones (Fig. 22A; ^{2nd high} ITG). ABAB secretion from the highest ABAB-secreting clone ( ^CRISPR -2) from ^{the high} ITG group was further enhanced by co-transformation of a second set of ABAB integration cassettes containing homologous recombination sequences and their corresponding guide sequences at CRISPR plasmid targeting site II (Fig. 22A). Two of the highest ABAB-secreting clones, ^CRISPR -3 and ^CRISPR -4, were selected. Figure 22B shows the ABAB secretion levels and stability over time for these four representative clones. Preliminary mouse CDI studies were conducted. However, ^CRISPR -4 was found to be no better than the previous M-/-:ABAB clones, and it showed a protective effect in many mouse CDI models (Figure 23).

While the invention has been described with reference to certain specific embodiments, those skilled in the art will understand that various modifications can be made without departing from the spirit and scope of the invention. The scope of the appended claims is not limited to the specific embodiments described.

References

All patents and publications mentioned in this specification indicate the level of skill of a person skilled in the art to which this invention pertains. Each cited patent and publication is incorporated herein by reference in its entirety. All of the following references are cited in this application:

1. Cloud, J. & Kelly, C.P. Update on Clostridium difficile associated disease. Curr Opin Gastroenterol 23, 4-9 (2007).

2.Kelly,C.P.&LaMont,J.T.Clostridium difficile--more difficult thanever.N Engl J Med 359,1932-1940(2008).

3. Barbut, F., et al. Epidemiology of recurrences or reinfections of Clostridium difficile-associated diarrhea. J Clin Microbiol 38, 2386-2388 (2000).

4. Hassoun, A. & Ibrahim, F. Use of intravenous immunoglobulin for the treatment of severe Clostridium difficile colitis. Am J Geriatr Pharmacother 5, 48-51 (2007).

5. Shahani, L. & Koirala, J. Intravenous immunoglobulin in treatment of Clostridium difficile colitis. BMJ Case Rep 2012(2012).

6.Saito,T.,et al.Evidence of intravenous immunoglobulin as a criticalsupportive therapy against Clostridium difficile toxin-mediated lethality inmice.J Antimicrob Chemother 66,1096-1099(2011).

7. Abougergi, M.S. & Kwon, J.H. Intravenous immunoglobulin for the treatment of Clostridium difficile infection: a review. Dig Dis Sci 56, 19-26 (2011).

8. Sokol, H., Maury, E., Seksik, P., Cosnes, J. & Beaugerie, L. Single immunoglobulin infusion can reverse hemodynamic failure associated with severe Clostridium difficile colitis. Am J Gastroenterol 104, 2649-2650 (2009).

9. Lowy, I., et al. Treatment with monoclonal antibodies against Clostridium difficile toxins. N Engl J Med 362, 197-205 (2010).

10. Louie, T.J., et al. Fidaxomicin versus vancomycin for Clostridiumdifficile infection. The New England journal of medicine 364, 422-431 (2011).

11. Rao, K. & Young, V. B. Fecal Microbiota Transplantation for the Management of Clostridium difficile Infection. Infectious disease clinics of North America 29, 109-122 (2015).

12. Vyas, D., Aekka, A. & Vyas, A. Fecal transplant policy and legislation. World journal of gastroenterology: WJG 21, 6-11 (2015).

13. Tonna, I. & Welsby, P.D. Pathogenesis and treatment of Clostridiumdifficile infection. Postgrad Med J 81, 367-369 (2005).

14. McFarland, L.V., Elmer, G.W. & Surawicz, C.M. Breaking the cycle: treatment strategies for 163 cases of recurrent Clostridium difficiledisease. The American journal of gastroenterology 97, 1769-1775 (2002).

15. O'Neill, G.L., Beaman, M.H. & Riley, T.V. Relapse versus reinfection with Clostridium difficile. Epidemiol Infect 107, 627-635 (1991).

16.Wilcox,M.H.,Fawley,W.N.,Settle,C.D.&Davidson,A.Recurrence ofsymptoms in Clostridium difficile infection--relapse or reinfection? J HospInfect 38,93-100(1998).

17. Johnson, S., Adelmann, A., Clabots, C.R., Peterson, L.R. & Gerding, D.N. Recurrences of Clostridium difficile diarrhea not caused by the originalinfecting organism. J Infect Dis 159, 340-343 (1989).

18. Tang-Feldman, Y., Mayo, S., Silva Jr, J., Jr. & Cohen, S.H. Molecular analysis of Clostridium difficile strains isolated from 18 cases of recurrent clostridium difficile-associated diarrhea. J Clin Microbiol 41, 3413-3414 (2003).

19. Johnson, S. Recurrent Clostridium difficile infection: causality and therapeutic approaches. Int J Antimicrob Agents 33 Suppl 1, S33-36 (2009).

20.Sun,X.,et al.Mouse relapse model of Clostridium difficileinfection.Infection and immunity 79,2856-2864(2011).

21. Seal, D., et al. Treatment of relapsing Clostridium difficilediarrhoea by administration of a non-toxigenic strain. Eur J Clin Microbiol 6, 51-53 (1987).

22.Kuehne,S.A.,et al.The role of toxin A and toxin B in Clostridiumdifficile infection.Nature(2010).

23.Lyras,D.,et al.Toxin B is essential for virulence of Clostridiumdifficile.Nature 458,1176-1179(2009).

24. Libby, J. M. & Wilkins, T. D. Production of antitoxins to two toxins of Clostridium difficile and immunological comparison of the toxins by cross-neutralization studies. Infection and immunity 35, 374-376 (1982).

25. Katchar, K., et al. Association between IgG2 and IgG3 subclass responses to toxin A and recurrent Clostridium difficile-associateddisease. Clinical gastroenterology and hepatology: the official clinical practice journal of the American Gastroenterological Association 5,707-713(2007).

26.Ghose, C., et al. Transcutaneous immunization with Clostridiumdifficile toxoid A induces systemic and mucosal immune responses and toxin A-neutralizing antibodies in mice. Infection and immunity 75, 2826-2832 (2007).

27. Sougioultzis, S., et al. Clostridium difficile toxoid vaccine inrecurrent C. difficile-associated diarrhea. Gastroenterology 128, 764-770 (2005).

28. Kyne, L., Warny, M., Qamar, A. & Kelly, C.P. Association between antibody response to toxin A and protection against recurrent Clostridium difficilediarrhoea. Lancet 357, 189-193 (2001).

29. Leav, B.A., et al. Serum anti-toxin B antibody correlates with protection from recurrent Clostridium difficile infection (CDI). Vaccine 28, 965-969 (2010).

30.Fernie,D.S.,Thomson,R.O.,Batty,I.&Walker,P.D.Active and passiveimmunization to protect against antibiotic associated caecitis inhamsters.Dev Biol Stand 53,325-332(1983).

31.Kim,P.H.,Iaconis,J.P.&Rolfe,R.D.Immunization of adult hamstersagainst Clostridium difficile-associated ileocecitis and transfer ofprotection to infant hamsters.Infection and immunity 55,2984-2992(1987).

32. Lyerly, D.M., Bostwick, E.F., Binion, S.B. & Wilkins, T.D. Passiveimmunization of hamsters against disease caused by Clostridium difficile by use of bovine immunoglobulin G concentrate. Infection and immunity 59, 2215-2218 (1991).

33. Wang, H., et al. A chimeric toxin vaccine protects against primary and recurrent Clostridium difficile infection. Infection and immunity 80, 2678-2688 (2012).

34. Hamers-Casterman, C., et al. Naturally occurring antibodies devoid of light chains. Nature 363, 446-448 (1993).

35.van der Linden,R.,et al.Induction of immune responses andmolecular cloning of the heavy chain antibody repertoire of Lama glama.JImmunol Methods 240,185-195(2000).

36. Harmsen, M.M. & De Haard, H.J. Properties, production, and applications of camelid single-domain antibody fragments. Appl Microbiol Biotechnol 77, 13-22 (2007).

37. Arbabi Ghahroudi, M., Desmyter, A., Wyns, L., Hamers, R. & Muyldermans, S. Selection and identification of single domain antibody fragments from camelheavy-chain antibodies. FEBS Lett 414, 521-526 (1997).

38. van der Linden, R.H., et al. Comparison of physical chemical properties of llama VHH antibody fragments and mouse monoclonalantibodies. Biochim Biophys Acta 1431, 37-46 (1999).

39. Dumoulin, M., et al. Single-domain antibody fragments with high conformational stability. Protein Sci 11, 500-515 (2002).

40. Hussack, G., Hirama, T., Ding, W., Mackenzie, R. & Tanha, J. Engineered single-domain antibodies with high protease resistance and thermalstability. PloS one 6, e28218 (2011).

41. Yang, Z., et al. A novel multivalent, single-domain antibody targeting TcdA and TcdB prevents fulminant Clostridium difficile infection in mice. The Journal of infectious diseases 210, 964-972 (2014).

42.McFarland, L.V. Systematic review and meta-analysis of Saccharomycesboulardii in adult patients. World J. Gastroenterol. 16, 2202-2222 (2010).

43.Elmer, G.W., McFarland, L.V., Surawicz, C.M., Danko, L. & Greenberg, R.N. Behavior of Saccharomyces boulardii in recurrent Clostridium difficiledisease patients. Aliment. Pharmacol. Ther. 13, 1663-1668 (1999).

44.Surawicz,C.M.,et al.The search for a better treatment forrecurrent Clostridium difficile disease:use of high-dose vancomycin combined with Saccharomyces boulardii.Clin.Infect.Dis.31,1012-1017(2000).

45. McFarland, L.V., et al. A randomized placebo-controlled trial of Saccharomyces boulardii in combination with standard antibiotics for Clostridium difficile disease. Jama 271, 1913-1918 (1994).

46.Czerucka,D.,Piche,T.&Rampal,P.Review article:yeast as probiotics--Saccharomyces boulardii.Alimentary pharmacology&therapeutics 26,767-778(2007).

47. Sougioultzis, S., et al. Saccharomyces boulardii produces a soluble anti-inflammatory factor that inhibits NF-kappaB-mediated IL-8 gene expression. Biochemical and biophysical research communications 343, 69-76 (2006).

48. Qamar, A., et al. Saccharomyces boulardii stimulates intestinal immunoglobulin A immune response to Clostridium difficile toxin A inmice. Infection and immunity 69, 2762-2765 (2001).

49. Chen,

50. Barc, M.C., et al. Molecular analysis of the digestive microbiota in a gnotobiotic mouse model during antibiotic treatment: Influence of Saccharomyces boulardii. Anaerobe 14, 229-233 (2008).

51. Gorlani, A., de Haard, H. & Verrips, T. Expression of VHHs in Saccharomyces cerevisiae. Methods in molecular biology 911, 277-286 (2012).

52. Khatri, I., et al. Gleaning evolutionary insights from the genome sequence of a probiotic yeast Saccharomyces boulardii. Gut Pathog. 5, 30 (2013).

53. Batista, T. M., Marques, E. T., Jr., Franco, G. R. & Douradinha, B. Draft Genome Sequence of the Probiotic Yeast Saccharomyces cerevisiae var. boulardiiStrain ATCC MYA-796. Genome announcements 2, 578-579 (2014).

54. Hamedi, H., et al. Generation of a uracil auxotroph strain of the probiotic yeast Saccharomyces boulardii as a host for the recombinant protein production. Avicenna journal of medical biotechnology 5, 29-34 (2013).

55.Hudson, L.E., et al. Functional heterologous protein expression by genetically engineered probiotic yeast Saccharomyces boulardii. PLoS One 9, e112660 (2014).

56.Douradinha,B.,et al.Novel insights in genetic transformation of the probiotic yeast Saccharomyces boulardii.Bioengineered 5,21-29(2014).

57. Fietto, J.L., et al. Molecular and physiological comparisons between Saccharomyces cerevisiae and Saccharomyces boulardii. Canadian journal of microbiology 50, 615-621 (2004).

58.Martins,F.S.,et al.Screening of yeasts as probiotic based oncapacities to colonize the gastrointestinal tract and to protect against enteropathogen challenge in mice.The Journal of general and appliedmicrobiology 51,83-92(2005).

59. Martins, F.S., Veloso, L.C., Arantes, R.M. & Nicoli, J.R. Effects of yeastprobiotic formulation on viability, revival and protection against infection with Salmonella enterica ssp.enterica serovar Typhimurium inmice. Lett. Appl. Microbiol. 49, 738-744 (2009).

60. Wu, D., Teng, D., Wang, X., Dai, C. & Wang, J. Saccharomyces boulardiiprevention of the hepatic injury induced by Salmonella Enteritidisinfection. Can.

61.Elmer,G.W.&Corthier,G.Modulation of Clostridium difficile inducedmortality as a function of the dose and the viability of the Saccharomycesboulardii used as a preventative agent in gnotobioticmice.Can.J.Microbiol.37,315-317(1991).

62.Chen,X.,et al.A mouse model of Clostridium difficile-associateddisease.Gastroenterology 135,1984-1992(2008).

63.Pawlowski,S.W.,et al.Murine model of Clostridium difficileinfection with aged gnotobiotic C57BL/6 mice and a BI/NAP1 strain.The Journal of infectious diseases 202,1708-1712(2010).

64. McFarland, L.V. Saccharomyces boulardii is not Saccharomyces cerevisiae. Clin. Infect. Dis. 22, 200-201 (1996).

65. Edwards-Ingram, L., et al. Genotypic and physiological characterization of Saccharomyces boulardii, the probiotic strain of Saccharomyces cerevisiae. Appl. Environ. Microbiol. 73, 2458-2467 (2007).

66. Panchal, C.J., Whitney, G.K. & Stewart, G.G. Susceptibility of Saccharomyces spp. and Schwanniomyces spp. to the aminoglycoside antibiotic G418. Appl. Environ. Microbiol. 47, 1164-1166 (1984).

67. Tsalik, E.L. & Gartenberg, M.R. Curing Saccharomyces cerevisiae of the2 micron plasmid by targeted DNA damage. Yeast 14, 847-852 (1998).

68. Chan, K. M., Liu, Y. T., Ma, C. H., Jayaram, M. & Sau, S. The 2 micron plasmid of Saccharomyces cerevisiae: a miniaturized selfish genome with optimized functional competence. Plasmid 70, 2-17 (2013).

69.Brake, A.J., et al. Alpha-factor-directed synthesis and secretion ofmature foreign proteins in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America 81, 4642-4646 (1984).

70. Porro, D., Sauer, M., Branduardi, P. & Mattanovich, D. Recombinant protein production in yeasts. Molecular biotechnology 31, 245-259 (2005).

71. Chung, B.H., Nam, S.W., Kim, B.M. & Park, Y.H. Highly efficient secretion of heterologous proteins from Saccharomyces cerevisiae using inulinase signalpeptides. Biotechnology and bioengineering 49, 473-479 (1996).

72. Hahm, M. S. & Chung, B. H. Secretory expression of human growth hormone in Saccharomyces cerevisiae using three different leaders sequences. Biotechnology and Bioprocess Engineering 6, 306-309 (2001).

73. Gietz, R.D. & Schiestl, R.H. High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method. Nat. Protoc. 2, 31–4 (2007).

74. Konishi, T. & Harata, M. Improvement of the transformation efficiency of Sacchaaromyces cerevisiae by altering carbon sources in pre-culture. Biosci. Biotechnol. Biochem. 78, 1090–1093 (2014).

75. Gatignol, A., Baron, M. & Tiraby, G. Phleomycin resistance encoded by theble gene from transposon Tn 5 as a dominant selectable marker in Saccharomyces cerevisiae. MGG Mol. Gen. Genet. 207, 342–348 (1987).

76. Macdonald, C. & Piper, R. C. Puromycin-and methotrexate-resistancecassettes and optimized Cre-recombinase expression plasmids for use inyeast. Yeast 32, 423–438 (2015).

77. Boeke, J.D., LaCroute, F. & Fink, G.R. A positive selection for mutantslacking orotidine-5’-phosphate decarboxylase activity in yeast: 5-fluoro-orotic acid resistance. Mol Gen Genet 197, 345–346 (1984).

78. Yang, Z. et al. Intravenous adenovirus expressing a multi-specific, single-domain antibody neutralizing TcdA and TcdB protects mice from Clostridium difficile infection. Pathog. Dis. 74, ftw078 (2016).

79. DiCarlo, J.E. et al. Genome engineering in Saccharomyces cerevisiaeusing CRISPR-Cas systems. Nucleic Acids Res. 41, 4336–43 (2013).

80. Bao, Z. et al. Homology-Integrated CRISPR-Cas (HI-CRISPR) System for One-Step Multigene Disruption in Saccharomyces cerevisiae. ACS Synth. Biol. 4, 585–594 (2015).

81.Liu,J.-J.et al.Metabolic engineering of a probiotic Saccharomycesboulardii.Appl.Environ.Microbiol.AEM.00057-16(2016).doi:10.1128/AEM.00057-16

82.Hao,H.et al.Large fragment deletion using a CRISPR/Cas9 system inSaccharomyces cerevisiae.(2016).doi:10.1016/j.ab.2016.07.008.This

83. Shi, S., Liang, Y., Zhang, M. M., Ang, E. L. & Zhao, H. A highly efficient single-step, markerless strategy for multi-copy chromosomal integration of large biochemical pathways in Saccharomyces cerevisiae. Metab. Eng. 33, 19–27 (2016).

84. Kim, J.M. et al. Transposable Elements and Genome Organization: A Comprehensive Survey of Retrotransposons Revealed by the Complete Saccharomyces cerevisiae. Genome Research. 8(5):464–478(1998).

85.Bao Z, Xiao H, Liang J, Zhang L, Xiong

86. DiCarlo JE, Norville JE, Mali P, Rios X, Aach J, Church GM. Genomeengineering in Saccharomyces cerevisiae using CRISPR-Cas systems. NucleicAcids Res 41:4336–43(2013).

87.Liu J-J, Kong II, Zhang G-C, Jayakody LN, Kim H, Xia P-F, Kwak S, SungBH, Sohn J-H, Walukiewicz HE, Rao C V., Jin Y-S. Metabolic engineering of aprobiotic Saccharomyces boulardii.Appl Environ Microbiol.82(8):2280-7(2016).

88. Shi S, Liang Y, Zhang MM, Ang EL, Zhao H. A highly efficient single-step, markerless strategy for multi-copy chromosomal integration of largebiochemical pathways in Saccharomyces cerevisiae. Metab Eng 33:19–27(2016).

Claims (10)

1. A yeast engineered strain that produces a tetraspecific tetramer binder, wherein the binder comprises linked first, second, third, and fourth _VHH peptide monomers, and wherein the _VHH peptide monomers independently have binding specificity to epitopes of Clostridium difficile toxin A (TcdA) or toxin B (TcdB). 2. The yeast engineered strain according to claim 1, wherein the first, second, third and fourth _VHH peptide monomers each have binding specificity for different epitopes. 3. The yeast engineered strain according to claim 1, wherein the two _VHH peptide monomers have binding specificity to the epitope of TcdA, and the two _VHH peptide monomers have binding specificity to the epitope of TcdB. 4. The yeast engineered strain according to claim 1, wherein the _VHH peptide monomer independently has binding specificity to epitopes in the glucosyltransferase domain, cysteine protease domain, translocation domain, or receptor-binding domain of TcdA or TcdB. 5. The engineered yeast strain according to claim 1, The binding agent is ABAB, wherein the first and third monomers are specific for binding to the epitopes of TcdA, and the first and third monomers are _VHH peptide monomers AH3 (SEQ ID NO: 7) and AA6 (SEQ ID NO: 5), respectively. The second and fourth monomers are specifically binding to the epitopes of TcdB. The second and fourth monomers are _VHH peptide monomers 5D (SEQ ID NO: 1) and E3 (SEQ ID NO: 3), respectively. 6. The yeast engineered strain according to claim 5, wherein the ABAB binding agent comprises the amino acid sequence shown in SEQ ID NO: 19, or a sequence variant having at least 95% sequence identity with it, and wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity, or both. 7. The yeast engineered strain according to claim 5, wherein the ABAB binding agent further comprises an N-terminal secretion signal selected from AT secretion signal (MRFPSIFTAVLFAASSALA (SEQ ID NO: 99)) and IVS secretion signal (MLLQAFLFLLAGFAAKISA (SEQ ID NO: 103)). 8. The yeast engineered strain according to claim 5, wherein the ABAB binding agent is expressed by a plasmid within the yeast, wherein the ABAB binding agent comprises the amino acid sequence shown in SEQ ID NO: 107, or a sequence variant having at least 95% sequence identity therewith, and wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity, or both. 9. The yeast strain according to claim 8, wherein the plasmid is pCEV-URA3-TEF-AT-yABAB-cMyc (SEQ ID NO: 88). 10. The yeast engineered strain of claim 5, wherein the ABAB binding agent coding sequence is integrated into the yeast chromosome, wherein the ABAB binding agent comprises the amino acid sequence shown in SEQ ID NO: 109, or a sequence variant having at least 95% sequence identity with it, and wherein the sequence variant retains TcdA and/or TcdB binding specificity, or the sequence variant retains toxin neutralizing activity, or both.