Metabolically engineered paenibacillus for improved production of colistin

WO2025176169A8PCT designated stage Publication Date: 2025-09-25VERSITECH LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2025/078278
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2024-04-08
Filing Date
2025-02-20
Publication Date
2025-09-25

AI Technical Summary

Technical Problem

The production of colistin is challenging due to the presence of impurities from endogenous secondary metabolites and limited production potential in native producers, and current metabolic engineering methods lack a comprehensive approach considering metabolic flux, precursor supply, and gene expression regulation.

Method used

Genetically engineered Paenibacillus bacteria with inactivated non-essential biosynthetic gene clusters and increased expression of L-DABA pathway genes, using promoters like PH4, to enhance colistin production and purity.

Benefits of technology

The mutant bacteria produce colistin with titers up to 2000 mg/L and a clean metabolite background, surpassing the production of other polymyxins and achieving high purity.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

Colistin is an important last-resort antibiotic, but its production involves significant challenges due to the presence of impurities resulting from various analogs and the limited number of native producers. In response, disclosed are mutant Paenibacillus bacteria that produce Colicin A more efficiently and / or Colicin A that is of higher purity and / or has a cleaner metabolite background. The disclosed mutant bacteria are genetically engineered to inactivate expression of one or more biosynthetic gene clusters (BGCs), increase expression of one or more other BCGs, and / or increase expression of one or more L-DABA pathway genes.
Need to check novelty before this filing date? Find Prior Art

Description

METABOLICALLY ENGINEERED PAENIBA CILL US FOR IMPROVED PRODUCTION OF COLISTINFIELD OF THE INVENTION

[0001] The disclosed invention is generally in the field of metabolically engineered bacteria and specifically in the area of metabolically engineering bacteria for improved antibiotic production.

[0002] DEPOSIT OF BACTERIAL SPECIMENS

[0003] The bacterial specimen Paenibacillus polyxma strain P14 was deposited by The University of Hong Kong under terms in accordance with the Budapest Treaty with the China General Microbiological Culture Collection Center ( “CGMCC” ) , having an address of Institute of Microbiology Chinese Academy of Sciences, No. 1 West Beichen Road, Chaoyang District, Beijing 100101, China, on February 26, 2024, with the following deposit number: CGMCC 29887. The bacterial specimen Paenibacillus polyxma strain P19 was deposited by The University of Hong Kong under terms in accordance with the Budapest Treaty with the China General Microbiological Culture Collection Center ( “CGMCC” ) , having an address of Institute of Microbiology Chinese Academy of Sciences, No. 1 West Beichen Road, Chaoyang District, Beijing 100101, China, on February 26, 2024, with the following deposit number: CGMCC 29888.BACKGROUND OF THE INVENTION

[0004] Colistin, also known as polymyxin E, is a last-resort antibiotic used to treat infections caused by multidrug-resistant gram-negative bacteria. However, the production of this compound poses significant challenges due to the presence of impurities resulting from various analogs, a series of endogenous secondary metabolites, and the limited number of native producers. Consequently, downstream isolation of Colistin for clinical usage becomes a challenging task. Recent advances in synthetic biology techniques have made rational metabolic engineering of colistin natural producers a promising strategy to overcome this challenge.

[0005] Polymyxins are a class of cationic lipopeptides that consist of a fatty acyl N-terminus, a tripeptide side chain, and a heptapeptide [1] . These potent antibiotics were first discovered in 1947 [2] and are widely reported from various strains of Paenibacillus polymyxa. They are classified as polymyxin A, B, C, D, E (Colistin) , M, P, S and T based on their amino acids at positions 3, 6 and 7 [3-6] . Polymyxin B and Colistin are considered as the last-line antibiotics in therapy due to their ability to combat multidrug-resistant Gram-negative bacteria [7, 8] . However, their complex chemical structures make them challenging to synthesize economically, and their production relies on large-scale fermentations of producer microorganisms.

[0006] Polymyxins are nonribosomal peptides biosynthesized by the pmx biosynthetic gene cluster (BGC) in Paenibacillus species [9] . However, its native producers, Paenibacillus, also harbor diverse BGCs of other secondary metabolites, often generating unwanted impurities. Additionally, these producers are limited in production potential due to native regulatory systems and restricted metabolic flux. Traditional methods for enhancing production have limitations in eliminating impurities and redirecting metabolic flux. Furthermore, the lack of genetic manipulation and genome editing tools for P. polymyxa has previously restricted the rational metabolic engineering of polymyxins [9, 10] . However, recent advances in synthetic biology, such as the emergence of CRISPR / Cas9-based genetic editing, have enabled notable progress in the genetic manipulation tools of P. polymyxa [11-14] . This progress allows for rational metabolic modifications in P. polymyxa, exemplified by recent metabolic engineering for 2, 3-butanediol and indole-3-acetic acid [15, 16] .

[0007] Recently, scientists have obtained ideal chassis strains for laboratory or industrial applications by deleting non-essential biosynthetic gene clusters (BGCs) from the genome for other bacterial strains. These strains have a clean metabolite background

[0017] , better growth

[0018] , and reduced precursor competition [19, 20] . Furthermore, some BGCs in bacteria act as signal molecules that regulate physiological and metabolic processes, potentially impacting the production of target products [21-23] . By deleting these BGCs and enhancing the supply of precursor amino acids, researchers have increased the production of natural product compounds such as bacitracin, surfactin, and erythromycin in various bacterial strains [24-28] . However, current metabolic engineering efforts for P. polymyxa strains have been limited to one or a few gene modifications, lacking a comprehensive approach that considers metabolic flux, precursor amino acid supply, and gene expression regulation.

[0008] Any discussion of documents, acts, materials, devices, articles or the like which has been included in the present specification is not to be taken as an admission that any or all of these matters form part of the prior art base or were common general knowledge in the field relevant to the present disclosure as it existed before the priority date of each claim of this application.

[0009] Throughout this specification the word “comprise, ” or variations such as “comprises” or “comprising, ” will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or group of elements, integers or steps.

[0010] BRIEF SUMMARY OF THE INVENTION

[0011] Disclosed are mutant Paenibacillus bacteria that produce Colistin A more efficiently and / or Colicin A that is of higher purity and / or has a cleaner metabolite background.

[0012] Disclosed are mutant Paenibacillus bacteria in which the expression of one or more biosynthetic gene clusters (BGCs) is inactivated. In some forms of the mutant bacteria, one of the inactivated BGCs is BGC4. Typically, the pmx BGC, which is responsible for Colistin production, is not inactivated.

[0013] In some forms of the mutant bacteria, expression of one or more BGCs is increased. In some forms, expression of at least one of the BGCs is increased by genetically engineering the BGC to be expressed using the PH4 promoter.

[0014] In some forms of the mutant bacteria, expression of one or more L-DABA pathway genes is increased. In some forms, expression of at least one of the L-DABA pathway genes is increased by genetically engineering the BGC to be expressed using the PH4 promoter.

[0015] In some forms, the mutant bacteria are progeny of a first parental bacterium genetically engineered to inactivate expression of one or more biosynthetic gene clusters (BGCs) . Typically, the pmx BGC, which is responsible for Colistin production, is not inactivated.

[0016] In some forms, the mutant bacteria are also progeny of a second parental bacterium genetically engineered to increase expression of one or more BGCs.

[0017] In some forms, the mutant bacteria are also progeny of a third parental bacterium genetically engineered to increase expression of one or more L-DABA pathway genes.

[0018] In some forms, the mutant bacteria have a clean metabolite background. In some forms, the mutant bacteria are Paenibacillus polymyxa.

[0019] In some forms, the mutant bacteria, when cultured, produce a colistin titer of at least 300 mg / L. In some forms, the mutant bacteria, when cultured, produces a colistin titer of at least 600 mg / L.

[0020] In some forms, the mutant bacteria, when cultured, produces a Colistin A titer of at least 200 mg / L. In some forms, the mutant bacteria, when cultured, produces a Colistin A titer of between about 200 mg / L and 2000 mg / L. In some forms, the mutant bacteria, when cultured, produces a Colistin A titer of about 200 mg / L, about 300 mg / L, about 400 mg / L, about 500 mg / L, about 600 mg / L, about 700 mg / L, about 800 mg / L, about 900 mg / L, about 1000 mg / L, about 1200 mg / L, about 1400 mg / L, about 1600 mg / L, about 1800 mg / L, and about 2000 mg / L. In some forms, the Colistin A titer is a greater than the titer of any other polymyxin. In some forms, the Colistin A titer is greater than the total titer of any other polymyxins. In some forms, Colistin A is at least 90%by weight of total polymyxin.

[0021] In some forms, the mutant bacteria are engineered strain P8, P9, P10, P14, P15, P16, P17, P18, or P19.

[0022] Also disclosed are methods of producing Colistin A by culturing the disclosed mutant bacteria and collecting culture liquid, where the culture liquid comprises Colistin A.

[0023] Also disclosed are compositions comprising Colistin A. In some forms, the Colistin A of the compositions is produced from the disclosed mutant bacteria or by the disclosed methods of producing Colistin A.

[0024] Additional advantages of the disclosed method and compositions will be set forth in part in the description which follows, and in part will be understood from the description, or can be learned by practice of the disclosed method and compositions. The advantages of the disclosed method and compositions will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings illustrate several embodiments of the disclosed method and compositions and together with the description, serve to explain the principles of the disclosed method and compositions.

[0026] Figures 1A-1D illustrate identification of colistin (polymyxin E) analogs produced by P. polymyxa ATCC 842. Figure 1A shows the biosynthesis gene cluster (BGC) of Colistin (polymyxin E) . Figure 1B shows the structure of Colistin (polymyxin E) .Figure 1C shows HPLC analysis of supernatant of P. polymyxa ATCC 842 culture liquid, colistin A (polymyxin E1) , colistin B (polymyxin E2) , and co-injection analysis. Figure 1D shows HRMS analysis of two peaks produced by P. polymyxa ATCC 842.

[0027] Figure 2 shows HPLC analysis of fermentation supernatant diluted by same volume of methanol from P. polymyxa ATCC 842 culture liquid, polymyxin E1 (colistin A) , polymyxin E2 (colistin B) , polymyxin E2 -I (Ile) , polymyxin E3, polymyxin E4 and polymyxin E7.

[0028] Figures 3A-3E show the structure of known natural products of NRPS or Ripps BGCs in P. polymyxa ATCC 842.

[0029] Figures 4A and 4B show the effect of single inactivation of non-essential BGCs on colistin production. Figure 4A shows transcription analysis of NRPS or Ripps BGCs with the potential for peptide-based molecules biosynthesis in P. polymyxa ATCC 842. Data are presented with 3 biological replicates. Figure 4B shows colistin production in single BGC inactivation at 48h. Data are presented as mean values of 3 biological replicates.

[0030] Figure 5 is a line graph showing the shows the growth curve of P. polymyxa ATCC 842 WT, P3, P14.

[0031] Figures 6A-6C show iterative knock-out of non-essential BGCs in P. polymyxa. Figure 6A shows transcriptional level of genes pmxA, pmxB and pmxE in P. polymyxa ATCC 842 WT, P1 and P3 at log phase. Figure 6B shows colistin production in multiple BGCs inactivation mutants at 48h. Figure 6C shows metabolites background of P. polymyxa ATCC 842 WT, mutant P3 and P14 under 200nm PDA detector. Data of Figures 6A and 6B are presented as mean values of 3 biological replicates. The HPLC analysis of Figure 6C was conducted with biological replicates, and a representative sample of each experimental group was presented here.

[0032] Figures 7A-7C show enhancement of the supply of L-DABA to increase colistin titer. Figure 7A shows the L-2, 4-DABA pathway (L-Lysine pathway) . Figure 7B shows promoter characterization using sfGFP as a reporter. Figure 7C shows the effect of overexpressing genes involved in the L-2, 4-DABA pathway on colistin production. Data of Figures 7B and 7C are presented as mean values of 3 biological replicates.

[0033] Figure 8 is a bar graph showing the Colistin production of BGC 15 overexpression strain.

[0034] Figures 9A-9C show colistin production of P. polymyxa P19 in 5-L fermenter. Figure 9A shows colistin titer during fermentation of P. polymyxa P19 in a 5-L fermenter. Figure 9B shows the chemical structures of Colistin A and Colistin B. Figure 9C shows the HPLC profile of supernatant of P19 fermentation liquid (a) and a Colistin standard product (b) under 200nm PDA detector.DETAILED DESCRIPTION OF THE INVENTION

[0035] The disclosed method and compositions can be understood more readily by reference to the following detailed description of particular embodiments and the Example included therein and to the Figures and their previous and following description.

[0036] I. DEFINITIONS

[0037] The use of the terms "a, " "an, " "the, " and similar referents in the context of describing the presently claimed invention (especially in the context of the claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context.

[0038] Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein.

[0039] Use of the term "about" is intended to describe values either above or below the stated value in a range of approx. + / -10%; in other embodiments the values may range in value either above or below the stated value in a range of approx. + / -5%; in other embodiments the values may range in value either above or below the stated value in a range of approx. + / -2%; in other embodiments the values may range in value either above or below the stated value in a range of approx. + / -1%. The preceding ranges are intended to be made clear by context, and no further limitation is implied. All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g., "such as" ) provided herein, is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the invention.

[0040] “Colistin” or “polymycin E” as used herein, is a cyclic lipopeptide with a hydrophobic N-terminal fatty acyl group, a tripeptide exocyclic chain, a heptapeptide ring, and a polymyxin antibiotic produced by certain strains of Paenibacillus polymyxa. Colistin interacts with the bacterial cytoplasmic membrane, changing its permeability therefore causing leakage of the cell contents.

[0041] The term "titer" refers to the concentration or activity of a specific enzyme or a biological product in a given sample. It is a quantitative measure that indicates the amount or potency of the enzyme produced by bacteria. The term “colistin titer” refers to the concentration of colistin produced by the mutant bacterium. Specifically, it indicates the amount of colistin, measured in milligrams per liter (mg / L) , that is present in the culture or sample of the bacterial mutant. For example, a colistin titer of at least 300 mg / L suggests that the mutant bacterium is capable of producing colistin at a concentration of 300 milligrams or more per liter of the culture medium. This measurement is indicative of the productivity or potency of the mutant strain in terms of colistin production.

[0042] The term “total titer” refers to the combined concentration of all variants or forms of a particular substance, such as the total titer of antibodies, enzymes, or other biological products in a culture or solution. For example, as used herein, “total titer” refers to the combined or overall concentration of polymyxins in a sample.

[0043] The term “inhibition” or “inhibit” when referring to the activity of a gene or protein refers to prevention of, or reduction in the rate of growth. Inhibition and or inhibit when referring to the activity of an enhancer of antimicrobial agent refers to the prevention or reduction of activity of a gene or gene product, that when inactivated potentiates the activity of the antimicrobial agent.

[0044] As used herein, the term "expressing" in relation to a gene or protein refers to making an mRNA or protein which can be observed using assays such as microarray assays, antibody staining assays, and the like. The term “expressing” also represented as “+” means, with respect to a cell protein level, detectable protein expression compared to a cell that is not expressing the protein, for example as measured by FACS analysis. The term “fold change” in relation to expression, refers to how much a quantity changes between an original and a subsequent measurement. Fold change is considered to be the ratio between the two quantities, e.g., quantities A and B the fold change of B with respect to A is B / A. For example, a change from 30 to 60 is defined as a fold-change of 2.This is also referred to as a "one-fold increase" . For a given comparison, a positive fold change value indicates an increase of expression, while a negative fold change indicates a decrease in expression. This value is typically reported in logarithmic scale (base 2) . For example, log2 fold change of 1.5 for a specific gene in the “WT vs KO comparison” means that the expression of that gene is increased in WT relative to KO by a multiplicative factor of 2^1.5 ≈ 2.82.

[0045] As used herein, the term "differentially expressed" as applied to a gene, refers to the production of the mRNA transcribed from the gene, or the protein product encoded by the gene that is different compared to normal or control. A differentially expressed gene may be overexpressed or under expressed as compared to the expression level of a normal or control cell. In one aspect, it refers to a differential that is at least 1.5 times, or at least 2.5 times, or alternatively at least 5 times, or alternatively at least 10 times higher or lower than the expression level detected in a control sample. The term "differentially expressed" also refers to nucleotide sequences in a cell or tissue which are expressed silent in a control cell or not expressed where expressed in a control cell.

[0046] “Genetically engineered” , “genetically engineer” , and “genetic engineering” refers to alteration of the DNA and / or RNA of a living cell by human intervention. In some forms, the alteration is mediated by the introduction of one or more expression vectors. In some forms, functionally equivalent alterations are achieved by mutagenesis. In some forms, the bacterium is genetically engineered to up-regulate the expression of genes in a biochemical pathway. “Up-regulation of a gene” refers to (i) genetically engineering a gene so that it has, relative to a wild-type gene, different control sequences that result in increased amounts of a gene product (RNA and, if the RNA is an mRNA, the protein encoded by the mRNA) in a cell; (ii) genetically engineering a bacterial cell so that it has, relative to a wild-type bacterial cell, more copies of a gene and increased amounts of the corresponding gene product; and / or (iii) genetically engineering the coding sequence of a gene to either increase the stability and / or activity of the gene product (i.e., if the increase the stability of an RNA gene product, increase translation of an mRNA gene product, and / or increase the level of enzymatic activity of a protein encoded by the mRNA gene product, i.e., by making the protein more stable or more active, also be referred to as up-regulation of an enzymatic product. An upregulated gene is the product of increased expression of a gene by any of the foregoing methods. Up-regulation of a gene thus results in increased expression of a gene.

[0047] As used herein, "strain" refers to a bacterial isolate or a group of isolates that exhibit phenotypic and / or genetic traits that are of the same lineage, but different from other isolates or strains of the same species.

[0048] As used herein, the term “purified” and like terms relate to the molecule or compound in a form that is substantially free (at least 60%free, preferably 75%free, and most preferably 90%free) from other components normally associated with the molecule or compound in a native environment.

[0049] Mutations of nucleic acids and genes include insertion mutations, deletion mutations, and substitution mutations. Insertions and deletions can be of one or more nucleotide residues of the nucleic acid being mutation. Mutations of peptides, polypeptides, and proteins can include insertion, deletion or substitution of one or more amino acid residues in the polypeptide. A mutation in a gene can be any one or more mutations of nucleotide residue (s) or of part (s) or all of the gene. A deletion in a gene can be any one or more deletions of nucleotide residue (s) or of part (s) or all of the gene. A deletion of a gene is a deletion of all of the gene.

[0050] As used herein, "disruptive mutation" refers to a mutation that reduces or eliminates (i.e., "disrupts" ) the expression or activity of a gene or biosynthesis gene cluster. Disruptive mutations can cause a gene or biosynthesis gene cluster to be partially inactivated, completely inactivated, or deleted. In some forms, the disruptive mutation is a knockout (KO) mutation. In some forms, the disruptive mutation is any mutation that reduces, prevents, or blocks the biosynthesis of the product produced by the biosynthesis gene cluster. Disruptive mutations include, for example, mutations in the gene encoding the enzyme, mutations in the genetic regulatory elements involved in the expression of the gene encoding the enzyme, introduction of nucleic acids that produce proteins that attenuate or inhibit the activity of the enzyme, or introducing nucleic acids (eg, antisense RNA, siRNA, CRISPR) or proteins that inhibit the expression of the enzyme. Disruptive mutations can be introduced using any method known in the art. An "inactivated" gene does not produce any gene product that functions like the native, natural, or endogenous gene. For example, the functional activity of an inactivated gene product is reduced to 10%, 20%, 50%, or 100%, 1-fold, 2-fold, 5-fold, or 10-fold less than the functional activity of the native, natural, or endogenous gene product.

[0051] The phrase “clean metabolite background” refers to a low level of extraneous or unwanted metabolites in addition to the desired products. For example, as used herein, a "clean metabolite background" is desirable because it means that the mutant bacteria are focusing their metabolic activities on producing the intended metabolite or compound without significant by-products or impurities.

[0052] The term “culture liquid” refers to the liquid medium in which microorganisms, such as bacteria, fungi, or cells, are cultivated or grown and which provides the necessary nutrients, salts, and conditions to support the growth and reproduction of the microorganisms.

[0053] The term “collecting” , as used herein, refers to the process of retrieving the liquid medium in which the bacteria have been cultivated, and which contains specific compounds or products of interest, e.g., Colistin A. Methods of collecting liquid medium and the products of interest are known in the art and include but are not limited to centrifugation and filtration e.g., vacuum filtration and centrifugal filtration.

[0054] The term “polymyxins” refer to the secondary metabolites of Paenibacillus or similar species, formed as a group of non-ribosomally synthesized cyclic lipopeptides with the typical cyclic structure attached to a fatty acid tail or sometimes amino acids of bacterial origin.

[0055] The term “variant” or “mutant, ” as used herein refer to an artificial outcome that has a pattern that deviates from what occurs in nature.

[0056] It is to be understood that the disclosed method and compositions are not limited to specific synthetic methods, specific analytical techniques, or to particular reagents unless otherwise specified, and, as such, can vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting.

[0057] II. COMPOSITIONS

[0058] Genetically modified Paenibacillus bacteria, and compositions containing the genetically modified Paenibacillus bacteria have been established. The disclosed genetically modified Paenibacillus bacteria are produced using a series of integrated bioengineering strategies that facilitate the production of high yields of Colistin A with low metabolite impurities.

[0059] Although not intended to be limiting, the following are exemplary features of the disclosed genetically modified Paenibacillus bacteria: (1) utilization of genome mining to identify Colistin biosynthetic gene clusters (BGCs) that harbor a condensation starter domain having a specific fatty acid side chain to produce Colistin A; (2) genetic modification of seven non-essential biosynthesis gene clusters in a Paenibacillus bacterial chassis strain by introducing disruptive mutations reduces the production of metabolite impurities associated with these gene clusters; (3) strengthened supply of L-2, 4, -diaminobutyric acid (L-DABA) , a crucial precursor in Colistin A biosynthesis via metabolic engineering to identify suitable promoters in the Paenibacillus bacterial chassis strain; (4) utilization of a 5-Liter (5L) fermenter bioreactor for the fermentation process, thereby reinforcing the ideal growth conditions of the genetically modified Paenibacillus bacteria and maximizing the production of highly purified Colistin A; and (5) utilization of an enzyme given the name CUU60_RS12930. The amino acid sequence for enzyme CUU60_RS12930 is represented by SEQ ID NO: 7.

[0060] CUU60_RS12930 Amino sequence:

[0061] The enzyme, as encoded by the CUU60_RS12930 amino acid sequence designated by SEQ ID NO: 7, plays a crucial role in the biosynthesis of Colistin A. Specifically, this enzyme catalyzes the reaction of L-2, 4-diaminobutyric acid, facilitating the formation of a peptide bond with 6-methyloctanoyl. This reaction is a key step in the synthesis pathway, leading to the generation of an intermediate product that is specific to Colistin A rather than the intermediate products of Colistin B or other Colistin analogs. Through this enzymatic action, the pathway is directed towards the production of Colistin A.

[0062] Also provided is an isolated peptide having SEQ ID NO: 8. The isolated peptide is a fragment of the enzyme CUU60_RS12930 provided above. In some forms, the isolated peptide has the following amino acid sequence:

[0063] EQYGLTQAQRRIWFMEIMNPGTSITMLSATYQITGEINTQLLKQAAAEIVKTYDVFRIRISGDLQNPTQWFEEPENVQARISRLEIGTTEQFYAWVKEVSEKPASVFDEHLYQFTIIHFANGQVWLNLTVNHIIADGLSVNALLHAVMEKYLELRKGISSSYQPPSYLDYISSEREYEQSQRYQKGKEYWLTKYNTLPETTGIKSYPPFSIGSESNKLSITLDGSRYERILAFSEQYQVSLYTLFLSAMYALLYKLTDSTDVPVGTVFANRTSKKEKETIGMFVSTVATRICLNPDG (SEQ ID NO: 8) . In some forms, the isolated peptide is a variant having at least 70%sequence similarity to SEQ ID NO: 8. In some forms, the isolated peptide is a variant having from about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, or about 99%sequence similarity to SEQ ID NO: 8.

[0064] A. Genetically Modified Paenibacillus bacteria

[0065] As described in the non-limiting Examples, a Paenibacillus polymyxa bacterial strain was identified as a chassis strain and genetically modified via metabolic engineering to generate genetically modified Paenibacillus bacteria. It has been demonstrated that alterations in gene expression within specific biosynthetic gene clusters significantly increase the yield and purity of Colistin A. It has also been demonstrated that expression of specific promoters facilitates increased synthesis of L-2, 4-diaminobutryic acid, a crucial precursor in Colistin A biosynthesis.

[0066] Thus, genetically modified Paenibacillus bacteria are disclosed. The genetically modified bacteria are Paenibacillus polymyxa bacteria that produce high yields of Colistin A with a clean metabolite background.

[0067] 1. Structure of Genetically Modified Paenibacillus bacteria

[0068] The genetically modified Paenibacillus bacteria typically have one or more of (i) disruptive mutations in genes of one or more biosynthetic gene clusters, (ii) a substituted and overexpressed promoter, and (iii) overexpression of molecules responsible for the synthesis of L-2, 4-DABA.

[0069] i. BGC Mutations

[0070] The genetically modified Paenibacillus bacteria (hereafter referred to as “mutant Paenibacillus bacteria” ) typically have mutations in genes in one or more biosynthetic gene clusters. Biosynthetic gene clusters (BGCs) typically include a gene encoding a core biosynthetic enzyme, most commonly a non-ribosomal peptide synthetase (NRPS) , polyketide synthase (PKS) , or terpene cyclase, that is responsible for the first metabolic step in product synthesis (Khaldi N, et al., Fungal Genet Biol 47: 736–741, (2010) ) . Additionally, BGCs include genes encoding so-called “tailoring” enzymes, such as P450 monooxygenases or methyltransferases, that modify the molecule produced by the core enzyme (Khaldi N, et al., Fungal Genet Biol., 47: 736–741, (2010) , Weber T, et al., Nucleic Acids Res., 43: W237–W243, (2015) ) . Many BGCs contain either genes encoding putative membrane transporters, which are required for metabolite efflux from the cell (Wang D-N, et al., Med Mycol 52: 506–518 (2014) ) , or additional “resistance” genes, which are necessary for detoxification / self-protection against the molecules produced (Schrettl M, et al., PLoS Pathog., 6: e1000952, (2010) ) .

[0071] As demonstrated in the non-limiting Examples, fifteen BGCs were found with the potential to biosynthesize peptide-based molecules using the bioinformatic tool antiSMASH version 5.1.0 (Blin et al., Nucleic Acids Research, 47 (W1) : W81-W87 (2019) ) , and 12 BGCs were predicted to produce peptide-based products. It was shown that disrupting expression in one or more BGCs with capacity of producing peptide-based products that potentially competed with or regulate colistin biosynthesis resulted in increased the yield and purity of colistin.

[0072] In some forms, the mutant Paenibacillus bacteria have mutations in or deletions in or of genes encoding peptide-based molecules of BGCs responsible for the synthesis of non-ribosomal peptides (NRPs) . Nonribosomal peptides is a diverse family of secondary metabolites with diverse properties as toxins, siderophores, pigments, antibiotics, cytostatics, immunosuppressants or anticancer agents; their synthesis is independent of ribosomal machinery ( and López, Sustain Chem Process 4, 13 (2016) ) . In some forms, the mutant Paenibacillus bacteria have mutations in non-ribosomal peptides involved in the synthesis of Tridecaptin (referred to as “BGC1” ) , Colistin (Polymyxin E) (referred to as “BGC5” ) , and Fusaricidin (referred to as “BGC7” ) .

[0073] In some forms, the mutant Paenibacillus bacteria have mutations in or deletions in or of genes encoding peptide-based molecules of BGCs responsible for the synthesis of ribosomally synthesized and post-translationally modified peptides (RiPPs) . In some forms, the ribosomally synthesized and post-translationally modified peptides are lanthipeptides. In some forms, the mutant Paenibacillus bacteria have mutations in one or both lanthipeptides, Paenibacillin (referred to as “BGC8” ) and Paenilan (referred to as “BGC13” ) . In some forms, the mutant Paenibacillus bacteria have mutations in one or more other lanthipeptides found in BGC6.

[0074] In some forms, the mutant Paenibacillus bacteria have mutations in or deletions in or of genes encoding peptide-based molecules of BGCs responsible for the synthesis of non-ribosomal peptide synthetase polyketide synthase (NRPS-PKS) (referred to as “BGC2” and “BGC4” , “BGC9” , and “BGC11” ) , lassopeptides (referred to as “BGC12” ) , and NRPS-like molecules (referred to as “BGC15” ) .

[0075] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in or deletions in or of one or more genes encoding molecules of a single BGC. In these forms, the mutant Paenibacillus bacteria have mutations in or deletions in or of one or more genes encoding BGC1, BGC2, BGC4, BGC7, BGC8, BGC12, or BGC13. In one form, the mutation (s) or deletion (s) is in one or more genes encoding molecules from BGC4 when a single BGC is mutated. In some forms, one or more of the following genes in BGC4 is mutated or deleted: CUU60_RS07225, CUU60_RS07220, CUU60_RS07215, CUU60_RS07210, CUU60_RS07205, CUU60_RS07200, CUU60_RS07195, CUU60_RS07190, CUU60_RS07185, CUU60_RS07180 and CUU60_RS07175. For example, one or more of the genes CUU60_RS07225, CUU60_RS07220, CUU60_RS07215, CUU60_RS07210, CUU60_RS07205, CUU60_RS07200, CUU60_RS07195, CUU60_RS07190, CUU60_RS07185, CUU60_RS07180 and CUU60_RS07175 in the BGC4 according to the antiSMASH prediction can be deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC4 includes 100, 435 nucleotides and is located from position 1, 494, 709 to 1, 595, 143 of the chassis strain NZ_CP024795.

[0076] In some forms, the mutant Paenibacillus bacteria have disruptive mutations or deletion (s) in one or more genes encoding molecules of multiple BGCs. In some forms, the mutant Paenibacillus bacteria have mutations or deletion (s) in one or more genes encoding molecules of two BGCs. In these forms, the mutations or deletion (s) are in genes encoding molecules of BGC1 and BGC4. In some forms, one or both of the following genes in BGC1 is mutated or deleted: CUU60_RS01235 and CUU60_RS01240. For example, one or both of the genes CUU60_RS01235 and CUU60_RS01240 in the BGC1 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC1 includes 93, 568 nucleotides and is located from position 256, 414 -349, 981 of the chassis strain NZ_CP024795. In some forms, one or more of the following genes in BGC4 is mutated or deleted: CUU60_RS07225, CUU60_RS07220, CUU60_RS07215, CUU60_RS07210, CUU60_RS07205, CUU60_RS07200, CUU60_RS07195, CUU60_RS07190, CUU60_RS07185, CUU60_RS07180 and CUU60_RS07175.

[0077] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of three BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, and BGC4. In some forms, one or more of the following genes in BGC2 is mutated or deleted: CUU60_RS02235, CUU60_RS02240 and CUU60_RS02245. For example, one or more of the genes CUU60_RS02235, CUU60_RS02240 and CUU60_RS02245 in the BGC2 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC2 includes 105, 581 nucleotides and is located from position 507, 910 -613, 490 of the chassis strain NZ_CP024795.

[0078] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of four BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, and BGC7. In some forms, one or more of the following genes in BGC7 is mutated or deleted: CUU60_RS16865, CUU60_RS16870, CUU60_RS16875, CUU60_RS16880, CUU60_RS16885 and CUU60_RS16890. For example, one or more of the genes CUU60_RS16865, CUU60_RS16870, CUU60_RS16875, CUU60_RS16880, CUU60_RS16885 and CUU60_RS16890 in the BGC7 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC7 includes 65, 394 nucleotides and is located from position 3, 694, 410 -3, 759, 803 of the chassis strain NZ_CP024795.

[0079] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of five BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, BGC7 and BGC8. In some forms, one or more of the following genes in BGC8 is mutated or deleted: CUU60_RS20960, CUU60_RS20965, CUU60_RS20970 and CUU60_RS20975. For example, one or more of the genes CUU60_RS20960, CUU60_RS20965, CUU60_RS20970 and CUU60_RS20975 in the BGC8 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC8 includes 23, 101 nucleotides and is located from position 4, 631, 262 -4, 654, 362 of the chassis strain NZ_CP024795.

[0080] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of six BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, BGC7, BGC8, and BGC12. In some forms, one or more of the following genes in BGC12 is mutated or deleted: CUU60_RS22345, CUU60_RS22350, CUU60_RS22355, CUU60_RS22360 and CUU60_RS22365. For example, one or more of the genes CUU60_RS22345, CUU60_RS22350, CUU60_RS22355, CUU60_RS22360 and CUU60_RS22365 in the BGC12 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC12 includes 24, 068 nucleotides and is located from nucleotide position 5,025,669 to nucleotide position 5,049,736 of the chassis strain NZ_CP024795.

[0081] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of seven BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, BGC7, BGC8, BGC12, and BGC13. In some forms, one or more of the following genes in BGC13 is mutated or deleted: CUU60_RS24780, CUU60_RS24785, CUU60_RS24790, CUU60_RS24795 and CUU60_RS24800. For example, one or more of the genes CUU60_RS24780, CUU60_RS24785, CUU60_RS24790, CUU60_RS24795 and CUU60_RS24800 in the BGC13 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, BGC13 includes 27, 010 nucleotides and is located from position 5, 463, 080 -5, 490, 089 of the chassis strain NZ_CP024795.

[0082] In preferred forms, the mutant Paenibacillus bacteria have disruptive mutations and / or deletions in genes encoding molecules in BGC4. In some forms, the mutant Paenibacillus bacteria have one or more mutations or deletions in genes affecting the transcription levels of BGC4, which encode nonribosomal peptide-synthetase responsible for colistin biosynthesis.

[0083] In some forms, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxA by about 1.20-fold to about 2.20-fold. In some forms, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxA by about 1.20-fold, about 1.40-fold, about 1.60-fold, about 1.80-fold, about 1.90-fold, about 2.00-fold, or about 2.20 fold. In one exemplary form, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxA by about 1.63-fold.

[0084] In some forms, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxB by about 2.0-fold to about 3.5-fold. In some forms, the one or more mutations or deletions in BGC gene increase transcriptional levels of pmxB by about 2.0-fold, about 2.20-fold, 2.40-fold, 2.60-fold, 2.80-fold, 3.00-fold, 3.20-fold, 3.40-fold, or about 3.60 fold. In one exemplary form, the one or more mutations or deletions in BGC genes increases transcriptional levels of pmxB by about 2.89-fold.

[0085] In some forms, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxE by about 3.0-fold to about 6.00-fold. In some forms, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxB by about 3.00-fold, 3.50-fold, 4.00-fold, 4.20-fold, 4.50-fold, 5.00 fold, 5.50 fold, or about 6.00 fold. In one exemplary form, the one or more mutations or deletions in BGC genes increase transcriptional levels of pmxE by about 4.20-fold.

[0086] In some forms, the mutant Paenibacillus bacteria have mutations in one or more pmx BGC genes which enhance the production of colistin. In some forms, pmx mutation (s) is in genes encoding one or more of pmxA, pmxB, and pmxE. In some forms, the pmx mutation (s) increases transcriptional levels of pmxA by about 1.20-fold to about 2.20-fold. In some forms, the pmx mutation (s) increases transcriptional levels of pmxA by about 1.20-fold, about 1.40-fold, about 1.60-fold, about 1.80-fold, about 1.90-fold, about 2.00-fold, or about 2.20-fold. In one exemplary form, the pmx mutation (s) increases transcriptional levels of pmxA by about 1.63-fold. The open reading frame for pmxA contains 14, 994 nucleotides, spans from nucleotide position 2, 838, 404 to nucleotide position 2, 853, 397 of the chassis strain NZ_CP024795 and encodes the amino acid sequence represented by SEQ ID NO: 19.

[0087] pmxA Amino Acid Sequence:

[0088] In some forms, the pmx mutation (s) increases transcriptional levels of pmxB by about 2.0-fold to about 3.5-fold. In some forms, the pmx mutation (s) increases transcriptional levels of pmxB by about 2.0-fold, about 2.20-fold, 2.40-fold, 2.60-fold, 2.80-fold, 3.00-fold, 3.20-fold, 3.40-fold, or about 3.60-fold. In one exemplary form, the pmx mutation (s) increases transcriptional levels of pmxB by about 2.89-fold. The open reading frame for pmxB contains 3, 309 nucleotides, spans from nucleotide position 2, 834, 941 to nucleotide position 2, 838, 249 of the genomic sequence of the chassis strain NZ_CP024795, and encodes the amino acid sequence represented by SEQ ID NO: 20.

[0089] pmxB Amino Acid Sequence:

[0090] In some forms, the pmx mutation (s) increases transcriptional levels of pmxE by about 3.0-fold to about 6.00-fold. In some forms, the pmx mutation (s) increases transcriptional levels of pmxB by about 3.00-fold, 3.50-fold, 4.00-fold, 4.20-fold, 4.50-fold, 5.00-fold, 5.50 fold, or about 6.00 fold. In one exemplary form, the pmx mutation (s) increases transcriptional levels of pmxE by about 4.20-fold. The open reading frame for pmxE contains 27, 355 nucleotides, spans from nucleotide position 2, 813, 810 to nucleotide position 2, 831, 164 of the genomic sequence of the chassis strain NZ_CP024795, and encodes the amino acid sequence represented by SEQ ID NO: 21.

[0091] pmxE Amino Acid Sequence:

[0092] In some forms, the mutant bacteria include genes from colistin BGCs that harbor a condensation starter domain having a specific fatty acid side chain to produce Colistin A. A Condensation starter domain (also known as a C-domain) are protein domains in the multi-domain non-ribosomal peptide synthases which catalyze a condensation reaction to form peptide bonds in non-ribosomal peptide biosynthesis. C-domains are large enzymes (~450 amino acids) , typically embedded in an NRPS protein or subunit, and catalyze amide bond formation between an amino acid tethered to the downstream T domain and an amino acid or peptide tethered to the upstream T domain, leading to elongation of the growing peptide chain (Bloudoff &Schmeing, Biochim. Biophys. Acta 1865: 1587–1604 (2017) ; Dekimpe, &Masschelein, Nat. Prod. Rep. 38: 1910–1937 (2021) ; reviewed in Folger et al., Nature chemical biology, 2024) . PmxA, containing 4, 953 amino acids, contains four modules and a C-domain. PmxE, a 6, 312-amino-acid polypeptide, has five modules and a C-domain. In some forms the C-domain for pmxE has the amino acid sequence represented by SEQ ID NO: 24. In some forms, the C-domain for pmxE has an amino acid sequence having at least 70%sequence similarity to SEQ ID NO: 24. In some forms, the C-domain for pmxE is a variant having from about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, or about 99%sequence similarity to SEQ ID NO: 24.

[0093] pmxE Starter C-domain Amino Acid Sequence:

[0094] QYGLTQAQRRIWFMEIMNPGTSITMLSATYQITGEINTQLLEQAVAEIVKTYDVFRIRISGDLQNPTQWFEEPENVQARISRLEIGTTEQFYAWVKEVSEKPASVFDEHLYQFTIIHFANGQVWLNLTVNHIIADGLSVNALLHAVMEKYLELRKGISSSYQAPSYLDYISAAFSEQYQVSLYTLFLSAMYALLYKLTDSTDVPVGTVFANRTSKKEKETIGMFVSTVATRIRLNPDGHVLSLIQTVSKENTADLRYQKYPYNQLIQDLREQHGRNDLSGLFRTSLEYL (SEQ ID NO: 24) (Shaheen et al., Chemistry and Biology, 18(12) : 1640-1648 (2011) ) .

[0095] ii. Overexpressed Promoters

[0096] In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress selected promoter genes. Overexpression of the selected promoter genes in part results in increased colistin yield (FIGs. 4A–4C) . In some forms, the mutant Paenibacillus bacteria contain genetic modifications which overexpress promoter H1. The nucleic acid sequence for promoter H1 is provided as SEQ ID NO: 3.

[0097] H1 Promoter Sequence:

[0098] In some forms, the mutant Paenibacillus bacteria contain genetic modifications which overexpress promoter H4. The nucleic acid sequence for promoter H4 is provided as SEQ ID NO: 4.

[0099] H4 Promoter Sequence:

[0100] In some forms, the mutant Paenibacillus bacteria contain genetic modifications which overexpress promoter H2. The nucleic acid sequence for promoter H2 is provided as SEQ ID NO: 9.

[0101] H2 Promoter Sequence:

[0102] In some forms, the mutant Paenibacillus bacteria contain genetic modifications which overexpress promoter H3. The nucleic acid sequence for promoter H3 is provided as SEQ ID NO: 10.

[0103] H3 Promoter Sequence:

[0104] In some forms, the mutant Paenibacillus bacteria contain genetic modifications which overexpress promoter M1. The nucleic acid sequence for promoter M1 is provided as SEQ ID NO: 11.

[0105] M1 Promoter Sequence:

[0106] In some forms, the mutant Paenibacillus bacteria contain genetic modifications which overexpress promoter M2. The nucleic acid sequence for promoter M2 is provided as SEQ ID NO: 12.

[0107] M2 Promoter Sequence:

[0108] iii. Overexpressed Molecules in the L-2, 4-DABA Pathway

[0109] In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress selected genes in the L-2, 4-DABA pathway. This in part increases the supply of amino acid precursors in the chassis strain or parent bacterium, thereby subsequently increasing the production of non-ribosomal peptide products.

[0110] The L-2, 4-DABA pathway shares the crucial metabolic pathway of L-Lysine biosynthesis (FIG. 4A) . L-2, 4-diaminobutyric acid (DAB, DABA) is a non-metabolized amino acid. It is a plant metabolite and a conjugate base of L-2, 4-diazaniumylbutyrate. It is functionally related to a butyric acid. The chemical structure of L-2, 4-DABA is shown in formula I below.

[0111] Formula I –Structure of L-2, 4-diaminobutyric acid

[0112] In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress one or more enzymes in the L-2, 4-DABA pathway. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress diaminobutyrate-2-oxoglutarate transaminase (ectB) . ectB catalyzes reversively the conversion of L-aspartate beta-semialdehyde (ASA) to L-2, 4-diaminobutyrate (DABA) by transamination with L-glutamate. The amino acid sequence for ectB contains 427 amino acids and is represented by SEQ ID NO: 5. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress variants having about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, about 99%sequence identity to SEQ ID NO: 5.

[0113] ectB Amino Acid Sequence:

[0114] In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress aspartate-semialdehyde dehydrogenase (asd) . asd catalyzes the NADPH-dependent formation of L-aspartate-semialdehyde (L-ASA) by the reductive dephosphorylation of L-aspartyl-4-phosphate. The amino acid sequence for asd contains 346 amino acids and is represented by SEQ ID NO: 6. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress variants having about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, about 99%sequence identity to SEQ ID NO: 6.

[0115] asd Amino Acid Sequence:

[0116] In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress one or both aspartate kinases (lysC and dapG) . The amino acid sequence for lysC contains 417 amino acids and is represented by SEQ ID NO: 13. The amino acid sequence for dapG contains 405 amino acids and is represented by SEQ ID NO: 14. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress variants having about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, about 99%sequence identity to SEQ ID NO: 13, SEQ ID NO: 14, or both SEQ ID NO: 13 and SEQ ID NO: 14.

[0117] lysC Amino Acid Sequence:

[0118] dapG Amino Acid Sequence:

[0119] In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress ectB and asd. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress ectB and lysC. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress ectB and dapG. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress asd and dapG. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress asd and lysC. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress ectB, lysC, and dapG. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress asd, lysC, and dapG. In some forms, the mutant Paenibacillus bacteria have genetic modifications which overexpress ectB, asd, lysC, and dapG.

[0120] 2. Function of Genetically Modified Paenibacillus bacteria

[0121] As described in the non-limiting Examples, the mutant Paenibacillus polymyxa bacteria demonstrate significantly increased production of colistin having a clean metabolite background.

[0122] The mutant Paenibacillus bacteria generally demonstrate increased production of colistin, and its analogs compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrates from about 20%to about 300%higher yields of total colistin compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrate about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 99%, about 100%, about 120%, about 140%, about 160%, about 180%, about 200%, about 220%, about 240%, about 260%, about 280%, or about 300%, higher yield of colistin compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrate production of about 200 mg / L to about 2000 mg / L colistin titer.

[0123] In some forms, the mutant Paenibacillus bacteria demonstrates from about 20%to about 300%higher yields of Colistin A and Colistin B compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrate about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 99%, about 100%, about 120%, about 140%, about 160%, about 180%, about 200%, about 220%, about 240%, about 260%, about 280%, or about 300%, higher yield of colistin compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrate production of about 200 mg / L to about 2000 mg / L colistin titer.

[0124] In some forms, the mutant Paenibacillus bacteria demonstrate increased production of Colistin A or Colistin B. The structures of Colistin A and Colistin B are represented by Formula II and Formula III respectively. In one preferred form, the mutant Paenibacillus bacteria demonstrate increased production of Colistin A. In another preferred form, the ratio of Colistin A to the ratio of Colistin B produced by the mutant Paenibacillus bacteria is about 13: 1.

[0125] The commercial marked standard for colistin compositions contains a higher ratio of Colistin B to Colistin A. As shown in Figure 5, compositions containing a higher ratio of Colistin A to Colistin B are achieved. Thus, in some forms, the mutant Paenibacillus bacteria produces Colistin having a higher ratio of Colistin A to Colistin B. In some forms, the amount of Colistin A produced by the mutant Paenibacillus bacteria accounts for about 60%to about 95%of the total Colistin produced by the mutant Paenibacillus bacteria. In some forms, the amount of Colistin A produced by the mutant Paenibacillus bacteria accounts for about 60%, about 70%, about 80%, about 90%, about 95%of the total Colistin produced by the mutant Paenibacillus bacteria.

[0126] In some forms, the amount of Colistin B produced by the mutant Paenibacillus bacteria accounts for about 5%to about 10%of the total Colistin produced by the mutant Paenibacillus bacteria. In some forms, the amount of Colistin A produced by the mutant Paenibacillus bacteria accounts for about 5%, about 6%, about 7%, about 8%, about 9%, or about 10%of the total Colistin produced by the mutant Paenibacillus bacteria.

[0127] The mutant Paenibacillus bacteria also demonstrates production of Colistin with a clean metabolite background i.e., a lower production of unwanted metabolites. This is in part due to the mutant Paenibacillus bacteria having mutations which prevent endogenous genetic regulation that potentially interferes with the production of colistin analogs, and which increases conservation of resources and energy. Exemplary unwanted metabolites include but are not limited to Tridecaptin, Fusaricidin, Paenibacillin, and Paenilan. The structures of Tridecaptin, Fusaricidin, Paenibacillin, and Paenilan are provided below as Formula IV, Formula V, Formula VI, and Formula VII respectively. In some forms, the mutant Paenibacillus bacteria demonstrate from about 30%to about 100%lower production of unwanted metabolites compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrate about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 99%, or about 100%, lower production of unwanted metabolites compared to the parent bacterium.

[0128] B. Formulations

[0129] Disclosed are formulations containing the genetically modified Paenibacillus bacteria. Generally, the formulations containing the genetically modified Paenibacillus bacteria demonstrate increased Colistin A production.

[0130] In some forms, the formulation is a fermentate of a live culture containing genetically modified Paenibacillus polymyxa. “Fermentate” refers to the cultured strain (that is after inoculation of strain in a medium and incubation) which includes both the live biomass and the metabolites, i.e., antimicrobial agents. Typically, the fermentate contains one or more of the genetically modified bacterial strains e.g., one or more of the P. polymyxa 842 variants, plus the medium in which the genetically modified bacterial strain has grown and secreted its metabolites. In some forms, the fermentate is capable of inhibiting the growth or activity of one or more Gram-negative bacteria, including but not limited to Pseudomonas aeruginosa, Klebsiella pneumoniae, or Acinetobacter, Escherichia coli 0157, Pseudomonas spp., Salmonella enterica serovars, and Yersinia enterocolitica e.g., Acinetobacter baumannii.

[0131] In some forms, the formulation is a crude extract of a live culture containing genetically modified Paenibacillus polymyxa, strain 842. A crude extract is partially purified active antimicrobial agent (s) , derived from the fermentate or from the separated cells.

[0132] In one exemplary form, the bacteria are preserved via freeze-drying (lyophilization) . This involves freezing the bacteria and then removing the water content through sublimation, which directly converts the ice into vapor. The freeze-dried bacteria are then packaged in vials or ampules with moisture-proof seals to prevent rehydration. This method allows for long-term (several years) storage at room temperature.

[0133] 1. Liquid Formulations

[0134] In some forms, the disclosed formulations are in liquid form. The liquid compositions include an effective amount of a liquid stabilizer such as 20%glycerol or 3.5%NaCl. Non-limiting examples of liquid formulations include suspension concentrations and oil dispersions.

[0135] 2. Solid Formulations

[0136] In some forms, the genetically modified Paenibacillus bacteria are present in dry compositions such as dry powder, grains or granules or pellets. Dry compositions can be prepared by adding an osmoprotectant (such as sugar, egg trehalose and / or maltodextrin) to the microbial composition in solution in the desired ratio. This solution is combined with a dry carrier or absorbent such as an organic or inorganic carrier (discussed further below) at the desired concentration of the microorganism composition (such as 2-30%, e.g., 2.5-10%, 5-15%, 7.5-20%or 15-30%) . Granules can be created by including clay or polymer binders that serve to hold the granules together or impart specific physical or degradable properties. Granules can be formed using rotary granulation, mixer granulator or extrusion, as several possible methods. In other examples, dry compositions can be prepared by spraying or soaking a liquid microorganism composition on  / in a solid carrier. Additional methods for preparing dry compositions comprising one or more microorganism species are known to one of ordinary skill in the art, for example, as described in Formulation of Microbial Biopesticides: Beneficial Microorganisms, Nematodes and Seed Treatments, Burges, ed., Springer Science, 1998; Bashan, Biotechnol. Adv. 16: 729-770, 1998; Ratul et al., Int. Res. J. Pharm. 4: 90-95, 2013.

[0137] In some forms, the bacterial cells can be immobilized on a solid support to enhance delivery efficiency. Immobilization of cells offers numerous advantages, including maintenance of their metabolic activity owing to better-controlled chemical conditions and physical protection. Also, slow release of cells from a solid support can reduce the need for constant reinoculation. Additionally, immobilized cells will disperse less widely compared with free-living cells, thereby effectively reducing potential issues concerning escaped microorganisms affecting or interacting with nontarget organisms. Microorganisms can be immobilized in carriers made from three types of material: inorganic, organic, or a composite of both. Organic carriers include naturally occurring products (e.g., coconut husk, chitin, alginate, agar, and guar gum) and synthetic polymers (e.g., polypropylene, polyurethane, and acrylamide) . Inorganic carriers include charcoal, clay, silica, ceramics, and celite, among others. Whereas inorganic carriers are usually more resistant, less costly, and easier to handle and reuse, organic carriers are more adaptable and typically biocompatible and therefore safer for other organisms. Composite carriers, which combine the advantages of both organic and inorganic carriers, provide some level of compromise between these respective (dis) advantages and include substances such as polyvinyl alcohol, sodium alginate, and polyvinyl alcohol–guar gum. Another option is bioencapsulation, where living carriers, such as the live feed organisms Artemia spp., rotifers, or copepods, are used to successfully deliver nutrients, immune stimulants, and probiotics to several aquatic organisms.

[0138] III. METHODS OF MAKING

[0139] There is a critical need for antibiotics to treat infections caused by multidrug-resistant Gram-negative bacteria. Existing methods employ Polymyxin B and Colistin as last-line antibiotics. However, the complex chemical structures of these antibiotics render the process to economically synthesize them on a large scale difficult, and the resulting antibiotics are riddled with metabolite impurities. Thus, improved methods of producing mutant Paenibacillus bacteria are also provided.

[0140] As demonstrated in the in the non-limiting Examples, (1) metabolic engineering was used to genetically disrupt seven non-essential biosynthesis gene clusters in a Paenibacillus bacterial chassis strain by introducing disruptive mutations reduces the production of metabolite impurities associated with these gene clusters, overexpress selected promoters and overexpress genes encoding rate-limiting enzymes in the L-2, 4, -diaminobutyric acid (L-DABA) pathway; and (2) a 5-Liter (5L) fermenter bioreactor was used to stabilize and improve the cultivation conditions for the fermentation process, thereby reinforcing the ideal growth conditions of the genetically modified Paenibacillus bacteria and maximizing the production of highly purified Colistin A.

[0141] Thus, the improved methods of producing mutant Paenibacillus bacteria include combining genetic engineering and metabolic engineering to generate mutant Paenibacillus bacteria which demonstrates higher colistin production compared to wild-type Paenibacillus bacteria. For example, in some forms, the improved methods utilize CRISPR / Cas9 to metabolically engineer mutant Paenibacillus bacteria with high colistin production.

[0142] In an exemplary method, the steps for generating mutant Paenibacillus bacteria include:

[0143] (i) isolating and cloning antibiotic biosynthetic genes,

[0144] (ii) transfecting parent Paenibacillus bacterium with expression vectors containing the cloned antibiotic biosynthetic genes,

[0145] (iii) selection and screening transformed progeny bacteria for the desired mutant Paenibacillus bacteria, and

[0146] (iv) culturing the mutant Paenibacillus bacteria under conditions suitable for production of higher yields of Colistin with clean metabolite backgrounds.

[0147] In another exemplary method, the steps for generating mutant Paenibacillus bacteria include increasing the Colistin-producing ability of an Paenibacillus bacteria; said method includes one or both of the following steps:

[0148] 1) transforming Paenibacillus bacteria which produces Colistin or a Colistin precursor by means of an antibiotic biosynthetic pathway with a recombinant DNA cloning vector or portion thereof, said vector or portion thereof comprising an antibiotic biosynthetic gene that codes for the expression of an enzyme or other gene product that is rate-limiting in said antibiotic biosynthetic pathway, and

[0149] 2) culturing Paenibacillus bacteria transformed with said vector under conditions suitable for cell growth, expression of said antibiotic biosynthetic gene, and production of Colistin or a Colistin precursor, subject to the limitation that said culturing procedure selected in step (2) provides for an increase in the Colistin-producing ability of the mutant Paenibacillus bacteria.

[0150] A. Metabolic Engineering of Parental Bacterium

[0151] The disclosed methods include metabolic engineering of a parent bacterium to genetically alter biosynthetic genes and pathway genes that are responsible for the production of antibiotics, e.g. Colistin. Metabolic engineering is achieved by introducing disruptive mutations in a Paenibacillus bacterial chassis strain which reduces the production of metabolite impurities associated with these gene clusters, overexpressing selected promoters and overexpressing genes encoding rate-limiting enzymes in the L-2, 4, -diaminobutyric acid (L-DABA) pathway.

[0152] In some forms, the methods include selecting a suitable parent bacterium as the host bacterium, identifying and cloning antibiotic biosynthetic genes of interest, preparing expression vectors that includes the cloned antibiotic genes of interest under the control of selected promoters, transforming the parent bacterium with the expression vectors, and screening the progeny bacterium to select progeny which have successfully integrated the genes of interest.

[0153] 1. Sources of Parental Bacterium

[0154] Parental bacterium refers to a bacterium used for the purpose of producing a bacterium. The parental bacterium can be a naturally occurring bacterium (i.e., a wild-type bacterium) or a previously modified bacterium (i.e., a mutant, or recombinant bacterium) . The bacterial progeny derived from the parental bacterium express or overexpress one or more enzymes that were not expressed or were overexpressed in the parent bacterium. Similarly, the parental bacterium can be modified to produce bacterial progeny containing one or more genes that the parent bacterium did not contain.

[0155] In preferred forms, the parental bacterium is a Paenibacillus bacterium. In some forms, the parental bacterium is a species selected from Paenibacillus polymyxa, Paenibacillus amylolyticus, Paenibacillus alvei, Paenibacillus thiaminolyticus, Paenibacillus dendritiformis, and Paenibacillus peoriae. Other species of Paenibacillus that can be used as the parental bacterium include but are not limited to Paenibacillus xylanexedens, Paenibacillus sambharensis, Paenibacillus aquistagni, Paenibacillus barcinonensis, Paenibacillus jamilae, and Paenibacillus barcinonensis. In one preferred form, the parental bacterium is Paenibacillus polymyxa.

[0156] Paenibacillus polymyxa bacteria that can be used as the parental bacteria are commercially available. In an exemplary form, the Paenibacillus polymyxa bacteria used as the parental bacteria is Paenibacillus polymyxa, ATCC strain No. 842. In some forms, other strains of Paenibacillus polymyxa bacteria can be mutated in a similar manner as described in the Non-limiting Examples to produce higher levels of Colistin, e.g., Colistin A. Exemplary strains of Paenibacillus polymyxa bacteria that can be used as the parental bacteria include Paenibacillus polymyxa ATCC Strain No. 21830, Paenibacillus polymyxa ATCC Strain No. 15970, Paenibacillus polymyxa ATCC Strain No. 7070, Paenibacillus polymyxa ATCC Strain No. 43865, Paenibacillus polymyxa ATCC Strain No. 10401, Paenibacillus polymyxa ATCC Strain No. 31037, 27955, Paenibacillus polymyxa ATCC Strain No. 8519, Paenibacillus polymyxa ATCC Strain No. 8524, and Paenibacillus polymyxa DMS 32871.

[0157] 2. Nucleic acid constructs

[0158] The disclosed methods include designing and producing nucleic acid constructs that are capable of disrupting existing genes (knock-out) or inserting new genes (knock-in) in the parent bacterium.

[0159] In some forms, the nucleic acid constructs are introduced into the parent bacterium to produce mutant Paenibacillus bacteria which have disruptive mutations in or deletions in or of one or more genes encoding molecules of a single BGC. In these forms, the mutant Paenibacillus bacteria have mutations in or deletions in or of one or more genes encoding BGC1, BGC2, BGC4, BGC7, BGC8, BGC12, or BGC13. In one form, the mutation (s) or deletion (s) is in one or more genes encoding molecules from BGC4 when a single BGC is mutated. In some forms, one or more of the following genes in BGC4 is mutated or deleted: CUU60_RS07225, CUU60_RS07220, CUU60_RS07215, CUU60_RS07210, CUU60_RS07205, CUU60_RS07200, CUU60_RS07195, CUU60_RS07190, CUU60_RS07185, CUU60_RS07180 and CUU60_RS07175. For example, one or more of the genes CUU60_RS07225, CUU60_RS07220, CUU60_RS07215, CUU60_RS07210, CUU60_RS07205, CUU60_RS07200, CUU60_RS07195, CUU60_RS07190, CUU60_RS07185, CUU60_RS07180 and CUU60_RS07175 in the BGC4 according to the antiSMASH prediction can be deleted via marker-less genetic engineering method based on CRISPR / Cas9 system.

[0160] In some forms, the mutant Paenibacillus bacteria have disruptive mutations or deletion (s) in one or more genes encoding molecules of multiple BGCs. In some forms, the mutant Paenibacillus bacteria have mutations or deletion (s) in one or more genes encoding molecules of two BGCs. In these forms, the mutations or deletion (s) are in genes encoding molecules of BGC1 and BGC4. In some forms, one or both of the following genes in BGC1 is mutated or deleted: CUU60_RS01235 and CUU60_RS01240. For example, one or both of the genes CUU60_RS01235 and CUU60_RS01240 in the BGC1 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system. In some forms, one or more of the following genes in BGC4 is mutated or deleted: CUU60_RS07225, CUU60_RS07220, CUU60_RS07215, CUU60_RS07210, CUU60_RS07205, CUU60_RS07200, CUU60_RS07195, CUU60_RS07190, CUU60_RS07185, CUU60_RS07180 and CUU60_RS07175.

[0161] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of three BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, and BGC4. In some forms, one or more of the following genes in BGC2 is mutated or deleted: CUU60_RS02235, CUU60_RS02240 and CUU60_RS02245. For example, one or more of the genes CUU60_RS02235, CUU60_RS02240 and CUU60_RS02245 in the BGC2 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system.

[0162] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of four BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, and BGC7. In some forms, one or more of the following genes in BGC7 is mutated or deleted: CUU60_RS16865, CUU60_RS16870, CUU60_RS16875, CUU60_RS16880, CUU60_RS16885 and CUU60_RS16890. For example, one or more of the genes CUU60_RS16865, CUU60_RS16870, CUU60_RS16875, CUU60_RS16880, CUU60_RS16885 and CUU60_RS16890 in the BGC7 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system.

[0163] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of five BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, BGC7 and BGC8. In some forms, one or more of the following genes in BGC8 is mutated or deleted: CUU60_RS20960, CUU60_RS20965, CUU60_RS20970 and CUU60_RS20975. For example, one or more of the genes CUU60_RS20960, CUU60_RS20965, CUU60_RS20970 and CUU60_RS20975 in the BGC8 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system.

[0164] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of six BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, BGC7, BGC8, and BGC12. In some forms, one or more of the following genes in BGC12 is mutated or deleted: CUU60_RS22345, CUU60_RS22350, CUU60_RS22355, CUU60_RS22360 and CUU60_RS22365. For example, one or more of the genes CUU60_RS22345, CUU60_RS22350, CUU60_RS22355, CUU60_RS22360 and CUU60_RS22365 in the BGC12 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system.

[0165] In some forms, the mutant Paenibacillus bacteria have disruptive mutations in one or more genes encoding molecules of seven BGCs. In these forms, the mutations are in genes encoding molecules of BGC1, BGC2, BGC4, BGC7, BGC8, BGC12, and BGC13. In some forms, one or more of the following genes in BGC13 is mutated or deleted: CUU60_RS24780, CUU60_RS24785, CUU60_RS24790, CUU60_RS24795 and CUU60_RS24800. For example, one or more of the genes CUU60_RS24780, CUU60_RS24785, CUU60_RS24790, CUU60_RS24795 and CUU60_RS24800 in the BGC13 according to the antiSMASH prediction were deleted via marker-less genetic engineering method based on CRISPR / Cas9 system.

[0166] In one preferred form, the mutant Paenibacillus bacteria have disruptive mutations and / or deletions in genes encoding molecules in BGC4. In some forms, the mutant Paenibacillus bacteria have one or more mutations or deletions in genes affecting the transcription levels of pmx BGC genes, which encode nonribosomal peptide-synthetase responsible for colistin biosynthesis. In some forms, the pmx mutation (s) is in one or more of pmxA, pmxB, pmxE. In some forms, the pmx mutation (s) increases transcriptional levels of pmxA by about 1.20-fold to about 2.20-fold.

[0167] An exemplary nucleic acid construct includes a plasmid for the deletion of BGC1 i.e., pJOE8999-842-BGC1-gRNA1-frag. In this exemplary form, the nucleic acid sequence of the plasmid is represented by SEQ ID NO: 25.

[0168] 3. Expression Elements

[0169] One or more of a variety of expression elements can be included in the nucleic acid construct to modify expression of the genes of interest in the bacterial cells whose genome is modified by integration of the nucleic acid construct. Suitable expression elements include but are not limited to promoters, enhancers, termination and splicing signal elements.

[0170] In some forms, a termination sequence is included in the nucleic acid construct. An exemplary termination sequence is represented by SEQ ID NO: 15.

[0171] Termination sequence:

[0172] In some forms, one or more selection markers is included in the nucleic acid construct. An exemplary selection marker is the KanR resistance gene. The sequence for the KanR resistance gene is represented by SEQ ID NO: 16.

[0173] Selection marker (KanR) sequence:

[0174] In some forms, the nucleic acid construct includes an origin of replication sequence. Exemplary origin of replication sequences are represented by SEQ ID NO: 17 and SEQ ID NO: 18.

[0175] Origin of replication sequences:

[0176] The nucleic acid construct can include one or more selection marker sequences. Selection marker sequences are genetic elements that are used to identify and select cells that have incorporated the expression construct or the foreign gene of interest. Cells harboring a component of the model generating system such as the CRISPR-Cas expression vector or the transposon expression vector, may include a marker construct, to enable monitoring of target gene knockdown or overexpression, and the production of colistin.

[0177] In some forms, the selection marker sequence encodes a gene which confers resistance to a selective agent e.g., an antibiotic. Exemplary antibiotic resistance genes include but are not limited to kanamycin resistance gene (KanR) , neomycin resistance gene (NeoR) , tetracycline resistance genes (TetR) such as tetA, tetB, tetM, and tetO; hygromycin resistance gene (HygR) , blasticidin resistance gene (BsdR) , puromycin resistance gene (Pac) , zeocin resistance gene (Sh ble) , phleomycin resistance gene (BleoR) , erythromycin resistance gene (ErmR) , streptomycin resistance gene (StrR) , gentamicin resistance gene (GmR) , trimethoprim resistance gene (TmpR) , rifampicin resistance gene (RifR) , β-lactamase (bla) enzymes such as blaTEM, blaSHV, blaCTX-M, and blaNDM; methicillin resistance genes such as mecA and mecC; vancomycin resistance genes such as vanA, vanB, and vanC; erm genes such as ermB, ermC, and ermF; quinolone resistance (qnr) genes such as qnrA, qnrB, and qnrS; sulfonamide (sul) resistance genes such as sul1, sul2, and sul3; and dihydrofolate reductase (dfr) enzymes such as dfrA1, dfrA12, and dfrA17.

[0178] In some forms, the selection marker is a fluorescent marker, fluorophore or fluorescent dye, molecules that enable the identification of cells by absorbing light at specific wavelengths and emit light at longer wavelengths. Suitable fluorescent markers include but are not limited to green fluorescent protein (GFP) , enhanced yellow fluorescent protein (EYFP) , enhanced cyan fluorescent protein (ECFP) , mCherry, fluorescein isothiocyanate (FITC) , tetramethylrhodamine isothiocyanate (TRITC) , Texas red, alexa fluor dyes such as Alexa Fluor 488, Alexa Fluor 546, Alexa Fluor 594, and Alexa Fluor 647, DAPI (4', 6-diamidino-2-phenylindole) , Renilla Reniformis green fluorescent protein, GFPmut2, GFPuv4, yellow fluorescent protein (YFP) , cyan fluorescent protein (CFP) , blue fluorescent protein (BFP) , enhanced blue fluorescent protein (EBFP) , citrine and red fluorescent protein from discosoma (dsRED) , orange fluorescent proteins such as mOrange and mOrange2, mApple and mStrawberry, tdTomato, TagRFP, TagRFP-T, and TagRFP657, and Dendra2.

[0179] B. Fermentation Process for Generation of Mutant Bacteria

[0180] In some forms, batch fermentation is used to generate large quantities of mutant Paenibacillus bacteria. The steps of an exemplary batch fermentation process include:

[0181] i. Filling a 5 L stainless steel fermenter with an initial volume of 3.5 L;

[0182] ii. Selecting the first-round seed cultures by picking the single clone from a fresh plate into 30 mL TSB medium;

[0183] iii. Incubating the culture at a temperature of about 30℃ and a rotational speed of about 200 rpm for 1 day;

[0184] iv. Transferring the culture was to 200 ml of TSB medium at a ratio of about 1: 25 (v / v) ;

[0185] v. Incubating the first-round seed cultures for about 16 h at about 30℃ to produce the second-round seed culture;

[0186] vi. Inoculating the second round of seed cultures into 3.3 L of GSC medium with 6.8 starting pH in the 5 L fermenter.

[0187] Conditions for fermentation include a temperature of about 30℃, constant aeration of about 0.2 vvm, average stirring speed of the stirrer of about 400rpm. Foam control can be performed by adding about 0.1%of DOWFAX DF 103T (Dow Inc., USA) . From 23h, carbon source can be supplemented at a rate of 1g / L / h with 500g / L glucose solution until the fermentation is complete.

[0188] IV. METHODS OF USE

[0189] The disclosed compositions, formulations, and methods provide a readily available source of mutant Paenibacillus bacteria that are valuable for antibiotic production as well as drug discovery, tissue engineering, and research. The disclosed mutant Paenibacillus bacteria are valuable for the production of antibiotics for the treatment of infections caused by gram-negative bacteria. In preferred forms, the mutant Paenibacillus bacteria is a mutant Paenibacillus polymyxa bacteria.

[0190] A. Providing Cells for Propagating High Titers of Bacterial Cells for Antibiotic Production

[0191] It has been established that the mutant Paenibacillus bacteria are capable of producing higher yields of antibiotics with a clean metabolite background. Thus, the disclosed compositions and methods are useful for efficient propagation and harvesting of high titers of mutant Paenibacillus bacteria for antibiotic production. In one preferred form, the antibiotic is Colistin, preferably Colistin A.

[0192] Typically, the mutant Paenibacillus bacteria are produced by (1) isolating and cloning antibiotic biosynthetic genes, (2) transfecting parent bacterial cells with expression vectors containing the cloned antibiotic biosynthetic genes, (3) selection and screening transformed progeny bacteria for the desired mutant Paenibacillus bacteria, and (4) culturing the mutant Paenibacillus bacteria under conditions suitable for production of higher yields of Colistin with clean metabolite backgrounds.

[0193] The term “transfection” or “transfecting” refers to the introduction of a foreign nucleic acid into a cell so that the host cell will express the introduced gene or sequence to produce a desired polypeptide, coded by the introduced gene or sequence. The introduced gene or sequence may also be called a “cloned” or “foreign” gene or sequence, may include regulatory or control sequences, such as start, stop, promoter, signal, secretion, or other sequences used by a cell's genetic machinery. The gene or sequence may include nonfunctional sequences or sequences with no known function. A host cell that receives and expresses introduced DNA or RNA has been “transformed” and is a “transformant” or a “clone. ” The DNA or RNA introduced to a host cell can come from any source, including the same genus or species as the host cell, or from a different genus or species.

[0194] The mutant Paenibacillus bacteria generally produces higher yields of Colistin and its analogs compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria produces about 30%to about 300%higher yields of colistin compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria produce about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 99%, about 100%, about 120%, about 140%, about 160%, about 180%, about 200%, about 220%, about 240%, about 260%, about 280%, or about 300%, higher yield of colistin compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria produces about 200 mg / L to about 1000 mg / L colistin titer. In some forms, the mutant Paenibacillus bacteria produces about 200 mg / L, about 300 mg / L, about 400 mg / L, about 500 mg / L, about 600 mg / L, about 700 mg / L, about 800 mg / L, about 900 mg / L, or about 1000 mg / L total colistin titer.

[0195] In some forms, the mutant Paenibacillus bacteria produce higher yields of Colistin A, Colistin B, or Colistin A and Colistin B. In one preferred form, the mutant Paenibacillus bacteria produce higher yields of Colistin A. In another preferred form, the ratio of Colistin A to the ratio of Colistin B produced by the mutant Paenibacillus bacteria is about 13: 1. In some forms, the amount of Colistin A produced by the mutant Paenibacillus bacteria accounts for about 60%to about 95%of the total Colistin produced by the mutant Paenibacillus bacteria. In some forms, the amount of Colistin A produced by the mutant Paenibacillus bacteria accounts for about 60%, about 70%, about 80%, about 90%, about 95%of the total Colistin produced by the mutant Paenibacillus bacteria.

[0196] In some forms, the amount of Colistin B produced by the mutant Paenibacillus bacteria accounts for about 5%to about 10%of the total Colistin produced by the mutant Paenibacillus bacteria. In some forms, the amount of Colistin A produced by the mutant Paenibacillus bacteria accounts for about 5%, about 6%, about 7%, about 8%, about 9%, or about 10%of the total Colistin produced by the mutant Paenibacillus bacteria.

[0197] The mutant Paenibacillus bacteria also demonstrates production of Colistin with a clean metabolite background i.e., a lower production of unwanted metabolites. This is in part due to the mutant Paenibacillus bacteria having mutations which prevent endogenous genetic regulation that potentially interferes with the production of colistin analogs, and which increases conservation of resources and energy. Exemplary unwanted metabolites include but are not limited to Tridecaptin, Fusaricidin, Paenibacillin, and Paenilan. The structures of Tridecaptin, Fusaricidin, Paenibacillin, and Paenilan are provided above as Formula IV, Formula V, Formula VI, and Formula VII respectively.

[0198] In some forms, the mutant Paenibacillus bacteria demonstrate from about 30%to about 100%lower production of unwanted metabolites compared to the parent bacterium. In some forms, the mutant Paenibacillus bacteria demonstrate about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 99%, or about 100%, lower production of unwanted metabolites compared to the parent bacterium.

[0199] B. Research Purposes

[0200] Millions of lives are tragically lost each year due to the devastating impact of multi-drug resistant (MDR) Gram-negative bacterial infections. In combating this urgent global health crisis, Colistin has emerged as a crucial "last-line antibiotic" with low drug resistance rates against Gram-negative pathogens. By utilizing innovative techniques to generate mutant Colistin producer bacteria and produce Colistin with exceptional titer and purity, effective Active Pharmaceutical Ingredient (API) urgently needed to combat MDR Gram-negative bacterial infections could be produced efficiently.

[0201] Thus, in some forms, the mutant Paenibacillus bacteria can be used for a variety of research purposes. Research involving the mutant Paenibacillus bacteria spans various areas, including but not limited to understanding antibiotic resistance, improve therapeutic strategies, and explore novel antimicrobial approaches.

[0202] In some forms, the mutant Paenibacillus bacteria can be used in research to gain better understanding of the mechanisms of action of Colistin. For example, the mutant Paenibacillus bacteria can be used in research to investigate how Colistin interacts with the membranes of Gram-negative bacteria to cause cell death. This includes exploring its binding to lipopolysaccharides and disrupting the bacterial outer membrane.

[0203] In some forms, the mutant Paenibacillus bacteria can be used to develop new antimicrobial drug strategies. For example, the mutant Paenibacillus bacteria and the colistin compositions can be used to investigate synergistic drug combinations. For example, in some forms, studies can be done exploring combinations of Colistin with other antibiotics or non-antibiotic drugs to enhance efficacy, reduce the likelihood of resistance development, and mitigate toxicity.

[0204] In some forms, the colistin compositions provided herein can be used in studies investigating the modification of Colistin molecules to develop derivatives with reduced toxicity and improved efficacy against resistant bacterial strains.

[0205] In some forms, the colistin compositions provided herein can be used in studies directed to optimizing Colistin for clinical use. For example, the Colistin compositions provided herein can be used in pharmacokinetics and pharmacodynamics (PK / PD) studies to helps optimize dosing strategies to improve efficacy while minimizing adverse effects. In some forms, the Colistin compositions provided herein can be used in research directed towards mitigating toxicity e.g., identifying factors that contribute to Colistin's nephrotoxicity to develop strategies that reduce these side effects.

[0206] C. Providing Antibiotics for the Treatment of Infections

[0207] The disclosed compositions and methods provide increased yields of Colistin with high purity for the treatment of infections caused by gram-negative bacteria. The disclosed compositions and methods address the critical need for improved antibiotics for the treatment of bacterial infections caused by gram-negative bacteria, which are notorious for their resistance to conventional antibiotics.

[0208] Exemplary infections that can be treated using the Colistin compositions and formulations include but are not limited to Pneumonia, urinary tract infections, bloodstream infections, wound infections, and sepsis. In some forms, the Colistin compositions and formulations are suitable for treating infections caused by various bacteria. For example, the Colistin compositions and formulations are suitable for treating infections caused by Acinetobacter baumannii, Klebsiella pneumoniae, Escherichia coli, Pseudomonas aeruginosa, and various Gram-negative multidrug-resistant bacteria.

[0209] Examples

[0210] Example 1: Metabolic Engineering of “Last-Line Antibiotic” Colistin in Paenibacillus polymyxa

[0211] In this study, genome mining was conducted to identify polymyxin BGCs present in the genus Paenibacillus, revealing Paenibacillus polymyxa ATCC 842 as the optimal chassis strain. By applying CRISPR / Cas9 gene editing, the chassis strain ATCC 842 was engineered to efficiently produce Colistin A and B with a clean metabolites background. An engineered strain was generated with a significantly improved colistin titer by inactivating non-essential biosynthetic gene clusters. Iterative inactivation of additional gene clusters led to an engineered strain with a clean metabolites background and an 82%increase in colistin titer compared to the starting strain. We also strengthened the expression of genes involved in the L-2, 4-diaminobutyric acid (L-DABA) pathway using a specific promoter, resulting in a 649.3 mg / L colistin titer, which was improved by 269%compared to the starting strain. Finally, in 5 L bioreactor continuous fermentation, the final colistin titer achieved 1711.7 mg / L.

[0212] This study is the first attempt to engineer a colistin natural producer using rational metabolic engineering. The approach resulted in the construction of a colistin producer with a clean metabolites background, which efficiently produces colistins with the desired analogs ratio suitable for clinical usage.

[0213] Significant results of this study:

[0214] P. polymyxa ATCC 842 was identified as a natural producer of Colistin, capable of producing colistin A and B in the acceptable pharmaceutical ratio;

[0215] By inactivating 7 non-essential BGCs, colistin production was improved by 82%, and unwanted by-products were eliminated during Fermentation;

[0216] Combining BGC inactivation with strengthening of the L-2, 4-DABA pathway resulted in a 269%improvement in colistin production compared to the wild-type strain, achieving a yield of 649.3 mg / L in shake-flask fermentation; and

[0217] Continuous fermentation of engineered strain P19 in 5 L bioreactor enable 1711.7 mg / L colistin titer with 88%Colistin A among its Colistin products.

[0218] In this study, we aimed to improve colistin production through metabolic engineering using CRISPR / Cas9-based genetic editing techniques. Based upon genome mining of polymyxin BGCs from genus Paenibacillus, we selected the native producer strain P. polymyxa ATCC 842 as chassis strain to generate engineered strain with a clean metabolite background. Firstly, we constructed eight mutants with a single inactivation of endogenous BGCs and analyzed the effect of each BGC inactivation on colistin production. Among these BGCs, we identified that mutant P8 with a single inactivation of BGC4 resulted in a remarkable increase in colistin production. To further improve colistin production, we iteratively inactivated seven non-essential BGCs in the chassis strain and generated an engineered strain, P14, with a clean metabolite background and a colistin titer of 320.8 mg / L. We also characterized nine promoters in P. polymyxa ATCC 842 and applied promoter PH4 to strengthen the expression of the genes involved in the precursor amino acid L-2, 4-diaminobutyric acid (L-DABA) pathway, achieving a colistin titer of 649.3 mg / L. In further testing, we scaled up to a 5 L bioreactor and achieved a final colistin titer of 1711.7 mg / L in continuous fermentation. Our results demonstrate the successful metabolic engineering of P. polymyxa ATCC 842 for efficient and clean colistin production, paving the way for further improvements in colistin biosynthesis.

[0219] Materials and methods

[0220] Strains and culture conditions

[0221] All strains used or constructed in this study are listed in Table 1.

[0222] Table 1 Strains Used in This Study

[0223] Escherichia coli strain Trans5α, used for constructing and propagating the plasmid, was acquired from Sangon Biotech, Shanghai, China. The E. coli strains were cultivated in Luria–Bertani (LB) medium (5 g / L yeast extract, 10 g / L tryptone, and 10 g / L NaCl) at 37 ℃. When the E. coli strains contained the pJOE8999 or its derived plasmids, 50 μg / mL kanamycin was supplemented to the medium.

[0224] P. polymyxa ATCC 842 was acquired from the Bacillus Genetic Stock Center (BGSC) . The P. polymyxa strains were cultivated in Tryptic Soy Broth (TSB) medium (Item No: 22092, Sigma-Aldrich, USA) at 30 C. When the P. polymyxa strains contain the pJOE8999 or its derived plasmids, 35 μg / mL kanamycin was additionally supplemented to the medium.

[0225] Plasmids construction

[0226] All plasmids used in this study were listed in Table 2.

[0227] Table 2 Plasmids used in this study

[0228] Two gRNAs were designed by Benchling online platform (https:  / / benchling. com) to construct knock-out vectors for each target gene or BGC.

[0229] Table 3: gRNAs Used in this Study

[0230] The vectors were constructed as previously described. Firstly, the plasmid pJOE8999 was digested by the BsaI enzyme (New England Biolabs, USA) . A pair of complementary oligonucleotides (primers) were heated to 98℃ for 5 minutes and annealed to form the gRNA fragment with sticky ends, and the gRNA fragment was cloned into BsaI-digested pJOE8999 fragment by T4 ligation. The pJOE8999-gRNA plasmid was next digested by the SfiI enzyme (New England Biolabs, USA) . The upstream and downstream homology arms (~500-1000bp) of the target sequence to be knocked out were amplified by PCR, digested by SfiI enzyme, purified, and then cloned into SfiI-digested pJOE8999-gRNA fragment by T4 ligation (Vazyme, China) .

[0231] The plasmid backbone with a kanamycin resistance gene, replication origin in E. coli, and replication origin in P. polymyxa was amplified from plasmid pJOE8999 to construct vectors for promoter characterization. A terminator fragment TrnnB and a sfGFP fragment (with upstream RBS sequence and downstream terminator sequence) were amplified and cloned into the above backbone by HiFi DNA Assembly (New England Biolabs, USA) , thereby constructing the plasmid pMJ (pModified JOE8999) -TrnnB-blank-sfGFP. Different fragments containing potential promoter sequences were amplified by PCR and cloned to the sfGFP upstream of plasmid pMJ-TrnnB-blank-sfGFP by HiFi DNA Assembly.

[0232] For the construction of gene knock-in vectors, the steps were similar to the steps of knock-out vector construction, and the DNA fragment (promoter-ORF-terminator) to be knocked in was cloned into the middle of the upstream and downstream homology arms by HiFi DNA Assembly.

[0233] Genetic manipulation of P. polymyxa via CRISPR / Cas9-based gene editing

[0234] For transformation of plasmids into P. polymyxa, the P. polymyxa strain was firstly cultured in Tryptic Soy Broth-Sorbitol (TSBS) medium (Tryptic Soy Broth with the addition of 91.08g / L sorbitol) at 37℃ for activation. The overnight cultures would be transferred into fresh TSBS medium at a ratio of 1: 50 (v / v) . The cells were harvested at OD600 0.6-0.8, cooled in ice for 30 min, and then washed 3-4 times with cold Wash Buffer (250 mmol / L sucrose, 10%glycerol, 1mmol / L MgCl2, 1mmol / L HEPES) on ice to prepare competently for electro-transformation. In the following step, ~1μg plasmid was added into the competent cells, and the mix was incubated on ice for 5-10min before electro-transformation. After the electroshock (1.8kV) in MicroPulser Electroporation Cuvettes (0.1 cm gap, Bio-Rad, USA) , the bacteria was added with 1 mL TSBS medium, recovered at 30℃ for 3 h and subsequently spread on a Tryptic Soy Agar (TSA) plate containing 35 μg / mL kanamycin. Finally, the strains were cultured at 30℃ for 24 h, and positive colonies were screened by colony PCR.

[0235] To achieve markerless genetic knock-out or knock-in of P. polymyxa, the chassis strains were first transformed with the corresponding knock-out or knock-in vectors and screened for positive colonies. Subsequently, the positive colonies were induced to express Cas9 protein, controlled by a mannose-inducible promoter, in a liquid TSB medium with an additional 20 g / L of mannose at 30℃ for 24 h. The Cas9-induced cultures were then streaked onto fresh TSA plates and screened for positive knock-out / knock-in colonies by colony PCR. To remove the knock-out / knock-in plasmids, positive knock-out / knock-in clones were cultured in antibiotic-free TSB at 25℃ for 2-3 days. These cultures were then streaked onto TSA plates devoid of antibiotics and incubated at 37℃ to eliminate the knock-out / knock-in plasmids.

[0236] Colistin production and analysis

[0237] A two-round seed culture protocol was employed for colistin production in the P. polymyxa ATCC 842 strain and its derived engineered mutants. The first-round seed culture was conducted by picking the single clones of the strains into TSB medium, incubating the culture at a temperature of 30℃ and a rotational speed of 200 rpm for 1 day. The culture was then transferred into 10 ml of TSB medium at a ratio of 1: 50 (v / v) and were allowed to incubate for another day as the second-round seed cultures. The second round of seed cultures were subsequently inoculated into 20 ml of GSC medium (20 g / L dextrose, 20 g / L starch, 20 g / L (NH4) 2SO4, 10 g / L yeast extract, 2.6 g / L K2HPO4, 0.1 g / L FeSO4·7H2O, 0.5 g / L MgSO4, 0.25 g / L NaCl, 9.0 g / L CaCO3) with a starting OD600 of 0.05, and cultured at 30℃, 200 rpm for 2 days. The fermentation liquid was harvested at 48h as samples for subsequent analysis.

[0238] To quantify colistin production, the fermentation liquid was centrifuged at 12,000 rpm for 10 min and the supernatant was collected and diluted four-fold with final concentration of 15%acetonitrile (ACN) as samples for analysis. Quantitative analysis was performed by a 1525 BINARY HPLC system (Waters, USA) . 50 μL of diluted supernatant sample was injected into a Kinetex XB-C18 LC Column (5 μm,  250 x 4.6mm) (Phenomenex, USA) , with following the HPLC gradient: 0 -5 min, 15%B phase; 5 -50 min, 15-50%B phase; 50-55 min, 100%B; 55-60 min, 15%B phase (A=0.5%trifluoroacetic acid in water, B = 0.5%trifluoroacetic acid in acetonitrile) . Colistin was detected at 200 nm on a 2998 Photodiode Array Detector (Waters, USA) .

[0239] The concentration of Colistin in the fermentation liquid was calculated based on the determined correlation equation between the peak area and the concentration of the standard Colistin Sulfate Salt sample (Macklin, Shanghai, China) . All measured colistin concentration data were presented as mean values of 3 biological replicates.

[0240] Biomass determination and promoter strength characterization

[0241] To determine the biomass of P. polymyxa ATCC 842 and its engineered mutants, the cells were cultured under the same conditions as colistin production. At various time points, culture samples were taken, and the cells were harvested by centrifugation and dried. The resulting cell dry weight was then measured to determine the biomass of each strain.

[0242] For the characterization of the strength of promoters in P. polymyxa, the P. polymyxa ATCC 842 strain and its derived engineered mutants were cultured in the same way as for colistin production. The culture liquid at different time points was sampled with 200μL, and the fluorescence intensity of sfGFP expressed by different promoters was measured by BioTek Synergy HTX Multimode Reader (Angilent, USA) , with an excitation wavelength of 470nm and an emission wavelength of 515nm. All quantitative measured data were presented as mean values of 3 biological replicates.

[0243] Transcriptome sequencing

[0244] The P. polymyxa strains were cultured and sampled at two time points (12h and 24h) . The RNA isolation, cDNA synthesis, quality control, and sequencing library construction or real-time quantitative PCR were performed by Shanghai Personalbio Technology Co., Ltd. In brief, for transcriptome sequencing, the stains were grown in 20 mL of GSC medium with shaking at 200 rpm, and the cell culture was sampled at 12h and 24h, respectively. Total RNA was extracted using TRIzolTM Reagent (Thermo Fisher, USA) , followed by cDNA synthesis with PrimeScriptTM 1st stand cDNA Synthesis Kit (Takara Bio, Japan) . The synthesized cDNA was subjected to the sequencing library construction with NEBNext Ultra II RNA Library Prep Kit (NEB, USA) . The constructed library was then sequenced on the NovaSeq 6000 platform (Illumina, USA) with a 150 bp paired-end protocol. Raw sequencing reads were quality controlled by fastp

[0029] with parameters “--detect_adapter_for_pe -l 50 -5 3 -3 3” , including removal of adapter and low-quality reads. Clean reads were then mapped to the reference genome of P. polymyxa ATCC 842 (NCBI accession number: NZ_CP024795) using Bowtie2

[0030] with default parameters. FeatureCounts

[0031] was used to compute mapped reads, with parameters "-p -B -C --fracOverlap 0.2" . Transcripts per million (TPM) was adopted for quantifying the gene expression level, with normalizing for sequencing depth and gene length. The abundance of one BGC was defined as the average abundance of core genes within the BGC. For real-time quantitative PCR, the strains were cultured and sampled at log phase. The housekeeping 16S rRNA gene fragment was used as an internal control for the transcriptional level determination of related genes.

[0245] Batch fermentation

[0246] Batch fermentations were conducted in 5 L stainless steel fermenter (Bailun, China) with an initial volume of 3.5 L. The first-round seed cultures was conducted by picking the single clone from a fresh plate into 30 mL TSB medium, incubating the culture at a temperature of 30℃ and a rotational speed of 200 rpm for 1 day. The culture was then transferred into 200 ml of TSB medium at a ratio of 1: 25 (v / v) by following incubation for 16 h at 30℃ as the second-round seed culture.

[0247] The second round of seed cultures were subsequently inoculated into 3.3 L of GSC medium with 6.8 starting pH in 5 L fermenter. Fermentation was performed at 30℃and constant aeration of 0.2 vvm, while the average stirring speed of the stirrer was maintained at 400rpm. Foam control was performed by adding 0.1%of DOWFAX DF 103T (Dow Inc., USA) . From 23h, carbon source was supplemented at a rate of 1g / L / h with 500g / L glucose solution until the fermentation is complete.

[0248] Genome mining of polymyxin BGCs

[0249] 747 genomes of the genus Paenibacillus were collected from the RefSeq database (downloaded in Aug. 2021) with their BGCs predicted by antiSMASH 6.0. NRPS BGCs were filtered by the following rules to search for polymyxin-like products: 1. BGC must contain at least 9 NRPS modules, responsible for 9 amino acid residues. 2. At least 3 NRPS modules must have their predicted substrate as 2, 4-diaminobutyric acid (Dab) . 3. These 3 NRPS modules must be at positions 1, 4, 5, 8, 9 (1-indexed) , with any permutation of NRPSs, while the order of modules within one NRPS is fixed. 4. BGC must contain a condensation starter domain responsible for loading fatty acyl tail. BGCs that met these rules were considered putative polymyxin-producing BGCs. The detailed result could be referred in the supplementary information.

[0250] Results and Discussion

[0251] Genome mining and metabolic analysis reveal P. polymyxa ATCC 842 as an ideal chassis strain for production of colistins

[0252] Polymyxins belong to a group of lipopeptide compounds containing an N-terminal fatty acyl, a tripeptide exocyclic chain, and a heptapeptide ring (Figure 1B) . These compounds are produced by Paenibacillus spp. and can be further categorized into a series of analogs called polymyxins A -T, which differ in their amino acid compositions (Table 3) . In addition, the same type of polymyxins will be further classified by assigning subscript number based on the fatty acid groups, such as polymyxin E1 -E7 (Table 4) .

[0253] Table 3. Amino acid of position 3, 6, 7, 10 of various polymyxin analogs.

[0254] Table 4. Structure and mass-to-charge ratio of polymyxin E (colistin) analogs.

[0255] To identify an appropriate chassis strain for colistin production, we first explored polymyxins' biosynthetic capacity, a diverse family of nonribosomal peptides (NRPs) found in bacterial genomes. Our approach involved using a rule-based genome-mining method, where we utilized antiSMASH to effectively target polymyxin biosynthetic gene clusters (BGCs) by analyzing the chemical features of predicted NRPs. We analyzed 747 Paenibacillus genomes collected from the RefSeq database in August 2021, and filtered the predicted NRPS BGCs based on their amino acid residues selected by A domains. Our analysis revealed the presence of 70 putative BGCs capable of producing polymyxin products. Interestingly, we also discovered four BGCs that had the potential to produce novel polymyxin products based on their antiSMASH-predicted amino acid residues, which were distinct from previously reported cases. Notably, the predicted amino acid residues and positions 3, 6, 7, and 10 of these 7 biosynthetic gene clusters (BGCs) match well with those of colistins, suggesting that they can encode colistins (Figure 1A and Table 5) .

[0256] Table 5. Amino acid of position 3, 6, 7, 10 of potential novel polymyxin analogs discovered by genome mining. X denotes an unpredicted amino acid.

[0257] Subsequently, we conducted traditional fermentation-based isolation, specifically identifying putative Paenibacillus strains that potentially produce colistins. Among the Bacilli strains examined, P. polymyxa ATCC 842 exhibited active biosynthetic gene clusters (BGCs) of interest, as determined through genome mining and metabolic analysis. Consequently, we selected this strain for further investigation in our study. In order to identify the specific polymyxin analogs produced by P. polymyxa ATCC 842, the supernatant of fermentation liquid was compared to chemically synthesized pure compounds of polymyxin E1, E2, E2-I (Ile) , E3, E4, and E7 using HPLC analysis. Results indicated that the peak retention times of polymyxin analogs produced by ATCC 842 were identical to those of polymyxin E1 and E2 (Figure 1C) , while being clearly distinct from other analogs (Figure 2) . Co-injection analysis further confirmed that the two major peaks of polymyxin analogs produced by ATCC 842 corresponded to polymyxin E1 and E2. This finding was further supported by LC-HRMS analysis, which showed that the mass-to-charge ratios of the two molecules were consistent with the calculated values of polymyxin E1 ( [M+3H] 3+=390.5958) and polymyxin E2 ( [M+3H] 3+=385.9239) (Figure 1D) .

[0258] According to the guidelines of the British Pharmacopeia, the proportions of colistin A (polymyxin E1) and B (polymyxin E2) have to be 50%–75%and 5%–20%, respectively, of the colistin product

[0032] , while the European Pharmacopoeia requires that the sum of polymyxins E1, E2, E3, E1-I and E1-7MOA in the components is greater than 77.0%

[0033] . The ratio of colistin analogs A and B produced by ATCC 842 was about 91%and 9%, respectively, based on the peak areas of HPLC. These proportions closely align with the desired ratios for pharmaceutical colistin products stated in the British Pharmacopeia and European Pharmacopoeia. This suggests that ATCC 842 is an ideal chassis strain for producing pharmaceutical colistins.

[0259] Respective inactivation of non-essential BGCs in P. polymyxa ATCC 842

[0260] In order to minimize the consumption of resources and energy caused by unwanted metabolites, we aimed to interrupt the biosynthesis of non-essential biosynthetic gene clusters (BGCs) in P. polymyxa ATCC 842. This particular strain contains several BGCs that encode diverse secondary metabolites, including Fusaricidin, Tridecaptin, Paenibacillin, Paenilan, and others [34, 35] , which may provide advantage for the strain's survival in natural or unique environments. However, these metabolites are undesirable in industrial settings and have a negative impact on the purity and yield of Colistin.

[0261] We employed antiSMASH 5.1.0

[0036] to detect 15 BGCs within the genome, 12 of which have the potential to biosynthesize peptide-based molecules (Table 6) . Of these, BGCs 1, 5, and 7 are accountable for the synthesis of non-ribosomal peptides (NRPs) , including Tridecaptin

[0037] , Colistin (Polymyxin E)

[0038] , Fusaricidin [39, 40] , respectively. BGCs 8 and 13 Paenibacillin

[0041] and Paenilan

[0042] , respectively, which are ribosomally synthesized and post-translationally modified peptides (RiPPs) . The metabolites encoded by the remaining unknown BGCs are yet to be characterized. To further investigate these BGCs, we performed transcriptome sequencing on wild-type strains at two time points (12h and 24h) . Our results indicate that the core biosynthesis genes of these BGCs were actively transcribed (TPM > 2) , albeit at varying levels among different BGCs (Figure 4A) .

[0262] Table 6. Type, product, and size of 12 NRPS or Ripps BGCs with the potential for peptide-based molecules biosynthesis in P. polymyxa ATCC 842.

[0263] To enhance colistin production, we sought to disrupt the expression of other BGCs as their production could potentially compete with or regulate colistin biosynthesis. Utilizing a CRISPR / Cas9-based marker-less deletion method

[0043] , we successfully inactivated eight BGCs (BGC 1, 2, 4, 7, 8, 12, 13, 15) , resulting in the Generation of P. polymyxa mutants P1 to P8 (Figure 4B) . As expected, the metabolites corresponding to the BGC mutants were eliminated. Interestingly, we observed a 79%improvement in colistin production upon eliminating BGC4, while eliminating BGC15 resulted in a significant reduction. These inactivations were necessary to eliminate the production of unwanted metabolites, conserve resources and energy, and prevent endogenous genetic regulation that could interfere with the production of colistin analogs.

[0264] We examined the growth and transcriptional changes of the mutant strain P3, using P1 and WT as controls, to understand the mechanism behind the observed increase in colistin production due to BGC4 inactivation. First, to determine if the increase in colistin titer was due to a difference in cell growth, we measured the biomass (cell dry weight, CDW) of P3 and the wild-type strain over 48 hours (Figure 5) . Our findings showed no significant difference in the biomass between mutant P3 and the wild-type strain, indicating that the inactivation of BGC4 did not directly affect cell growth.. Next, we analyzed the transcription levels of pmx BGC (i.e., pmxA, pmxB, and pmxE) , which encode the nonribosomal peptide-synthetase responsible for colistin biosynthesis. The qPCR quantification showed that the expression levels of these three genes in mutant P3 were significantly enhanced by 1.63-fold, 2.89-fold, and 4.20-fold, respectively, compared to the wild-type strain (Figure 6A) . Interestingly, we observed that the inactivation of BGC4 increased the production of Colistin and significantly enhanced the production of other secondary metabolites.

[0265] Multiple inactivation of 7 non-essential BGCs

[0266] To further enhance colistin production, we sequentially disrupted multiple BGCs (excluding BGC15 due to its negative impact) , starting with the mutant strain P3 with BGC4 single inactivation. The data revealed that the titer of Colistin was significantly improved in the double and triple BGC inactivation mutants (P9 and P10) (Figure 6B) . Strains P9 and P10 exhibited a 107%and 121%increase in colistin titer, respectively, compared to the wild-type strain. The inactivation of tridecaptin (P9) may have led to increased colistin production by reducing competition for precursors, L-2, 4-DABA, and fatty acids, while the inactivation of BGC2 may reduce metabolic burden.

[0267] However, in subsequent mutant strains with BGC7, BGC8, and BGC12 inactivation, there was no significant improvement in colistin titer, while the knock-out of BGC13 resulted in a slight decrease in colistin production. Considering that secondary metabolites produced by endogenous non-essential BGCs can affect the purity of the product and create unnecessary burdens in downstream processes, we inactivated seven non-essential BGCs to generate the mutant strain P14. After inactivating these BGCs, most peaks of secondary metabolites were eliminated without any impact on the growth status of mutant strain P14 compared to P4 (Figures 4 and 5C) . Consequently, we developed an engineered strain, P14, with a colistin titer of 320.8 mg / L, which is 82%higher than the wild-type strain and has a clean metabolite background.

[0268] Promoters characterization and overexpression of BGC15

[0269] As engineered strain P14 has possessed a clean metabolite background compared with the wild-type strain, its yield of Colistin still needs to be further improved. Since the inactivation of BGC 15 (P8) significantly reduced the production of Colistin, we attempted to characterize promoters and overexpress it.

[0270] According to the previous transcriptome data, a series of upstream sequences of an open reading framework (ORF) with high transcription levels were cloned as potential promoters, and sfGFP was used as a reporter to determine their expression intensity (Figure 7B) . The result showed that promoter H1 possessed the highest transcription intensity. The sfGFP controlled by H1 reached the highest fluorescence intensity in 12h and remained at this level. Meanwhile, the sfGFP controlled by promoter H4 showed a low expression level at the early stage of culture but gradually increased with time and reached a close fluorescence intensity at 48h.

[0271] Subsequently, we selected promoter H4 rather than H1 to express BGC15, a carboxylate reductase, to avoid excessive expression intensity at the early growth stage. The strain expressing sfGFP with promoter H4 was used as the control for comparison (Figure 8) . Surprisingly, overexpression of BGC15 did not increase Colistin yield but reduced it by 35.4%. This unexpected result suggested that BGC15 and its products may be involved in a comprehensive regulatory process rather than acting as an activator for upregulating secondary metabolite production.

[0272] Enhancing L-2, 4-DABA Pathway to Improved Colistin Production

[0273] To enhance the colistin titer, it is essential to increase the supply of amino acid precursors in the chassis strain in addition to promoter refactoring

[0044] . Previous studies have shown that enhancing the supply of precursor amino acids can effectively improve the production of NRP products [27, 28, 45, 46] . This study aimed to strengthen the L-2, 4-DABA pathway, which shares the crucial metabolic pathway of L-Lysine biosynthesis (Figure 7A) . We overexpressed the diaminobutyrate-2-oxoglutarate transaminase (ectB) , aspartate-semialdehyde dehydrogenase (asd) , and two predicted aspartate kinase (lysC and dapG) in the chassis strain ATCC 842, by genomic integration at the location of BGC 8, generating mutant strains P15 to P18 derived from P14. The colistin titer of these mutants was analyzed, and the results showed that the overexpression of ectB resulted in the best improvement of colistin titer, increasing it by 54.2%. Overexpression of lysC or dapG improved colistin titer by 34.2%and 25.6%, respectively. These results suggested that the expression level of diaminobutyrate-2-oxoglutarate transaminase and aspartate kinase might be the rate-limiting step for L-2, 4-DABA biosynthesis. Therefore, by respectively knocking in the target genes at the location of BGC8 and BGC13, we constructed strain P19 with both ectB and lysC overexpression, which achieved a colistin titer of 649.2 mg / L, 269%higher than the initial wild-type strain (Figure 7C) .

[0274] Colistin production of P19 in a 5-L fermenter.

[0275] To evaluate the colistin production capacity of the engineered strain P19, P19 was cultured in a 5-L fermenter, and its colistin production was determined. As the data showed (Figure 9A) , the strain P19 initiated colistin production at 12h and rapidly accumulated Colistin from 24h to 36h. At 72h, the final colistin titer of engineered strain P19 achieved 1711.7 mg / L, 2.64 folds of its titer under shake-flask cultivation. However, the colistin production rate of strain P19 decreased from 36h suddenly. As the colistin titer reached 1247.8 mg / L at 36h, it only increased to 1711.7 mg / L by 72h. This may be due to the fact that after reaching the stationary stage, the metabolic rate of the chassis strain was not as vigorous as that in the log phase, and meanwhile, the accumulation of by-products, such as extracellular polysaccharides (EPS) and butanediol (BDO) , at the late stage of fermentation [15, 47] may also shift the metabolic flux towards Colistin.

[0276] In addition to colistin yield, the metabolites background and proportion of Colistin components of fermentation liquid were also analyzed. The supernatant of the fermentation liquid, without being extracted or purified, was directly injected for HPLC analysis. As the BGCs of a series of peptide-based natural products in the chassis strain were disrupted through genetic manipulation in early engineering, Colistin has been the main product in the supernatant of the fermentation liquid (Figure 9C) . Furthermore, according to peak area, the primary components of its colistin products were Colistin A (~88%) and Colistin B (~6%) (Figure 9B) , which summed up to about 94%among all colistin components.

[0277] These results indicate that engineered strain P. polymyxa P19 has a remarkable colistin production capacity and is a promising chassis strain for producing clinical colistin products with high purity.

[0278] Discussion

[0279] This work identified a P. polymyxa ATCC 842 capable to produce Colistin A. Through BGCs inactivation, promoter characterization, and overexpression of genes involved in L-2, 4-DABA pathway, an engineered strain with a 3.33-fold increase in colistin titer and clean metabolite background was generated. This is the first attempt to perform metabolic engineering for pharmaceutical colistin production in colistin natural producer P. polymyxa.

[0280] This study found that P. polymyxa ATCC 842 can produce polymyxin with a high ratio (~90%) of specific fatty acyl chains (6-Methyloctanoyl, ) . The impact of the length and isomery of fatty acyl chains on the bioactivity and toxicity of NRP compounds has been widely reported, as seen in Daptomycin

[0048] , Polymyxin

[0049] , Surfactin

[0050] , Mycosubtilin

[0051] . However, the selection of fatty acyl chains in NRP production is often non-specific, resulting in mixtures with diverse fatty acyl chains, making it challenging to obtain NRP products with specific fatty acyl chains at high purity [50, 51] . As the selectivity of the starter condensation domain of nonribosomal peptide synthetases mainly affects the fatty acyl chains of nonribosomal lipopeptides, understanding the selectivity mechanism is of extensive interest

[0052] . Thus, we believe that investigating the catalytic selectivity of the condensation domain in the pmx BGC of P. polymyxa ATCC 842 is of research significance.

[0281] We engineered a strain with a higher colistin titer and a clean metabolite background by inactivating BGCs. Interestingly, inactivating BGC 4 and BGC 15 significantly affected colistin production. BGC 4 encodes a complex non-ribosomal peptide synthetase-polyketide synthase (NRPS-PKS) pathway; its products are unstable and hard to isolate. Its inactivation affects other natural product production, suggesting a potential regulatory mechanism among these BGCs that remains to be uncovered. BGC 15 is an NRPS-like BGC that contains a single carboxyl reductase (CAR) , which has been reported to encode diverse signaling molecules [22, 23, 53, 54] associated with virulence or global regulation in bacteria. However, the products of the CAR enzyme in ATCC 842 are diverse and low-yield (unpublished data) , making it challenging to identify the unknown signal molecules. Overexpression of CAR did not directly increase colistin production, indicating an indirect and unknown regulatory mechanism. Unraveling the cryptic network between unexplored BGCs is challenging but essential for metabolic engineering in future studies.

[0282] In addition, we tried to characterize promoters and overexpress the genes of L-2, 4, -DABA pathway in P. polymyxa ATCC 842, and found that overexpression of ectB and lysC would boost colistin production. A previous study has reported that L-2, 4, -DABA was a crucial precursor for polymyxin production, and ectB was the gene directly responsible for its biosynthesis [38, 44] . Meanwhile, lysC is also a gene verified as a rate-limiting step in L-Lysine biosynthesis [55, 56] . Our results are consistent with these studies.

[0283] Despite the potential of P. polymyxa to efficiently produce various value-added compounds [34, 57] , metabolic engineering of this strain has been limited due to its non-model status and suboptimal transformation efficiency [15, 16] . Furthermore, the lack of well-characterized promoters for dynamic regulation or fine-tuning of gene expression in P. polymyxa has hindered the development of synthetic biology tools for this strain. However, as a chassis strain with significant potential for value-added compounds, we believe that further research into its genetic background and developing well-characterized synthetic biology tools will be crucial in promoting systematic and efficient metabolic engineering in P. polymyxa. This will allow for the optimization of metabolic pathways and the production of desired compounds with higher yields and purity. Overall, we anticipate that continued efforts in understanding and improving P. polymyxa's genetic and synthetic biology toolset will pave the way for the development of novel bioprocesses for the production of valuable products.

[0284] Conclusion

[0285] In this investigation, we utilized P. polymyxa ATCC 842 as a chassis strain to perform metabolic engineering for colistin production. The strain naturally produces colistin A with high purity, making it an ideal candidate for engineering. Our initial approach involved developing eight mutants with single inactivation of endogenous NRPS or RiPPs BGCs, and evaluating the effect of each BGC inactivation on colistin production. The most promising result was achieved with BGC4 inactivation, which increased colistin production by 78%. We then iteratively performed inactivation of the remaining 7 BGCs in the chassis strain, which generated an engineered strain, P14, capable of producing 320.8 mg / L colistin titer with a clean metabolite background.

[0286] To further boost colistin production, we characterized nine promoters in the strain and applied the PH4 promoter to increase gene expression involved in the biosynthesis of precursor amino acid L-2, 4-DABA pathway. This led to a colistin titer of 649 mg / L in engineered strain P19. Additionally, we measured the time course of growth and colistin production and found that the engineered strain P19 achieved a 3.33-fold improvement in colistin production compared to the starting strain. Our results demonstrate the first successful case of rational metabolic engineering in P. polymyxa for efficient production of clinical colistin products. This provides new insight for the metabolic engineering of NRP products in non-model chassis strains.

[0287] References

[0288] 1. Storm, D. R., K. S. Rosenthal, and P. E. Swanson, Polymyxin and related peptide antibiotics. Annu Rev Biochem, 1977. 46: p. 723-63.

[0289] 2. Ainsworth, G. C., A. M. Brown, and G. Brownlee, Aerosporin, an antibiotic produced by Bacillus aerosporus Greer. Nature, 1947. 159 (4060) : p. 263.

[0290] 3. Shoji, J., et al., Isolation of two new polymyxin group antibiotics. (Studies on antibiotics from the genus Bacillus. XX) . J Antibiot (Tokyo) , 1977. 30 (12) : p. 1029-34.

[0291] 4. Kato, T. and T. Shoji, The structure of octapeptin D (studies on antibiotics from the genus Bacillus. XXVIII) . J Antibiot (Tokyo) , 1980. 33 (2) : p. 186-91.

[0292] 5. DeCrescenzo Henriksen, E., D. R. Phillips, and J. B. Peterson, Polymyxin E production by P. amylolyticus. Lett Appl Microbiol, 2007. 45 (5) : p. 491-6.

[0293] 6. Shaheen, M., et al., Paenibacillus polymyxa PKB1 produces variants of polymyxin B-type antibiotics. Chem Biol, 2011. 18 (12) : p. 1640-8.

[0294] 7. Nang, S. C., et al., Rescuing the Last-Line Polymyxins: Achievements and Challenges. Pharmacol Rev, 2021. 73 (2) : p. 679-728.

[0295] 8. Rabanal, F. and Y. Cajal, Recent advances and perspectives in the design and development of polymyxins. Nat Prod Rep, 2017. 34 (7) : p. 886-908.

[0296] 9. Shen, M., et al., Two different restriction-modification systems for degrading exogenous DNA in Paenibacillus polymyxa. Biochem Biophys Res Commun, 2018. 504 (4) : p. 927-932.

[0297] 10. Brito, L. F., et al., Magnesium aminoclay-based transformation of Paenibacillus riograndensis and Paenibacillus polymyxa and development of tools for gene expression. Appl Microbiol Biotechnol, 2017. 101 (2) : p. 735-747.

[0298] 11. Rutering, M., et al., Tailor-made exopolysaccharides-CRISPR-Cas9 mediated genome editing in Paenibacillus polymyxa. Synth Biol (Oxf) , 2017. 2 (1) : p. ysx007.

[0299] 12. Brito, L. F., et al., CRISPR interference-based gene repression in the plant growth promoter Paenibacillus sonchi genomovar Riograndensis SBR5. Appl Microbiol Biotechnol, 2020. 104 (11) : p. 5095-5106.

[0300] 13. Kim, M. S., et al., Cytosine Base Editor-Mediated Multiplex Genome Editing to Accelerate Discovery of Novel Antibiotics in Bacillus subtilis and Paenibacillus polymyxa. Front Microbiol, 2021. 12: p. 691839.

[0301] 14. Meliawati, M., C. Teckentrup, and J. Schmid, CRISPR-Cas9-mediated Large Cluster Deletion and Multiplex Genome Editing in Paenibacillus polymyxa. ACS Synth Biol, 2022. 11 (1) : p. 77-84.

[0302] 15. Schilling, C., et al., Engineering of the 2, 3-butanediol pathway of Paenibacillus polymyxa DSM 365. Metab Eng, 2020. 61: p. 381-388.

[0303] 16. Sun, H., et al., Identification and combinatorial engineering of indole-3-acetic acid synthetic pathways in Paenibacillus polymyxa. Biotechnol Biofuels Bioprod, 2022. 15 (1) : p. 81.

[0304] 17. Myronovskyi, M., et al., Generation of a cluster-free Streptomyces albus chassis strains for improved heterologous expression of secondary metabolite clusters. Metab Eng, 2018. 49: p. 316-324.

[0305] 18. Liu, J., et al., Rational construction of genome-reduced Burkholderiales chassis facilitates efficient heterologous production of natural products from proteobacteria. Nat Commun, 2021. 12 (1) : p. 4347.

[0306] 19. Culp, E. J., et al., Hidden antibiotics in actinomycetes can be identified by inactivation of gene clusters for common antibiotics. Nat Biotechnol, 2019. 37 (10) : p. 1149-1154.

[0307] 20. Wu, Q., Y. Zhi, and Y. Xu, Systematically engineering the biosynthesis of a green biosurfactant surfactin by Bacillus subtilis 168. Metab Eng, 2019. 52: p. 87-97.

[0308] 21. Wen, J., et al., Surfactin, a quorum sensing signal molecule, globally affects the carbon metabolism in Bacillus amyloliquefaciens. Metab Eng Commun, 2021. 12: p. e00174.

[0309] 22. Jenul, C., et al., Biosynthesis of fragin is controlled by a novel quorum sensing signal. Nat Commun, 2018. 9 (1) : p. 1297.

[0310] 23. Sieber, S., et al., Mitigation of Pseudomonas syringae virulence by signal inactivation. Sci Adv, 2021. 7 (37) : p. eabg2293.

[0311] 24. Palazzotto, E., et al., Synthetic biology and metabolic engineering of actinomycetes for natural product discovery. Biotechnol Adv, 2019. 37 (6) : p. 107366.

[0312] 25. Wu, F., et al., Modular metabolic engineering of lysine supply for enhanced production of bacitracin in Bacillus licheniformis. Appl Microbiol Biotechnol, 2019. 103 (21-22) : p. 8799-8812.

[0313] 26. Zhu, J., et al., Metabolic Engineering of Aspartic Acid Supply Modules for Enhanced Production of Bacitracin in Bacillus licheniformis. ACS Synth Biol, 2021. 10 (9) : p. 2243-2251.

[0314] 27. Zhu, J., et al., Enhancement of precursor amino acid supplies for improving bacitracin production by activation of branched chain amino acid transporter BrnQ and deletion of its regulator gene lrp in Bacillus licheniformis. Synth Syst Biotechnol, 2018. 3 (4) : p. 236-243.

[0315] 28. Cai, D., et al., Systematic engineering of branch chain amino acid supply modules for the enhanced production of bacitracin from Bacillus licheniformis. Metab

[0316] Eng Commun, 2020. 11: p. e00136.

[0317] 29. Chen, S., et al., fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics, 2018. 34 (17) : p. i884-i890.

[0318] 30. Langmead, B. and S. L. Salzberg, Fast gapped-read alignment with Bowtie 2. Nat Methods, 2012. 9 (4) : p. 357-9.

[0319] 31. Liao, Y., G. K. Smyth, and W. Shi, featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics, 2014. 30 (7) : p. 923-30.

[0320] 32. Kitteringham, G. R., The British Pharmacopoeia Commission. Prog Med Chem, 1969. 6: p. 1-10.

[0321] 33. Monographs for colistimethate sodium and colistin sulfate. European Pharmacopoeia 8.0, 2014. pg. 1961-1962.

[0322] 34. Langendries, S. and S. Goormachtig, Paenibacillus polymyxa, a Jack of all trades. Environ Microbiol, 2021. 23 (10) : p. 5659-5669.

[0323] 35. Zhou, L., et al., Pan-genome analysis of Paenibacillus polymyxa strains reveals the mechanism of plant growth promotion and biocontrol. Antonie Van Leeuwenhoek, 2020. 113 (11) : p. 1539-1558.

[0324] 36. Blin, K., et al., antiSMASH 5.0: updates to the secondary metabolite genome mining pipeline. Nucleic Acids Res, 2019. 47 (W1) : p. W81-W87.

[0325] 37. Lohans, C. T., et al., Biochemical, structural, and genetic characterization of tridecaptin A (1) , an antagonist of Campylobacter jejuni. Chembiochem, 2014. 15 (2) : p. 243-9.

[0326] 38. Choi, S. K., et al., Identification of a polymyxin synthetase gene cluster of Paenibacillus polymyxa and heterologous expression of the gene in Bacillus subtilis. J Bacteriol, 2009. 191 (10) : p. 3350-8.

[0327] 39. Li, J., et al., Use of PCR-targeted mutagenesis to disrupt production of fusaricidin-type antifungal antibiotics in Paenibacillus polymyxa. Appl Environ Microbiol, 2007. 73 (11) : p. 3480-9.

[0328] 40. Li, J. and S. E. Jensen, Nonribosomal biosynthesis of fusaricidins by Paenibacillus polymyxa PKB1 involves direct activation of a D-amino acid. Chem Biol, 2008. 15 (2) : p. 118-27.

[0329] 41. He, Z., et al., Isolation and identification of a Paenibacillus polymyxa strain that coproduces a novel lantibiotic and polymyxin. Appl Environ Microbiol, 2007. 73 (1) : p. 168-78.

[0330] 42. Park, J. E., et al., Identification of the biosynthesis gene cluster for the novel lantibiotic paenilan from Paenibacillus polymyxa E681 and characterization of its product. J Appl Microbiol, 2017. 123 (5) : p. 1133-1147.

[0331] 43. Altenbuchner, J., Editing of the Bacillus subtilis Genome by the CRISPR-Cas9 System. Appl Environ Microbiol, 2016. 82 (17) : p. 5421-7.

[0332] 44. Park, S. Y., et al., Efficient production of polymyxin in the surrogate host Bacillus subtilis by introducing a foreign ectB gene and disrupting the abrB gene. Appl Environ Microbiol, 2012. 78 (12) : p. 4194-9.

[0333] 45. Wang, C., et al., Enhancing surfactin production by using systematic CRISPRi repression to screen amino acid biosynthesis genes in Bacillus subtilis. Microb Cell Fact, 2019. 18 (1) : p. 90.

[0334] 46. Liu, J., et al., Engineering of an Lrp family regulator SACE_Lrp improves erythromycin production in Saccharopolyspora erythraea. Metab Eng, 2017. 39: p. 29-37.

[0335] 47. Okonkwo, C. C., et al., Inactivation of the Levansucrase Gene in Paenibacillus polymyxa DSM 365 Diminishes Exopolysaccharide Biosynthesis during 2, 3-Butanediol Fermentation. Appl Environ Microbiol, 2020. 86 (9) .

[0336] 48. Baltz, R. H., V. Miao, and S. K. Wrigley, Natural products to drugs: daptomycin and related lipopeptide antibiotics. Nat Prod Rep, 2005. 22 (6) : p. 717-41.

[0337] 49. Roberts, K. D., et al., Antimicrobial Activity and Toxicity of the Major Lipopeptide Components of Polymyxin B and Colistin: Last-line Antibiotics against Multidrug-Resistant Gram-negative Bacteria. ACS Infect Dis, 2015. 1 (11) : p. 568-575.

[0338] 50. Razafindralambo, H., et al., Foaming Properties of Lipopeptides Produced by Bacillus subtilis: Effect of Lipid and Peptide Structural Attributes. Journal of Agricultural and Food Chemistry, 1998. 46 (3) : p. 911-916.

[0339] 51. Fickers, P., et al., High-level biosynthesis of the anteiso-C (17) isoform of the antibiotic mycosubtilin in Bacillus subtilis and characterization of its candidacidal activity. Appl Environ Microbiol, 2009. 75 (13) : p. 4636-40.

[0340] 52. Zhong, L., et al., Engineering and elucidation of the lipoinitiation process in nonribosomal peptide biosynthesis. Nat Commun, 2021. 12 (1) : p. 296.

[0341] 53. Arrebola, E., et al., A nonribosomal peptide synthetase gene (mgoA) of Pseudomonas syringae pv. syringae is involved in mangotoxin biosynthesis and is required for full virulence. Mol Plant Microbe Interact, 2007. 20 (5) : p. 500-9.

[0342] 54. Vallet-Gely, I., et al., A secondary metabolite acting as a signalling molecule controls Pseudomonas entomophila virulence. Cell Microbiol, 2010. 12 (11) : p. 1666-79.

[0343] 55. Ikeda, M., et al., A genome-based approach to create a minimally mutated Corynebacterium glutamicum strain for efficient L-lysine production. J Ind Microbiol Biotechnol, 2006. 33 (7) : p. 610-5.

[0344] 56. Ning, Y., et al., Pathway construction and metabolic engineering for fermentative production of ectoine in Escherichia coli. Metab Eng, 2016. 36: p. 10-18.

[0345] 57. Daud, N. S., et al., Paenibacillus polymyxa bioactive compounds for agricultural and biotechnological applications. Biocatalysis and Agricultural Biotechnology, 2019. 18: p. 101092.

[0346] It is understood that the disclosed method and compositions are not limited to the particular methodology, protocols, and reagents described as these can vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims.

[0347] Disclosed are materials, compositions, and components that can be used for, can be used in conjunction with, can be used in preparation for, or are products of the disclosed method and compositions. These and other materials are disclosed herein, and it is understood that when combinations, subsets, interactions, groups, etc. of these materials are disclosed that while specific reference of each various individual and collective combinations and permutation of these compounds may not be explicitly disclosed, each is specifically contemplated and described herein. For example, if a gene is disclosed and discussed and a number of modifications that can be made to a number of components including the gene are discussed, each and every combination and permutation of the gene and the modifications that are possible are specifically contemplated unless specifically indicated to the contrary. Thus, if a class of molecules A, B, and C are disclosed as well as a class of molecules D, E, and F and an example of a combination molecule, A-D is disclosed, then even if each is not individually recited, each is individually and collectively contemplated. Thus, is this example, each of the combinations A-E, A-F, B-D, B-E, B-F, C-D, C-E, and C-F are specifically contemplated and should be considered disclosed from disclosure of A, B, and C; D, E, and F; and the example combination A-D. Likewise, any subset or combination of these is also specifically contemplated and disclosed. Thus, for example, the sub-group of A-E, B-F, and C-E are specifically contemplated and should be considered disclosed from disclosure of A, B, and C; D, E, and F; and the example combination A-D. Further, each of the materials, compositions, components, etc. contemplated and disclosed as above can also be specifically and independently included or excluded from any group, subgroup, list, set, etc. of such materials. These concepts apply to all aspects of this application including, but not limited to, steps in methods of making and using the disclosed compositions. Thus, if there are a variety of additional steps that can be performed it is understood that each of these additional steps can be performed with any specific embodiment or combination of embodiments of the disclosed methods, and that each such combination is specifically contemplated and should be considered disclosed.

[0348] It must be noted that as used herein and in the appended claims, the singular forms “a, ” “an, ” and “the” include plural reference unless the context clearly dictates otherwise. Thus, for example, reference to “a gene” includes a plurality of such genes, reference to “the gene” is a reference to one or more genes and equivalents thereof known to those skilled in the art, and so forth.

[0349] “Optional” or “optionally” means that the subsequently described event, circumstance, or material may or may not occur or be present, and that the description includes instances where the event, circumstance, or material occurs or is present and instances where it does not occur or is not present.

[0350] Unless the context clearly indicates otherwise, use of the word “can” indicates an option or capability of the object or condition referred to. Generally, use of “can” in this way is meant to positively state the option or capability while also leaving open that the option or capability could be absent in other forms or embodiments of the object or condition referred to. Unless the context clearly indicates otherwise, use of the word “may” indicates an option or capability of the object or condition referred to. Generally, use of “may” in this way is meant to positively state the option or capability while also leaving open that the option or capability could be absent in other forms or embodiments of the object or condition referred to. Unless the context clearly indicates otherwise, use of “may” herein does not refer to an unknown or doubtful feature of an object or condition.

[0351] Ranges can be expressed herein as from “about” one particular value, and / or to “about” another particular value. When such a range is expressed, also specifically contemplated and considered disclosed is the range from the one particular value and / or to the other particular value unless the context specifically indicates otherwise. Similarly, when values are expressed as approximations, by use of the antecedent “about, ” it will be understood that the particular value forms another, specifically contemplated embodiment that should be considered disclosed unless the context specifically indicates otherwise. It will be further understood that the endpoints of each of the ranges are significant both in relation to the other endpoint, and independently of the other endpoint unless the context specifically indicates otherwise. It should be understood that all of the individual values and sub-ranges of values contained within an explicitly disclosed range are also specifically contemplated and should be considered disclosed unless the context specifically indicates otherwise. Finally, it should be understood that all ranges refer both to the recited range as a range and as a collection of individual numbers from and including the first endpoint to and including the second endpoint. In the latter case, it should be understood that any of the individual numbers can be selected as one form of the quantity, value, or feature to which the range refers. In this way, a range describes a set of numbers or values from and including the first endpoint to and including the second endpoint from which a single member of the set (i.e. a single number) can be selected as the quantity, value, or feature to which the range refers. The foregoing applies regardless of whether in particular cases some or all of these embodiments are explicitly disclosed.

[0352] Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of skill in the art to which the disclosed method and compositions belong. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present method and compositions, the particularly useful methods, devices, and materials are as described. Nothing herein is to be construed as an admission that the present invention is not entitled to antedate such disclosure by virtue of prior invention. No admission is made that any reference constitutes prior art. The discussion of references states what their authors assert, and applicants reserve the right to challenge the accuracy and pertinency of the cited documents. It will be clearly understood that, although a number of publications are referred to herein, such reference does not constitute an admission that any of these documents forms part of the common general knowledge in the art.

[0353] Although the description of materials, compositions, components, steps, techniques, etc. can include numerous options and alternatives, this should not be construed as, and is not an admission that, such options and alternatives are equivalent to each other or, in particular, are obvious alternatives.

[0354] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the method and compositions described herein. Such equivalents are intended to be encompassed by the following claims.

Claims

1.A mutant Paenibacillus bacterium in which the expression of one or more biosynthetic gene clusters (BGCs) is inactivated, wherein one of the inactivated BGCs is BGC4, wherein, when cultured, the mutant bacterium produces a colistin titer of at least 300 mg / L.2.The mutant bacterium of claim 1, in which the expression of one or more BGCs is increased.3.The mutant bacterium of claim 2, in which the expression of at least one of the BCGs is increased by genetically engineering the BGC to be expressed using the PH4 promoter.4.The mutant bacterium of any one of claims 1-3, in which the expression of one or more L-DABA pathway genes is increased, wherein expression at least one of the L-DABA pathway genes is increased, and wherein, when cultured, the mutant bacterium produces a colistin titer of at least 600 mg / L.5.The mutant bacterium of claim 4, in which the expression of at least one of the L-DABA pathway genes is increased by genetically engineering the BGC to be expressed using the PH4 promoter.6.The mutant bacterium of any one of claims 1-5, wherein the mutant bacterium has a clean metabolite background.7.The mutant bacterium of any one of claims 1-6, wherein the mutant bacterium is Paenibacillus polymyxa.8.The mutant bacterium of any one of claims 1-7, wherein, when cultured, the mutant bacterium produces a Colistin A titer of at least 1700 mg / L.9.The mutant bacterium of claim 8, wherein the Colistin A titer is a greater than the titer of any other polymyxin.10.The mutant bacterium of claim 8 or 9, wherein the Colistin A titer is a greater than the total titer of any other polymyxins.11.The mutant bacterium of any one of claims 8-10, wherein Colistin A is at least 90%by weight of total polymyxin.12.The mutant bacterium of any one of claims 8-11, wherein the mutant bacterium is engineered strain P8, P9, P10, P14, P15, P16, P17, P18, or P19.13.A method of producing Colistin A, the method comprising:culturing the mutant bacterium of any one of claims 1-12 and collecting culture liquid, wherein the culture liquid comprises Colistin A.14.A composition comprising Colistin A, wherein the Colistin A is produced from the mutant bacterium of any one of claims 1-12 or by the method of claim 13.15.A mutant Paenibacillus bacterium, wherein the mutant bacterium is a progeny of a first parental bacterium genetically engineered to inactivate expression of one or more pmx biosynthetic gene clusters (BGCs) , wherein one of the inactivated BGCs is BGC4, wherein, when cultured, the mutant bacterium produces a colistin titer of at least 300 mg / L.16.The mutant bacterium of claim 15, wherein the mutant bacterium is also a progeny of a second parental bacterium genetically engineered to increase expression of one or more BCGs, wherein expression of one or more of the BCGs was increased by genetically engineering the genes to be expressed using the PH4 promoter.17.The mutant bacterium of claim 15 or 16, wherein the mutant bacterium is also a progeny of a third parental bacterium genetically engineered to increase expression of one or more L-DABA pathway genes, wherein expression of one or more of the L-DABA pathway genes was increased by genetically engineering the genes to be expressed using the PH4 promoter, and wherein, when cultured, the mutant bacterium produces a colistin titer of at least 600 mg / L.