Traitement en cours

Veuillez attendre...



Aller à Demande


Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique

[ EN ]



This patent application claims the benefit of U.S. Patent Application Serial No. 62/089,912, filed December 10, 2014, entitled "Genetically Modified Phenylpyruvate Decarboxylase, Processes to Prepare, and Uses Thereof," which is incorporated herein by reference in its entirety.

The invention relates to the field of using biological enzymes to produce C6-C10 compounds such as alcohols, carboxylic acids and alkanes in microbial organisms. More particularly, it relates to the field of using one or more engineered thiamin dependent decarboxylase enzymes to convert a given 2-keto-acid substrate.

Samples of microorganisms expressing one particular embodiment of the genetically modified decarboxylase of the invention representing the M461V variant, as described hereinbelow, have been deposited at the American Tissue Type Collection (ATCC) Patent Repository, 10801 University Blvd., Manassas, VA 20110, on December 9, 2015.

Geopolitical and environmental concerns have sparked researchers around the world in the pursuit of producing petrochemical based products using renewable avenues, including but not limited to fermentation using microorganisms. However, because microorganisms often fail to produce many of the petrochemical based products at economically viable rates or yields, metabolic engineering has been extensively employed, either to build pathways and/or to channel metabolites toward the pathway of interest. Currently, ethanol is the most common biochemical made using microorganisms. However, economically viable methods for producing longer chain alcohols and carboxylic acids are being actively pursued in both the biofuel and chemical industries.

The success in the production of natural amino acids by microbial fermentation has generated significant interest specifically in utilizing the amino acid biosynthetic pathways for producing chemicals of interest, including the longer chain alcohols and carboxylic acids. See, e.g., Becker, J.; Wittmann, C. "Systems and synthetic metabolic engineering for amino acid production - the heartbeat of industrial strain development," Curr. Opin Biotechnol., 2012, 23:718-726; and Becker, J.; Wittmann, C. "Bio-based production of chemicals, materials and fuels - Corynebacterium glutamicum as versatile cell factory," Curr. Opin. Biotechnol., 2012, 23:631-640. The 2-ketoacids, which are key intermediates during amino acid biosynthesis, are amenable to different types of modifications that can be exploited for the biosynthesis of chemicals inside the cells. See, e.g., Gronenberg, L.S.; Marcheschi, R.J.; Liao, J.C. "Next generation biofuel engineering in prokaryotes," Curr. Opin. Chem. Biol., 2013, 17:462-471.

In one example, U.S. Patent 8,232,089 describes a recombinant yeast that expresses an isobutanol-producing metabolic pathway including an Azospirillum brasilense decarboxylase that, when coexpressed with isobutanol producing genes, converts 2-ketoisovalerate to isobutyraldehyde.

In another example, U.S. Patent 8,298,798 describes production of both linear and branched chain alcohols in Escherichia coli (E. coli) cells through the decarboxylation of 2-ketoacids, followed by reduction of the generated aldehyde through expression of Lactobacillus lactis (L. lactis) keto-isovalerate decarboxylase and yeast alcohol dehydrogenase, ADH6. See also, Atsumi, S.; Hanai, T.; Liao, J.C. "Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels" Nature, 2008, 451:86-89; Marcheschi, .J.; Li, H.; Zhang, K.; Noey, E.L; Kim, S.; Chaubey, A.; Houk, K.N.; Liao, J.C. "Synthetic recursive pathway for carbon chain elongation," ACS Chem. Biol., 2012, 7:689-697; and Zhang, K.; Sawaya, M.R.; Eisenberg, D.S.; Liao, J.C. "Expanding metabolism for biosynthesis of nonnatural alcohols," Proc. Natl. Acad. Sci. USA, 2008, 105:20653-20658. The conversion of 2-ketoacid intermediates to carboxylic acids inside cells has been demonstrated via expression of decarboxylase and an aldehyde dehydrogenase. See, e.g., Xiong, M.; Deng, J.; Woodruff, A.P.; Zhu, M.; Zhou, J.; Park, S.W.; Li, H.; Yao, F. "A bio-catalytic approach to aliphatic ketones," 2012, Sci. Rep. 2:311; and Zhang, K.; Woodruff, A.P.; Xiong, M.; Zhou, J.; Dhande, .,K. "A synthetic metabolic pathway for production of the platform chemical isobutyric acid," ChemSusChem, 2011, 4:1068-1070.

The feasibility of extending the length of 2-ketoacids inside the cell via engineering of the LeuA gene product of E. coli has also expanded the range of biochemicals that can be produced from 2-ketoacids. See, e.g., Atsumi, S., ibid., and Zhang, K., ibid. In E. coli, LeuABCD genes extend the length of 2-ketoacids by one carbon unit, as observed during leucine biosynthesis, in which they work together to convert 2-ketoisovalerate (a 5-carbon acid) to 2-ketoisocaproate (a 6-carbon acid). Marcheschi, et al. ACS Chem. Biol., 2012, 7:689-697, describes the expansion of the active site of LeuA and extension of the C4 ketoacid, 2-ketobutyric acid [2-ketobutyrate], to a C9 ketoacid, 2-ketononanoic acid [2-keto-nonanoate].

While it is possible to produce alcohols and carboxylic acids of varied lengths in microorganisms using metabolic engineering, production of a particular C6-C8 alcohol or acid, preferably, in an amount of greater than 20 weight percent (wt%), more preferably greater than 30 wt%, based on total alcohols product, has not been demonstrated to date. Several factors appear to determine the specificity of the alcohol/acid produced from the 2-ketoacids inside the cells. The promiscuity of the decarboxylase in accepting 2-ketoacids of varied lengths leads to aldehydes of varied lengths, which are then oxidized or reduced by the respective coexpressed aldehyde or alcohol dehydrogenase. Thus, higher levels of promiscuity, i.e., lower levels of specificity, lead to higher numbers of products. This, in turn, may mean lower yields of particularly desired, specific products.

Specific production of an alcohol or carboxylic acid via 2-ketoacids may also be unfavorably affected by the level of expression of a decarboxylase with respect to the LeuABCD gene products. Higher levels of a decarboxylase having a broad substrate specificity tend to compete with the LeuA gene product for the 2-ketoacid intermediate and thereby limit the pathway's ability to elongate 2-ketoacids. This may result in formation of a shorter alcohol or carboxylic acid than may be desired, again resulting in an undesirable product and/or product mix.

In general, methods for the improvement of industrial microbial organisms range from the random approach of classical strain improvement (CSI) to the highly rational methods of metabolic engineering. CSI is generally effective for alleviating product inhibition or improving productivity, but is a far less effective approach at generating strains capable of producing entirely new products. Furthermore, CSI is intensive as to both time and resources. To obtain strains with high tolerance to inhibitory fermentation products, it is necessary to continuously screen and select mutants by successively culturing the strain in the media in the presence of increasing inhibitor concentrations. This is usually carried out in conjunction with induced mutagenesis using chemical mutagens and/or ultraviolet (UV) radiation. However, the conventional culture screening process is tedious, time-consuming, and often fruitless.

Metabolic modifications are generally more effective at creating strains that produce new products. This is because genes, and in some cases even entire pathways, can be transferred between organisms (recombinant methods) and/or enzymes can be modified (engineered methods). These methods avoid some of the disadvantages of CSI. Metabolic engineering, a term comprehending both recombinant and engineered methods, is a targeted and often faster approach that is widely used to design strains to achieve higher efficiencies in metabolite overproduction, through alterations in the metabolic flux distribution. Most of this work to date is related to the production of secondary metabolites (such as antibiotics), amino acids (e.g., lysine), and heterologous proteins using organisms with well-studied genetics and physiology (e.g., Escherichia coli, yeast, and hybridoma cells). Stoichiometric analysis of metabolic flux distributions provides a guide to appropriate metabolic modification, optimal medium formulation and feeding strategies, and bioprocess optimization. However, this approach still requires in-depth knowledge of the metabolic and regulatory networks in

the fermentation cells. Although these rational approaches have been successful in cases involving single gene or a few genes within a single gene cluster, they have often been ineffective in cases involving more complex or largely unknown metabolic pathways. This is because such usually target one gene at a time, and thus fail to predict complex interactions among multiple genes in a given pathway.

Enzyme modification is performed by modifying that portion of the genetic code, i.e., the organism's DNA, which corresponds to the expression of that enzyme. Modification of enzymes can lead to entirely new functionality or may be used to improve the specificity or efficiency of desired intermediates or products. Additionally, certain enzymes are known to be promiscuous and may be found performing tasks beyond their known natural roles. Such enzymes may also be modified to perform novel conversions, but to date the success of this approach has been frequently limited to product yields that are not commercially viable. See, e.g., Zhang, K., ibid. Modifying multiple enzymes in a pathway may theoretically be used as a technique to maximize specificity and/or catalytic efficiency.

One example of an organism known to produce octanol under certain conditions is Clostridium. Various species of Clostridium (e.g., C. acetobutylicum, difficile, and kluyveri) are employed in WO 2012135731. That publication describes production of a small amount of n-octanol, along with other products, by an engineered Clostridium species, and ascribes the poor specificity to n-octanol to the organism's ability to express or overexpress beta-ketothiolase (e.g., BktB), acetyl CoA acetyltransferase (e.g., AtoB), 3-hydroxybutyryl-CoA dehydrogenase (e.g., Hbd, from the Clostridium, or PaaHI), crotonase (e.g., Crt), and trans-enoyl-CoA reductase (e.g., Ter). In general, the engineered modifications are to the organism's CoA pathway for the production of higher alcohols, and this pathway avoids the butanol production pathway, found in many species of Clostridium, involving oxygen-sensitive enzymes and intermediates. The amount of n-octanol shown to have been produced via this invention is too small to be commercially viable. See also, e.g., Lee, J.Y.; Jang, Y.S.; Lee, J.; Papoutsakis, E.T.; Lee, S.Y. "Metabolic engineering of Clostridium acetobutylicum M5 for highly selective butanol production," Biotechnol. 2009, 4:1432-1440; and Wang, Y.; Blaschek, H.P. "Optimization of butanol production from tropical maize stalk juice by fermentation with Clostridium beijerinckii," Bioresour. Technol., 2011, 102:9985-9990.

One application of genetic engineering currently being explored is in the energy field. Concerns about the future scarcity, cost, and environmental impact of obtaining and using fossil fuels have stimulated interest in the exploitation of cheap, renewable biomass as alternative sources for both fuels and chemicals made from them. As crude oil prices have become more volatile, bio-based chemicals and industrial products have become attractive alternatives to their petroleum-derived counterparts. Fermentation processes using anaerobic microbial organisms offer a promising path for converting biomass and agricultural wastes into useful products, while at the same time remediating problems that may be encountered in disposal of low-value agricultural commodities and food processing byproducts/wastes. Some of the useful products that can be prepared from low-cost biomass feedstocks are organic acids and alcohols, including octanol. C6-C10 alcohols find particular use as a lower-cost starting material to prepare alkanes, alkenes and aldehydes which are highly desirable feedstock chemicals in a number of industries. These industries include uses as co-monomers for solution polymerizations, and the detergent industry, which uses these precursors to alkylate phenols to produce detergent precursors. These alcohols can also be used as surfactants; as emollients; as thickeners in the cosmetic and food industries; as pesticides; and in a variety of other applications.

In one embodiment the invention provides a process for genetically modifying a microorganism comprising (A) selecting a microorganism that produces a C7-C11 2-ketoacid; and (B) inserting a non-native nucleic acid sequence that encodes an amino acid sequence corresponding to SEQ. ID 4, 8, 14, 16, 18, 28, 30, 32, 34, 36, 38, 40, 42, 46, 52, 54, 56, 62, 64, 66, 68, or 76, or an amino acid sequence that is at least 90 percent homologous thereto; such that a non-native phenylpyruvate decarboxylase is expressed in the microoorganism..

In another embodiment, the invention provides the genetically modified microorganism.

In yet another embodiment, the invention provides a process to prepare a C6-C10 aldehyde, a C6-C10 alcohol, a C6-C10 carboxylic acid, or a C6-C10 alkane, comprising the steps of (A) contacting 2-ketobutyrate or 2-ketoisovalerate, isopropylmalate synthase, isopropylmalate isomerase, and isopropylmalate dehydrogenase, under conditions such that the 2-ketobutyrate or 2-ketoisovalerate is converted to a C7-C11 2-ketoacid; (B) contacting the C7-C11 2-ketoacid and a phenylpyruvate decarboxylase which is expressed by a non-native nucleic acid sequence that encodes an amino acid sequence corresponding to SEQ ID 4, 8, 14, 16, 18, 28, 30, 32, 34, 36, 38, 40, 42, 46, 52, 54, 56, 62, 64, 66, 68, or 76, or an amino acid sequence that is at least 90 percent homologous thereto; under conditions such that the C7-C11 2-ketoacid is converted to a C6-C10 aldehyde having one less carbon atom than the C7-C11 2-ketoacid being converted; and (C) optionally, contacting the C6-C10 aldehyde and (1) an alcohol dehydrogenase under conditions to form a C6-C10 alcohol; or (2) an aldehyde dehydrogenase under conditions to form a C6-C10 carboxylic acid; or (3) a fatty aldehyde decarbonylase under conditions to form a C6-C10 alkane; the process being carried out such that each step and substep occurs independently within or outside of a microbial organism and under aerobic or anaerobic conditions.

In still another embodiment, the invention provides a polypeptide comprising, consisting essentially of, or consisting of an amino acid sequence corresponding to SEQ ID 8, 14, 16, 18, 28, 30, 32, 34, 36, 38, 40, 42, 46, 52, 54, 56, 62, 64, 66, 68, or 76, or at least 90 percent homologous thereto.

In still another embodiment the invention provides a genetically-modified microorganism comprising (A) a source of a 2-ketoacid; (B) a wild type metabolic pathway that converts the 2-ketoacid to a C7-C11 aldehyde; and (C) a non-native phenylpyruvate decarboxylase represented by a nucleic acid sequence encoding an amino acid sequence corresponding with GenBank: Accession No. L26240, or an amino acid sequence that is at least 80 percent homologous thereto; such sequence having been optionally modified by (1) substituting Met-380 with valine; or (2) substituting Phe-385 with valine, leucine or isoleucine; or (3) substituting Met-461 with valine, leucine, alanine or, cysteine; or (4) substituting Phe-465 with valine or leucine; or (5) substituting Phe-532 with glycine, alanine, valine, or leucine; or (6) substituting Gln-536 with valine, leucine, isoleucine, alanine, or glycine; or (7) any combination of two or three substitutions as described in (l)-(6)

For the sequences described hereinafter, each odd-numbered sequence identification number (SEQ ID) shows the nucleotide, or nucleic acid, sequence, and each even-numbered SEQ ID shows the corresponding coded amino acid sequence. The nucleic acid sequence encodes the amino acid sequence, with the term "amino acid sequence" being equivalent to "polypeptide" or "protein." All references to the protein (even-numbered) SEQ IDs acknowledge the fact that it is possible to produce a given amino acid sequence using alternative codons.

SEQ IDs 1 and 2 represents the 13-amino acid histidine tag that is attached at the beginning of the amino acid sequences.

SEQ IDs 3 and 4 represent the wild type Azospirillum brasilense phenylpyruvate decarboxylase gene corresponding to GenBank: Accession No. L26240.

SEQ ID 5 and 6 represent the gene sequence of SEQ ID 3 and 4, but the Met-380 in AbPPDC is replaced with valine at position 380 (based on the amino acid sequence without the his-tag; with the 13-amino acid his-tag, this would be position 393). The modification carries the denomination M380V, therefore adhering to industry standard wherein amino acid modifications are defined as the original single letter amino acid code, followed by the amino acid position, followed by the new amino acid single letter code.

SEQ ID 7 and 8 represent F385L.

SEQ ID 9 and 10 represent F385V.

SEQ ID 11 and 12 represent F385I.

SEQ ID 13 and 14 represent M461C.

SEQ ID 15 and 16 represent M461V.

SEQ ID 17 and 18 represent M461L

SEQ ID 19 and 20 represent M461A.

SEQ ID 21 and 22 represent F465L

SEQ ID 23 and 24 represent F532A.

SEQ ID 25 and 26 represent F532G.

SEQ ID 27 and 28 represent F532V.

SEQ ID 29 and 30 represent F532L

SEQ ID 31 and 32 represent Q536G.

SEQ ID 33 and 34 represent Q536A.

SEQ ID 35 and 36 represent Q536L.

SEQ ID 37 and 38 represent Q536I.

SEQ ID 39 and 40 represent Q536V.

SEQ ID 41 and 42 represent F532V/Q536V.

SEQ ID 43 and 44 represent M380L/M461V.

SEQ ID 45 and 46 represent M380V/M461V.

SEQ ID 47 and 48 represent F385V/M461V.

SEQ ID 49 and 50 represent F385L/M461V.

SEQ ID 51 and 52 represent F532A/Q536V.

SEQ ID 53 and 54 represent F532V/Q536A.

SEQ ID 55 and 56 represent F385L/Q536V.

SEQ ID 57 and 58 represent F385V/Q536V.

SEQ ID 59 and 60 represent M461V/Q536V.

SEQ ID 61 and 62 represent M461L/Q536V.

SEQ ID 63 and 64 represent M461A/Q536V.

SEQ ID 65 and 66 represent M461V/F532V.

SEQ ID 67 and 68 represent F465L/Q536V.

SEQ ID 69 and 70 represent F465V/Q536V.

SEQ ID 71 and 72 represent F465L/F532V.

SEQ ID 73 and 74 represent F532A/Q536A.

SEQ ID 75 and 76 represent M461V/F532V/Q536V.

SEQ ID 77 and 78 represent M380V/M461V/Q536V.

SEQ ID 79 and 80 represent F385L/M461L/Q536V.

SEQ ID 81 and 82 represent M380V/F385V/M461V.

FIGURE 1 illustrates chain elongation by recursive (iterative) activity followed by decarboxylation to an aldehyde that is one carbon atom shorter than the 2-ketoacid preceding it in the iterative pathway.

FIGURE 2 illustrates linear C5-C8 alcohol production from 2-ketobutyrate in vitro using a combination of isopropylmalate synthase, isopropylmalate isomerase, isopropylmalate dehydrogenase, and alcohol dehydrogenase (ADH6) in combination with the F385L variant of AbPPDC (SEQ ID 8).

FIGURE 3 illustrates branched C5-C8 alcohol production from 2-ketoisovalerate in vitro using a combination of isopropylmalate synthase, isopropylmalate isomerase, isopropylmalate dehydrogenase, and alcohol dehydrogenase (ADH6) in combination with the F385L variant of AbPPDC (SEQ ID 8).

FIGURE 4 illustrates branched C5-C8 alcohol production from 3-methyl-2-ketopentanoate in vitro using a combination of isopropylmalate synthase, isopropylmalate isomerase, isopropylmalate dehydrogenase, and alcohol dehydrogenase (ADH6) in combination with the F385L variant of AbPPDC (SEQ ID 8).

FIGURE 5 illustrates the mean alcohol distributions for serum bottle fermentations of E. coli containing the "+1 pathway" enzymes in combination with AbPPDC wild type (WT), AbPPDC variants, KIVD WT (Lactococcus lactis keto-isovalerate decarboxylase), and no decarboxylase. ADH6 is also included in all strain constructs.

In general the present invention includes, among other things, two specific embodiments of a novel phenylpyruvate decarboxylase that may be, in the first embodiment, the expression of an amino acid sequence that is obtained from Azospirillum brasilense (A. brasilense), corresponding to GenBank: Accession No. L26240, or is at least 80 percent (%) homologous thereto. In a second embodiment, the present invention includes the previously defined genetically modified phenylpyruvate decarboxylase, but with further intentional genetic engineering to insert one, two or three modifications of specific amino acids within the sequence, which again serve to modify the catalytic efficiency of the

phenylpyruvate decarboxylase in ways that are in many cases advantageous in carrying out a variety of biosyntheses wherein the phenylpyruvate decarboxylase participates. In particular, either the wild type or the nucleic acid-modified Azospirillum brasilense phenylpyruvate decarboxylase enzymes can be used in combination with isopropylmalate synthase, isopropylmalate isomerase and isopropylmalate dehydrogenase enzymes to produce alcohols, carboxylic acids or alkanes.

As will be shown herein, novel phenylpyruvate decarboxylase enzymes with improved properties over the wild type enzyme of a selected host microorganism were created through genetic modification in one of a variety of ways that are described herein; or is an enzyme represented by an amino acid sequence that is at least 80 % homologous to the A. brasilense phenylpyruvate decarboxylase and includes the same modifications; processes to make it via recombinant, engineered, or technology combining both recombinant and engineered approaches; processes to make C6-C10 alcohols, carboxylic acids and alkanes using the wild type or novel phenylpyruvate decarboxylase; and a genetically modified microbial organism that can express or overexpress this enzyme and can be used to produce C6-C10 alcohols, carboxylic acids and alkanes. As the term is used herein, homology refers to identical or functional correspondence of 80 percent, or more, of the amino acids listed in the sequence, in their given positions.

The novel phenylpyruvate decarboxylase may be used or expressed as part of, in certain particular embodiments, a metabolic pathway that produces acetyl co-A via either an anabolic (e.g., Wood-Ljundahl) or catabolic (e.g., glycolysis, or a pentose phosphate pathway) route, and ultimately takes part in conversion of a C7-C11 2-ketoacid to form the corresponding C6-10 aldehyde having one less carbon. In some embodiments the C6-C10 aldehyde may be further reacted to form a C6-C10 alcohol, carboxylic acid or alkane. Because of the specific alterations in its amino acid sequence that are described herein, the genetically modified phenylpyruvate decarboxylases of the invention offer some significant differences in specificity to various substrates, and this alteration in specificity offers important advantages in terms of product yields and the reduction or elimination of undesirable and/or competing side products.

The invention includes a number of altered amino acid sequences of A. brasilense phenylpyruvate decarboxylase that have been identified as exhibiting improved decarboxylations of C7-C11 2-ketoacids in comparison with the wild type A. brasilense amino acid sequence corresponding to GenBank: Accession No. L26240, which is shown in SEQ. ID 4. Six sites within the wild type sequence have been identified as key to obtaining the improvements. These are: Met-380, Phe-385, Met-461,

Phe-465, Phe-532, Gln-536, and combinations thereof. In each alteration changes are made wherein either valine, leucine, alanine, glycine, or isoleucine are substituted at the identified site(s) for the wild type amino acid, with substitutions varying from single-site (i.e., single amino acid constituting three base pairs) substitution, to a wide variety of multiple-site (from 2 to 5 sites) substitutions defined as "combinations" of the identified sites, preferably from 2 to 3. SEQ. ID 3-82 show amino acid sequences for the many variations produced that include one or more of the substitutions as specified. The substitutions can be summarized as follows: (1) substituting Met-380 with valine; or (2) substituting Phe-385 with valine, leucine or isoleucine; or (3) substituting Met -461 with valine, leucine, alanine, or cysteine; or (4) substituting Phe-465 with valine or leucine; or (5) substituting Phe-532 with glycine, alanine, valine, or leucine; or (6) substituting Gln-536 with valine, leucine, isoleucine, alanine, or glycine; or (7) any combination of three substitutions as described in (l)-(6);

It will be understood by those skilled in the art that the inventive genetically modified phenylpyruvate decarboxylases may be used either in vivo, i.e., by a genetically modified microorganism, or in vitro. In view of this, the terms "genetically modified," or "modified," as used herein, refer to the group of inventive phenylpyruvate decarboxylases having an intentionally altered amino acid sequence, i.e., a "non-wild type" amino acid sequence, or a microbial organism (depending upon placement of either term as an adjective) having a genome that has been intentionally altered as to (at least) the specific, modified decarboxylase(s) described and defined as inventive herein. Such alteration may have been accomplished via recombinant technology, where one or more genes is transferred from a second, different microbial organism into a target microbial organism; or engineered technology, wherein the nucleic acids within the target microbial organism are altered, generally via site-directed mutagenesis, resulting in the conversion of at least one nucleic acid to a different nucleic acid and therefore modification of one or more enzymes. With today's DNA synthesis technologies, recombinant technology can also be accomplished using fully synthetic DNA that is transferred to the target microorganism using conventional methods. Combinations of any of the above methods may also be employed.

The invention further includes a process to prepare C6-C10 aldehydes, C6-C10 carboxylic acids, C6-C10 alkanes, and C6-10 alcohols such as hexanol, heptanol and/or 1-octanol, via contact between a starting substrate and a series of enzymes that include one or more of the genetically-modified phenylpyruvate decarboxylases of the invention to ultimately convert that substrate, using additional enzymes and steps, to the desired C6-C10 aldehyde, alcohol, carboxylic acid, or alkane. This process

may be carried out biosynthetically, in one of the described embodiments of a non-naturally occurring, i.e., genetically engineered, cell, i.e., in a non-naturally occurring microbial organism; or production of the C6-C10 alcohol(s), carboxylic acid(s), or alkane(s), may be carried out via in vitro methodology, typically beginning from a starting point that does not include a microbial organism.

In order to obtain the group of modified phenylpyruvate decarboxylases of the invention, it is desirable, in one embodiment, to perform a protocol similar to that described hereunder. In general the examples show genetic modification involving engineering to alter one or more nucleic acid base(s) in a given codon in order to alter the enzyme of which the nucleic acid base(s) is/are a part. Such may be used simply to produce altered enzyme for, e.g., in vitro assay purposes. In contrast, the genome of a host microbial organism may be preferably altered for a larger scale production strain.

The following methodology, designed for in vitro enzyme production, may be carried out as is generally understood by those skilled in the art. In general, a suitable database, such as GenBank, is used to obtain the genetic codes for the wild type enzyme(s), followed by identification of the codons suitable for modification. This identification may be used as the basis for art-known methods of protein engineering, wherein computer molecular modeling identifies and also enables differentiation of structural locations at which modifications of enzyme/substrate interfaces may be effectively employed. A given desirable modification is then performed, using a molecular biology technique wherein the alteration(s) of the nucleic acid base(s) is/are done via site-directed mutagenesis. The variant-type enzymes must then be subjected to purification to separate out non-targeted proteins, leaving a purified enzyme that will exhibit a higher-than-wild type catalytic efficiency. This can be appropriately assayed in vitro, according to the methodology most suitable for the given particular enzyme. An assayed enzyme that is shown to have a desirable level of catalytic efficiency is thereby confirmed to be the product of a desirable genetic modification, and may be used for in vitro production methods, such as for the in vitro conversion of a C7-C11 2-ketoacid to the corresponding C6-C10 aldehyde having one less carbon (e.g., converting 2-ketononanoate to octanal, or 2-ketooctanoate to heptanal) which can then be reduced, in one embodiment, by contact with an appropriate wild type or non-wild type alcohol dehydrogenase, to form the corresponding C6-C10 alcohol.

As noted hereinabove, the invention may be carried out either in vivo or in vitro. An in vivo approach may be preferred for commercial scale production, although in some cases an in vitro approach may be suitable for commercial scale production. Frequently, an in vitro approach may be particularly convenient for laboratory and general research purposes, such as to carry out enzymatic

assays. For example, desirable microbial organism, useful for large or commercial scale fermentative production of an enzyme-facilitated product, such as, in certain particular embodiments, a C6-C10 alcohol or combination of C6-C10 alcohols, may be prepared. Such preparation may be carried out by inserting the DNA, or pieces of DNA, which encode for the desired improved enzyme, from a first microbial organism into the genome of a second, "host" microbial organism known or believed to possess one or more desired metabolic pathways and/other desired features, such as inhibition-resistant fermentative capability, using recombinant technology. In general the in vivo approach employs such a microbial organism's wild type metabolic pathway(s), first to convert a suitable carbon-containing substrate to pyruvate, and then to convert the pyruvate to 2-ketobutyrate or, alternatively, to 2-ketoisovalerate, in a varying number of steps.

For example, in one embodiment a suitable carbon-containing substrate, such as a C5 or C6 sugar (e.g., glucose, sucrose, pentose, or a combination thereof), may be converted directly to pyruvate via one of the catabolic or anabolic pathways, such as a glycolysis or pentose phosphate pathway. Thereafter the pyruvate may be converted first to L-threonine, via PC (pyruvate carboxylase); AAT (aspartate aminotransferase); ThrABC (ThrA, which is a bifunctional aspartokinase/homoserine dehydrogenase; ThrB, which is homoserine kinase; ThrC, which is threonine synthase; and ASD, which is aspartate semialdehyde dehydrogenase). The L-threonine is then converted to 2-ketobutyrate via llva (threonine dehydratase). In an alternative embodiment, the pyruvate may be converted to 2-ketoisovalerate via the activities of HvBN/llvGM, llvC, and IlvD in leucine biosynthesis. See, also, Zhang, K.; Sawaya, M. .; et al., ibid.

From this point a wild type or genetically modified form of one or more of the three enzymes within the leucine biosynthetic pathway, that are involved in elongating 2-ketoacids, operate to convert the 2-ketobutyrate or 2-ketoisovalerate to a C7-C11 2-ketoacid. These enzymes are generally referred to, without reference to any specific microbial organism, as isopropylmalate synthase, isopropylmalate isomerase, and isopropylmalate dehydrogenase. However, in E. coli specifically, they are referred to as LeuA (GenBank:Accession No. NC 000913.3 Gene ID: 947465), LeuB (GenBank:Accession No. NC 000913.3 Gene ID: 944798), and LeuCD (GenBank:Accession No. NC 000913.3 Gene ID: 945076 and Gene ID: 945642), respectively. One example of this chain elongation is shown in FIGURE 1, wherein 2-ketobutyrate is converted, via a number of steps termed a "+1 pathway," to the C7-C11 2-ketoacid 2-ketononanoate.

In certain particular embodiments the wild type enzymes of leucine biosynthetic pathway involved in extending 2-ketoacids may be modified, in particular by inclusion of at least one exogenous enzyme, enzyme complex, or combination thereof, to convert 2-ketobutyrate first to 2-ketovalerate, then to 2-ketocaproate, then to 2-ketoheptanoate and continuing, if desired, to another elongated 2-ketoacid up to 2-ketoundecanoate, i.e., a desired C7-C11 2-ketoacid, as chain-lengthening occurs. However, it is optionally possible to modify only one or two of the enzymes, enzyme complex, or combination thereof, in order to obtain acceptable or desirable production of the C7-C11 2-ketoacid. These enzymes may include LeuA, LeuB and/or LeuCD, as mentioned hereinabove.

Particularly applicable to modification of this portion of the pathway is the disclosure of co-pending International Patent Application Serial No. PCT/US14/69438, filed December 10, 2014 (Attorney Docket No. 75413-WO-PCT), claiming the benefit of U.S. Provisional Patent Application No. 61/915,040, filed December 12, 2013 (Attorney Docket No. 75413-US-PSP), which are both incorporated herein in their entireties by reference. In certain embodiments, at least a modified isopropylmalate dehydrogenase variant (which is the product of the LeuB gene in E. coli) is selected, or in other embodiments at least a modified LeuA (LeuA') and LeuB' variant is included, preferably, but not necessarily, as described in one or both of the referenced patent applications. It is also preferable to employ other combinations of the LeuA', LeuB' and modified LeuCD (LeuCD') enzymes/enzyme complex. Again, it should be noted that the "Leu" + letter (A, B, CD) designations are specific names for the leucine pathway enzymes of isopropylmalate synthase, isopropylmalate isomerase, and isopropylmalate dehydrogenase in E. coli, while the same or equivalent enzymes in the leucine pathway of other organisms may have different names.

Finally, the inventive genetically modified phenylpyruvate decarboxylase may, in this particular embodiment, serve to convert the C7-C11 2-ketoacid to an aldehyde having one less carbon than the substrate 2-ketoacid. In various embodiments, the resulting C6-C10 aldehyde may find a wide variety of uses, as a product in itself or as a starting or intermediate product for the production of products including the C6-C10 alcohols. Preparation of C6-C10 alcohols may be accomplished via conversion of the C6-C10 aldehyde by an appropriate wild type or genetically modified alcohol dehydrogenase, but other products, such as C6-C10 alkanes, may also be prepared, via the action or expression of a fatty aldehyde decarbonylase, or C6-10 carboxylic acids may be prepared by the action or expression of an aldehyde dehydrogenase. See, e.g., Choi, Y.J.; Lee, S.Y. "Microbial production of short-chain alkanes," Nature, 2013, 502:571-574. Thus, the C6-10 aldehydes are industrially highly useful as excellent intermediate products for preparing a wide variety of other products.

Accordingly, it is anticipated that the inventive family of genetically modified phenylpyruvate decarboxylases will be applicable in a wide variety of industries. Such industries may include, for example, use in fuels, plastics, food, packaging, cosmetics, perfumes, pharmaceuticals, cleaning materials, pollution control, perfumes, drugs, and many others. While there are a number of possible amino acid sequences falling fully within the scope of the claims of the present invention, it is noted that certain amino acid sequences, identified by their sequence identification numbers (SEQ ID) as selected from SEQ ID 34, 36, 38, 40, 42, 46, 62, 68, and 76, are particularly well-suited and preferred for decarboxylating the C7-C11 2-ketoacids.

Example 1

Design of A brasilense phenylpyruvate decaroxylase (AbPPDC) variants with higher catalytic efficiency for 2-ketononanoic acid decarboxylation

A crystal structure model of the ternary complex of AbPPDC with 3-deaza-thiamine diphosphate and 5-phenyl-2-oxovaleric acid (PDB ID Code 2Q5Q) is used to identify residues lining the 2-ketoacid binding pocket within the active site of AbPPDC. See, e.g., Versees, W.; Spaepen, S.; Wood, M.D.; Leeper, F.J.; Vanderleyden, J.; Steyaert, J. "Molecular mechanism of allosteric substrate activation in a thiamine diphosphate-dependent decarboxylase," J. Biol. Chem., 2007, 282:35269-35278. The amino acid sites denominated as Met-380, Met-461, Phe-385, Phe-465, Gln-536 and Phe-532 are selected for substitution experimentation based on their relationship with 5-phenyl-2-oxovaleric acid. Substitutions of one or more sites are made as listed in Table 1 and the variants prepared.

Enzyme F532V replaces Phe-532 in AbPPDC with valine, while enzyme F532L replaces Phe-532 with leucine. Enzyme F385L/M461V replaces Phe-385 with leucine and Met-461 with valine. The remaining A. brasilense phenylpyruvate decarboxylase (AbPPDC) variants in the Table 1 are named according to the amino acid (first letter, with "F" representing "phenylalanine [Phe];" "M" representing "methionine" [Met]; and "Q" representing "glutamine" [Gin]), its position in the amino acid sequence (the number), and the amino acid used as a replacement (last letter, with "L" representing "leucine;" "V" representing "valine;" "A" representing "alanine"; "C" representing "cysteine"; "I" representing "isoleucine"; and "G" representing "glycine.")

Each of the modified AbPPDC variants is expressed and purified, and then tested for activity against the three substrates, which are 2-ketohexanoate (2-KH), 2-ketooctanoate (2-KO) and 2-keto-

nonanoate (2-KN). The 2-KH, 2-KO and 2-KN would be anticipated to form pentanal, heptanal and octanal, respectively, upon decarboxylation by AbPPDC.

The evaluation of the AbPPDC variants is performed in two steps using the high-throughput enzyme assay described hereinbelow. Initially, all the variants are tested for activity against a single high concentration (2 mM) of 2-KH and 2-KN (as shown in Table 1). Following the initial evaluation, detailed kinetic analysis is performed on a select number of variants to determine the maximal rate (kcat), substrate concentration yielding half maximal rate (K0.5, equivalent of KM for enzymes following Michaelis-Menten kinetics), and the catalytic efficiency of the enzyme (kcat/Ko.5) against 2-KO and 2-KN (as shown in Table 2). AbPPDC variants, having higher specificity (higher kcat/Ko.5) for 2-KN, will be efficient in producing octanal and chemicals derived from it inside the cells.

Table 1. Sequence listings and activity of AbPPDC variants

SEQ ID 4 is the amino acid sequence of A. brasilense phenylpyruvate

decarboxylase (GenBank: Accession No. L26240). SEQ ID 6-82 are sequences of proteins designed and expressed in this invention. All the proteins expressed in this invention have 13 additional amino acids at the N-terminus, added as the histidine-tag (shown in

SEQ ID 2).

Example 2

A. Heterologous expression of Azospirillum brasilense phenylpyruvate decarboxylase (AbPPDC) and its engineered variants in E. coli

To evaluate the substrate specificity of the wild type AbPPDC and its variants listed in Table 1, genes of all the proteins are expressed in E. coli cells separately and the protein products are isolated from the cells. The gene sequence of the Azospirillum brasilense phenylpyruvate decarboxylase (GenBank: Accession no. L26240) is downloaded from the NCBI database. Codons of 13 additional amino acids that include six (6) histidines (his) are added upstream of the Met-1 codon of the AbPPDC gene sequence. Such a modification allows expression of a Histidine-tagged AbPPDC having 13 additional amino acids on the N-terminus. The additional amino acids are attached as an aid for purifying the protein in a single step using Ni-NTA chromatography. The entire AbPPDC sequence with 13 additional amino acids (SEQ ID 2) is chemically synthesized and then cloned into the pRSFDuet-1 vector (EMD Biosciences) downstream of the T7 polymerase promoter by Synthetic Genomics, Inc. (San Diego, CA). The final vector is sequenced by Synthetic Genomics, Inc. before shipping.

Genes of the AbPPDC variants listed in Table 1 are either chemically synthesized or generated using New England Biolab's Q5 Site-directed Mutagenesis Kit ( E0554S) and cloned into the pRSFDuet-1 vector. The pRSFDuet-1 vector containing AbPPDC or the AbPPDC variant gene is transformed into E. coli, AbPPDC or its variant, then expressed and eventually purified, as described below.

E. coli expression studies are then conducted using the competent BL21(DE3) cells acquired from

EMD Biosciences. Transformations are performed as per the kit instructions and involve mixing a 50 microliter (μί) aliquot of competent cells with 1 μί of the vector. Cells harboring the AbPPDC expression vector are selected using kanamycin as the marker in the growth medium.

E. coli transformants harboring the AbPPDC or AbPPDC variant expression vector are then selected on Luria-Bertani (LB) broth agar plates containing 50 micrograms per milliliter ^g/mL) of kanamycin. The plates are incubated at 37 degrees Celsius (°C) for 16 hours (h). A starter culture is started by transferring a single colony of transformant into 50 milliliters (mL) of LB medium containing 50 ug/mL of kanamycin and incubated at 37 °C with shaking at 220 revolutions per minute (rpm) overnight. On the next day, 7 mL of

starter culture is inoculated into 800 mL of Terrific Broth (TB) and the culture is incubated at 37 °C until the culture reaches an optical density at 600 nanometers (OD600nm) of 0.5. Isopropyl β-D-l-thiogalacto-pyranoside (IPTG) at a final concentration of 1 mM is added to induce the expression of the AbPPDC or AbPPDC variant genes and the culture is transferred to a 15 °C incubator for 16 hours (h). At the end of 16 h, the culture is centrifuged at 8000 revolutions per minute (rpm) to pelletize the cells. The cell pellet is divided into two aliquots and stored at -80 °C overnight before purification.

An E. coli cell pellet taken from 400 mL of expression culture is suspended in B-PE reagent (Thermo Fisher Scientific, Inc., Rockford, IL) containing 1 μg/mL of DNAse (Thermo Fisher Scientific, Inc., Rockford, IL), 1 μg/mL of lysozyme (Thermo Fisher Scientific, Inc., Rockford, IL), 1 millimolar (mM) of dithiothreitol, and protease inhibitor cocktail (RPI Corp., Mount Prospect, IL). The suspension is rocked gently for 30 minutes (min) at room temperature and centrifuged at 15,000 times gravity (x g) for 20 min to pelletize cell debris. The supernatant is separated and incubated with 5 mL of Co-NTA resin (Thermo Fisher Scientific, Inc., Rockford, IL) that has been pre-equilibrated with an equilibration buffer (50 mM sodium phosphate, pH 8.0, containing 300 mM sodium chloride, 20 mM imidazole, 50 μί protease inhibitor cocktail, and 15 % glycerol). Following an incubation period of 1 h at 4 °C, the enzyme bound resin is washed with 5 volumes of equilibration buffer. AbPPDC or its variants are eluted from the Co-NTA resin with equilibration buffer containing 200 mM imidazole. The eluted proteins are dialyzed against phosphate buffered saline and stored as a 20 % glycerol solution at -20 °C.

B. Determination of the substrate specificity of AbPPDC and AbPPDC variants

The evaluation of the substrate specificities of AbPPDC variants is performed using the methods as described in detail in Example 1.

A high-throughput AbPPDC coupled enzyme assay is developed for evaluating the substrate specificity of AbPPDC variants. The assay involves reducing the aldehyde produced from AbPPDC mediated 2-ketoacid decarboxylation, using an alcohol dehydrogenase (ADH6, GenBank: Accession No. NP 014051.3). The initial velocities of the AbPPDC catalyzed reactions are determined from the rates of oxidation of reduced nicotinamide adenine dinucleotide phosphate (NADPH) occurring during the ADH6 catalyzed reduction of aldehyde.

The HTP screening assay involves incubating 2 mM 2-KH or 2 mM 2-KN with 0.5 mM thiamine diphosphate, 0.35 mM NADPH, 4.7 micrograms ^g) of yeast ADH6 (GenBank: Accession No. NP 014051.3) and 0.3 milligrams per milliliter (mg/mL) bovine serum albumin (BSA) in AbPPDC assay buffer (50 mM 3-(N-morpholino)propanesulfonic acid, pH 6.8, containing 2.5 mM magnesium chloride

(MgCI2)) at 30 °C. The reaction is started by addition at 30 °C of working enzyme stock containing from 0.5 μg to 3.5 μg of AbPPDC variant diluted in AbPPDC assay buffer containing 1 mg/mL BSA. The plate containing the 200 μί of reaction mixture is centrifuged at 2500 x g for 15 sec and the absorbance change of the reaction mixture followed spectrophotometrically at 340 nm on a BioTek™ plate reader, pre-equilibrated at 30 °C. Initial velocity of the enzyme reaction is calculated using the rate of NADPH consumption at 340 nm and the extinction coefficient of NADPH (6.22 mM^cm"1). The activity of all the variants is normalized with the amount of enzyme present in the reaction mixture and expressed as nanomoles per minute per milligram (nmol.min^.mg 1). Protein concentrations for normalizing the activities are determined using the 660 nm total protein assay kit from Pierce Biotechnology Inc., available from Thermo Fisher Scientific, Inc., using BSA as the standard.

The kinetic parameters of the decarboxylation of 2-ketooctanoate (2-KO) and 2-ketononanoate (2-KN) by AbPPDC and its variants are also determined using the same HTP AbPPDC coupled enzyme assay, except that the concentrations of 2-KO or 2-KN are varied from 0 to 4 mM.

For AbPPDC variants exhibiting substrate activation, as evident from a sigmoidal plot of initial velocities versus substrate concentration plot, the kinetic parameters (kcat/ Ko , and kcat/Ko ) of 2-keto-acid decarboxylation are obtained by fitting the data to the Hill equation (shown in the legend of Table 2) using nonlinear regression. For variants following normal saturation kinetics, the kinetic parameters (kcat/ KM, and kcat/KM) are obtained by fitting initial velocities to the Michaelis-Menten equation using nonlinear regression. Nonlinear regression is performed using the GraphPad Prism™ software. Table 2 lists the kinetic parameters of 2-KO and 2-KN decarboxylation by AbPPDC and its variants. The amount of enzyme in the reaction mixture is determined using the Pierce Biotechnology Inc.™ 660 nm total protein assay kit and using BSA as the standard.

Narrowing the substrate specificity of AbPPDC is expected to improve the accumulation of a specific aldehyde and its downstream products. In general AbPPDC prefers bulkier 2-ketoacids, such as 5-phenyl-2-ketopentanoate and phenylpyruvic acid, as evidenced by high catalytic efficiencies with respect to those substrates (See, e.g., Spaepen, S.; Versees, W.; Gocke, D.; Pohl, M.; Steyaert, J.; Vanderleyden, J. "Characterization of phenylpyruvate decarboxylase, involved in auxin production of Azospirillum brasilense," J. Bacteriol., 2007, 189:7626-7633).

AbPPDC and the variants listed in Table 1 are screened for activity against 2 mM 2-keto-hexanoate (2-KH) and 2 mM 2-ketononanoate (2-KN) as substrates. That screening reveals that the wild type AbPPDC catalyzes the decarboxylation of 2-KN, but exhibits poor activity against 2-KH under the assay conditions. All of the AbPPDC variants, also catalyze the decarboxylation of 2-KN, and exhibit relatively low activity against 2-KH (Table 1). Substitution of Gln-536 with alanine, valine, isoleucine or leucine increases the 2-KN decarboxylating activity over that of the wild type enzyme, but also improves activity against 2-KH as a substrate. These results suggest that all of the AbPPDC variants listed in Table 1 can be expressed in an active form in heterologous systems. Furthermore, all of them have significantly higher activity against 2-KN than 2-KH, suggesting that AbPPDC and the variants described herein prefer >C6 2-ketoacids. .

Detailed steady state kinetic analysis is performed on all the enzymes to determine the maximal rate and the catalytic efficiency of decarboxylating 2-ketooctanoate (2-KO) and 2-ketononanoate (2-KN). Both the substrates exhibit hyperbolic and non-hyperbolic kinetics as evident from Table 2. For AbPPDC variants showing non-hyperbolic kinetics, initial velocities of the decarboxylations of 2-KO and 2-KN are fitted to the Hill equation (Table 2 legend) and the maximal rate and the catalytic efficiencies (kcat/Ko ) calculated as shown in Table 2. A Hill coefficient greater than 1 suggests presence of substrate activation with 2-KO and 2-KN. Substrate activations have been reported with AbPPDC and with other decarboxylases. See, also, Spaepen, S., Ibid.

As evident from Table 2, the amino acid substitutions affect the catalytic efficiency of the variants in capturing 2-KO and 2-KN for catalysis in different ways. For some variants, for example, F532V, the catalytic efficiency of decarboxylation of 2-KN and 2-KO is 180 % and 45 %, respectively, in comparison with that of the wild type AbPPDC. This suggests that F532V substitution increases the substrate specificity for 2-KN while decreasing it for 2-KO. The preference of the AbPPDC variants for 2-KN over 2-KO is calculated by taking the ratio of the variant's catalytic efficiencies and is shown in Table 2. As evidenced in Table 2, the specificities of AbPPDC and F532V are 1.8 and 5.6, respectively, indicating that their catalytic efficiency of decarboxylating 2-KN is 1.8 and 5.6 times higher than that of decarboxylating 2-KO. This also indicates that the F532V variant is 3-fold more specific than AbPPDC in preferring 2-KN over 2-KO. Sim ilarly, the preference of F385L for 2-KN over 2-KO is 5-fold higher than that of AbPPDC. This data suggests that the F385L and F532V substitutions improve the substrate specificity for a longer 2-ketoacid (for example 2-KN) over shorter one (for exam ple 2-KO). Thus, the F385L and F532V variants would improve the accumulation of longer (C7-C10) aldehyde based products when 2-ketoacids are being elongated using the "+1 pathway" (FIGURE 1).

Similarly, the specificities of the M461L, F532L, Q536G, Q536L, F532V/Q536V, M380V/M461V, F532A/Q536V, F532V/Q536A, F385L/Q536V, M461V/F532V and M461V/F532V/Q536V variants for 2-KN, in comparison with the specificity of each variant for 2-KO, are 3.3, 4.3, 4.8, 2.7, 3.6, 2.7, 6.8, 4.6, 4.3, 5.4, and 2.1, respectively. This suggests that all of these variants are more specific than AbPPDC in capturing 2-KN for catalysis.

In addition to the specificity of the AbPPDC variant for 2-KN, maximal accumulation of octanal and biochemicals derived from it will also be dependent on the relative efficiencies of the 2-KN producing pathways versus that of the AbPPDC variant. For example, where the efficiency of the engineered 2-ketoacid chain extension pathway (involving the three enzymes, isopropylmalate synthase, isopropylmalate isomerase and isopropylmalate dehydrogenase) in producing 2-KN is relatively low compared to that for producing 2-KO, heptanal formation would result, due to the decarboxylation of 2-KO by AbPPDC variants in combination with reduction in the accumulation of octanal based chemicals inside the cells. Under such circumstances, AbPPDC variant such as F385L would be preferred decarboxylase based upon its relatively high specificity (9.1), coupled with its reduced efficiency as 2-KN decarboxylating catalyst (Table 2).

The results also show that substituting Gln-536 with a hydrophobic amino acid (i.e., glycine, alanine, valine, leucine, or isoleucine) improves the catalytic efficiency of AbPPDC and other specificity enhancing substitutions as shown in Table 2. The Q536V variant is 8- and 5.7-fold more efficient than the wild type enzyme in decarboxylating 2-KO and 2-KN, respectively (Table 2). Similarly, the M461V/F532V/Q536V variant is 27- and 10-fold more efficient than M461V/F532V variant in decarboxylating 2-KO and 2-KN, respectively (Table 2). The M461V/F532V/Q536V variant is about 17- and 20-fold more efficient enzyme than the wild type enzyme in decarboxylating 2-KO and 2-KN, respectively (Table 2). The higher catalytic efficiency of the M461V/F532V/Q536V variant allows effective decarboxylation of 2-KO at 17-fold lower intracellular levels than the wild type enzyme and promotes accumulation of heptanal-derived chemicals, such as heptanol (through coexpression with an alcohol dehydrogenase) or heptanoate (through coexpression of an aldehyde dehydrogenase) inside the cells.

Other substitutions of Gln-536, such as with glycine, alanine, leucine or isoleucine, which also improve the catalytic efficiency of decarboxylation , will also improve the catalytic efficiencies of specificity-enhancing substitutions. This is exhibited by Q536A substitution, which, when added into a F532V variant (with kcat/K0.s = 4.8 mM _1min 1 for 2-KO and kcat/Ka5 = 27 m M _1min 1 for 2-KN), gives rise to a F532V/Q536A variant (with kcat/Ka5 = 8.3 mM _1min 1 for 2-KO and kcat/Ka5 = 38 m M _1min 1 for 2-KN) having 72 % and 40 % higher catalytic efficiencies, respectively, against 2-KO and 2-KN.

In summary, results suggest that the expression of AbPPDC and its genetically modified variants allow efficient decarboxylation of C7-C11, and particularly C7-C9 in this example, 2-ketoacids in vivo, and

thereby allow accumulation of, for example, chemicals derived from aldehydes such as hexanal, heptanal, and/or octanal, inside the cells. Furthermore, modifications of F532, F385, Q.536, M380, M461, F465 either alone or in combination, may give rise to microbial organisms that exhibit specifically improved accumulation of, for example, similarly-derived chemicals inside the cells.

Table 2. Kinetic characterization of AbPPDC and its variants*

*lnitial velocity studies are determined using the HTP coupled assay described in the text. The initial velocities of kcat S^1

all the enzymes except those indicated by § are fitted to the Hill equation (v = h ' h ; v is the initial velocity at a given substrate concentration, 5) using the GraphPad Prism software. kcat, K05, h and kcat/K05 are the maximal velocity, substrate concentration at half the maximal velocity, Hill coefficient and catalytic efficiency respectively.

Results are the mean ± standard error of 2-3 independent experiments. Specificity of the AbPPDC variant is calculated by taking the ratio of catalytic efficiencies (/rcot// 05) of the variant for 2-KN to that for 2-KO.

1 The applied naming convention is that the first letter indicates the amino acid residue which has been altered. F

= phenylalanine [Phe]; Q = glutamine [Gin]; M = methionine [Met]. The number indicates the position in the amino acid sequence (shown are positions 380, 385, 461, 465, 532, and 536, accordingly). The last letter indicates the amino acid residue that is substituted at that location. G = glycine; A = alanine; I = isoleucine; V = valine; L = leucine.

Initial velocities of these variants are fitted to the classical Michaelis-Menton equation.

Example 3

In vitro synthesis of C5-C9 alcohols with the F385L variant (SEQ ID. 8) of Azospirillum brasilense phenylpyruvate decarboxylase (AbPPDC)

In vitro synthesis of linear alcohols with the F385L variant is performed by incubating 0.5 mM

2-ketobutyrate (2-KB) with 0.5 mM thiamine diphosphate, 2.5 mM NAD+, 0.2 milligrams per milliliter (0.2 mg/mL) bovine serum albumin (BSA), 5 mM acetyl coenzyme A, 0.036 mg/mL of the H97A/S139G/N167G/P169A/G462D variant of E. coli isopropylmalate synthase (reported by Marcheschi, . J., et al. "A Synthetic Recursive pathway for carbon chain elongation" ACS Chem. Biol. 7:689-697, 2012) , 0.16 mg/mL of LeuC subunit of isopropylmalate isomerase (GenBank Accession No. NC 000913.3 Gene ID: 945076) and 0.21 mg/mL of LeuD subunit of isopropylmalate isomerase (GenBank Accession No. NC 000913.3 Gene ID: 945642), 0.264 mg/mL of E. coli isopropylmalate dehydrogenase (LeuB; GenBank Accession No. NC_000913.3 Gene ID: 944798), 0.192 mg/mL of L96G/V198A variant of isopropylmalate dehydrogenase (reported in WO2015089127 Al), 0.025 mg/mL of Sacchromyces cerevisiae alcohol dehydrogenase (ADH6, GenBank: Accession No. NP_014051.3) and 0.0054 mg/mL of F385L variant (SEQ ID 8) in in vitro synthesis buffer (50 mM 2-[4-(2-hydroxyethyl)piperazin-l-yljethanesulfonic acid, pH 7.5, containing 30 mM potassium chloride (KCI) and 5 mM magnesium chloride (MgCI2)).

The reaction is initiated with the addition of 2-ketobutyrate to the rest of the reaction mixture. An equal volume of analytical grade toluene (CHROMOSOLVPIus™ for HPLC, > 99%, catalog number 650579) is overlaid on top of the reaction mixture and the solution is incubated at 30 °C. NADPH is added to the aqueous layer to a final concentration of 1 mM after 2.5 hours of incubation at 30 °C. Additional NADPH is added to the aqueous layer to a final concentration of 2 mM after 6 hours of incubation at 30 °C. The reaction is incubated an additional 18 hours at 30 °C, then stopped by freezing at -20°C for 30 minutes. Part of the toluene layer is removed and analyzed using a Gas Chromatograph equipped with a Flame Ionization Detector (FID).

In vitro synthesis of branched alcohols with the F385L variant is performed by replacing 2-ketobutyrate with 0.5 mM 2-ketoisovalerate (2-KIV) or 0.5 mM 3-methyl-2-ketopentanoate (3M-2KP) in the above reaction mixture and performing the experiment as described above.

Alcohols are quantified using a Hewlett Packard (HP) 6890 Series Gas Chromatograph equipped with a Flame Ionization Detector (FID), a model G1513A automatic injector, and a GC AutoSampler

Controller. The analytes are separated using an Agilent J&W DB-FFAP capillary GC column (30 m x 0.320 mm ID x 0.25 μΜ film thickness; catalog number 123-3232, Agilent Technologies, Inc., Wilmington, DE 19808). The initial GC oven temperature is 40 °C, which is held for 1.50 minutes, then is increased to 235 °C with a 40 °C/minute gradient. This gradient gives a total run time of 6.38 minutes. The column flow rate is 4.0 mL/minute, with helium as the carrier gas. The injection volume is 1 μί. The temperature settings for the injector and detector are 225 °C.

The alcohol titers produced from these in vitro synthesis reactions are shown in FIGURES 2, 3 and 4. The results indicate that the F385L variant, in combination with the H97A/S139G/N167G/-P169A/G462D variant of E. coli isopropylmalate synthase (LeuA), E. coli isopropylmalate isomerase (LeuCD), wild type and modified E. coli isopropylmalate dehydrogenase (LeuB), and alcohol dehydrogenase (ADH6), produces elongated C5-C9 alcohols upon incubation with 2-ketobutyrate, 2-ketoisovalerate, or 3-methyl-2-ketopentanoate. Moreover, the results demonstrate the specificity of the F385L variant for longer linear alcohols, wherein 1-octanol represents approximately 60 % of the total alcohols generated upon incubation with 2-ketobutyrate. With branched chain 2-ketoacids (2-ketoisovalerate, KIV, and 3-methyl-2-ketopentanoate, 3M-2-KP), approximately equivalent amounts of 5-methyl-l-hexanol and 6-methyl-l-heptanol are produced upon incubation with 2-ketoisovalerate, and approximately equivalent amounts of 3-methyl-l-pentanol and 5-methyl-l-heptanol are produced upon incubation with 3-methyl-2-ketopentanoate. These results demonstrate that the F385L variant accepts linear as well as branched chain 2-ketoacids as substrates and can produce corresponding linear and branched chain aldehydes which could subsequently be converted to other products such as alcohols, carboxylic acids, or alkanes.

Example 4

In vivo production of C5-C8 alcohols in engineered strains of E. coli using wild type AbPPDC and its variants in combination with the "+1 pathway" enzymes

Escherichia coli (E. coli) MG1655 is engineered to promote long-chain linear alcohol production and to enable gene expression from a T7 promoter. To improve linear alcohol production, ilvBN and ilvlH are inactivated via ARed-mediated homologous recombination as described by Datsenko, KA, Wanner, BL, "One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products," Proc. Natl. Acad. Sci. U.S.A. 2000, 97(12), 6640-6645. The ilvBN and IlvlH genes are involved in branched chain amino acid production, so the inactivation of these genes eliminates the production of branched chain alcohols. The ilvA gene, which is involved in the production of 2-ketobutyrate, is upregulated by replacing its native promoter and ribosome binding site with a strong constitutive promoter and strong ribosome binding site via ARed-mediated homologous recombination as described by Datsenko and Wanner, Ibid. To enable the expression of genes from T7 promoters, the DE3 lysogen is integrated into MG1655 using the ADE3 Lysogenization Kit (EMD Millipore Cat # 69734). The resulting strain genotype is MG1655(DE3) ΔϋνΒΝ ΔϋνΙΗ ilvAup.

C5-C8 alcohols are produced in the engineered E. coli strain through the expression of eight proteins: (1) E.coli isopropylmalate synthase (LeuA); (2) engineered isopropylmalate synthase (described by Marcheschi, et al. ACS Chem. Biol. 2012, 7, 689-697); (3) and (4) two subunits of E. coli isopropylmalate isomerase (LeuCD); (5) isopropylmalate dehydrogenase (LeuB); (6) L96G/V198A variant of E. coli isopropylmalate dehydrogenase (as described in greater detail in co-pending International Patent Application Serial No. PCT/US14/69438, filed December 10, 2014 (Attorney Docket No. 75413-WO-PCT), claiming the benefit of U.S. Provisional Patent Application No. 61/915,040, filed December 12, 2013 (Attorney Docket No. 75413-US-PSP), both of which are incorporated herein in their entireties by reference); (7) AbPPDC or its variants; and (8) S. cerevisiae alcohol dehydrogenase (ADH6). Eleven strains are created in total. One strain is created containing only wild type AbPPDC. As a negative control, a strain with no PPDC is also created. Eight strains containing AbPPDC variants F532V, F358L, F385V, F532V Q536V, M461C, M461V, F385V M461C and F385L M461V are also created. Lastly, a strain containing wild type Lactococcus lactis keto-isovalerate decarboxylase (KIVD; Gene Accession No. AJ746364) is created as a comparison, as prior work has shown that KIVD is capable of producing long-chain alcohols in combination with the "+1 pathway" enzymes. See, e.g., Marcheschi, et al. Ibid.

The Novagen Duet Vector system (EMD Millipore Cat # 71146, 71341, 71340, and 71147), which allows for the simultaneous expression of eight genes using four compatible plasmids, is used to express the genes mentioned above. Each of the four Duet vectors is cloned with two of the eight genes downstream of T7 promoters, and the four Duet vectors are transformed into the engineered E. coli strain. Recombinant strains bearing all of the plasmids are selected for using antibiotics (ampicillin at 25 micrograms per milliliter, μg/mL, chloramphenicol at 17 μg/mL, spectinomycin at 25 μg/mL, and kanamycin at 15 μg/mL) and confirmed with polymerase chain reaction (PCR) using methods known to those skilled in the art. Antibiotics are added at each solid and liquid cultivation step to ensure maintenance of the plasmids. After transformation, plate selection and PCR confirmation, strains are initially cultivated on a Luria-Bertani (LB) agar plate grown at 37 °C. A single agar plate colony is used to inoculate 50 mL of LB medium in a 250 mL shake flask which is cultivated aerobically at 37 °C using an incubator shaker set at 200 rpm.

After 12-16 hours of cultivation in the LB shake flasks, serum bottles are inoculated at 1% v/v to evaluate alcohol production. Serum bottle fermentation medium is prepared using deionized water according to the concentrations shown in Table 1. The medium is filter sterilized, and 20 mL of medium is added to butyl rubber-stoppered 125 mL serum bottles. Prior to media addition, serum bottles are pre-sterilized by autoclaving at 125 °C for 30 minutes using a Steris Amsco Century SV-160H Prevac Sterilizer.

Table 1. Medium composition used to demonstrate alcohol production from E. coli recombinantly engineered to contain the "+1 pathway" in combination with Azospirllum brasilense decarboxylase

(AbPPDC) or its variants.

After inoculation, serum bottle cultures are cultivated at 37 °C with shaking at 200 rpm in an incubator shaker. Approximately three hours after inoculation, the cultures are induced using 0.1 mM of Isopropyl β-D-l-thiogalactopyranoside (IPTG) to ensure expression of all genes. Fermentations are harvested for analysis 24 hours after induction.

At the end of the fermentation, serum bottles are immediately chilled to 4 °C by placing in a refrigerator for 20-30 minutes. Serum bottles are de-capped, and the fermentation broth is quickly poured into a 50 mL conical tube containing 1 mL of a saturated sodium chloride solution and 2 mL of analytical grade toluene (CH OMOSOLV Plus™ for HPLC, > 99.9%, catalog number 650579). The broth-sodium chloride-toluene mixture is vortexed for 30 seconds. A 300 μί aliquot of the toluene extract is then submitted for analysis using GC/FID as described in Example 3.

The mean alcohol distributions for the serum bottles are shown in Figure 5. The results indicate that expression of the wild type Azospirllum brasilense decarboxylase (AbPPDC) in combination with the "+1 pathway" genes and ADH6 results in a functional pathway for the production of linear alcohols ranging from pentanol to octanol. No C5-C8 alcohols are detected in strains without AbPPDC, confirming that the presence of this gene is essential for long-chain alcohol production. Furthermore, the results demonstrate that the strain containing wild type AbPPDC accumulated substantially more hexanol, heptanol and octanol than the strain containing KIVD. No hexanol, heptanol or octanol production is detected in the KIVD strain, but the AbPPDC WT strain produces >2 mg/L, >3 mg/L, and >0.1 mg/L of hexanol, heptanol and octanol, respectively. Approximately 50 % of alcohol production is heptanol and octanol, which is a significant improvement compared to previous work with other decarboxylases that result primarily in pentanol and hexanol production (Marcheschi, et al. ACS Chem. Biol. 2012, 7:689-697). Thus, the use of the AbPPDC decarboxylase appears to shift alcohol production to longer chain lengths, a result which is consistent with the in vitro data contained within Examples 1 and 2.

The additional data in Figure 5 demonstrates that all of the AbPPDC variants have the ability to produce C5-C8 alcohols. Three of the AbPPDC variants, M461C, M461V and F385L/M461V, demonstrate a significant improvement in C5-C8 alcohol production compared to the wild type AbPPDC. AbPPDC variant M461C produces >3 mg/L of hexanol and >5 mg/L of heptanol, representing more than a 40 % improvement compared to the wild type AbPPDC. Most impressively, AbPPDC variant M461V shows more than 2-fold improvements in pentanol, hexanol and heptanol production compared to wild type AbPPDC. The M461V variant also shows a 30 % improvement in terms of octanol titer relative to the wild type AbPPDC. This strain containing the M461V variant produces the highest titers with ~9 mg/L of hexanol and ~8.5 mg/L of heptanol. Variants F532V, F385L, F532V Q536V, M461C, M461V and F385L M461V all show improvements in 1-octanol titer compared to KIVD and the AbPPDC wild type gene. Lastly, AbPPDC variant F385L/M461V shows about 60 % improvement in hexanol and heptanol production compared to the AbPPDC wild type.