ALPHA (1,2) FUCOSYLTRANSFERASE SYNGENES FOR USE IN THE PRODUCTION OF FUCOSYLATED OLIGOSACCHARIDES

FIELD OF THE INVENTION

The invention provides compositions and methods for producing purified oligosaccharides, in particular certain fucosylated oligosaccharides that are typically found in human milk.

BACKGROUND OF THE INVENTION

Human milk contains a diverse and abundant set of neutral and acidic

oligosaccharides. More than 130 different complex oligosaccharides have been identified in human milk, and their structural diversity and abundance is unique to humans. Although these molecules may not be utilized directly by infants for nutrition, they nevertheless serve critical roles in the establishment of a healthy gut microbiome, in the prevention of disease, and in immune function. Prior to the invention described herein, the ability to produce human milk oligosaccharides (HMOS) inexpensively was problematic. For example, their production through chemical synthesis was limited by stereo-specificity issues, precursor availability, product impurities, and high overall cost. As such, there is a pressing need for new strategies to inexpensively manufacture large quantities of HMOS.

SUMMARY OF THE INVENTION

The invention features an efficient and economical method for producing fucosylated oligosaccharides. Such production of a fucosylated oligosaccharide is accomplished using an isolated nucleic acid comprising a sequence encoding a lactose-utilizing a (1,2)

fucosyltransferase gene product (e.g., polypeptide or protein), which is operably linked to one or more heterologous control sequences that direct the production of the recombinant fucosyltransferase gene product in a host production bacterium such as Escherichia coli (E. coif).

The present disclosure provides novel a (1,2) fucosyltransferases (also referred to herein as α(1,2) FTs) that utilize lactose and catalyzes the transfer of an L-fucose sugar from a GDP-fucose donor substrate to an acceptor substrate in an alpha- 1,2-linkage. In a preferred embodiment, the acceptor substrate is an oligosaccharide. The α(1,2) fucosyltransferases identified and described herein are useful for expressing in host bacterium for the production of human milk oligosaccharides (HMOS), such as fucosylated oligosaccharides. Exemplary fucosylated oligosaccharides produced by the methods described herein include 2'- fucosyllactose (2'FL), lactodifucotetraose (LDFT), lacto-N-fucopentaose I (LNF I), or lacto- N-difucohexaose I (LDFH I). The "α(1,2) fucosyltransferases'' disclosed herein encompasses the amino acid sequences of the α(1,2) fucosyltransferases and the nucleic acid sequences that encode the α(1,2) fucosyltransferases, as well as variants and fragments thereof that exhibit α(1,2) fucosyltransferase activity. Also within the invention is a nucleic acid construct comprising an isolated nucleic acid encoding a lactose-accepting a (1,2) fucosyltransferase enzyme, said nucleic acid being optionally operably linked to one or more heterologous control sequences that direct the production of the enzyme in a host bacteria production strain.

The amino acid sequence of the lactose-accepting <x(l,2) fucosyltransferases described herein is at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% identity to Helicobacter pylori 26695 alpha-(l,2) fucosyltransferase (futC or SEQ ID NO: 1). Preferably, the lactose-accepting α(1,2) fucosyltransferases described herein is at least 22% identical to H. pylori FutC, or SEQ ID NO: 1.

In another aspect, the amino acid sequence of the lactose-accepting α(1,2) fucosyltransferases described herein is at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% identity to Bacteroides vulgatus alpha-(l,2) fucosyltransferase (FutN or SEQ ID NO: 3). Preferably, the lactose-accepting <x(l,2) fucosyltransferases described herein is at least 25% identical to B. vlugatos FutN, or SEQ ID NO: 3.

Alternatively, the exogenous a (1,2) fucosyltransferase preferably comprises at least at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% identity to any one of the novel a (1,2)

fucosyltransferases disclosed herein, for example, to the amino acid sequences in Table 1.

Exemplary <x(l,2) fucosyltransferases include, but are not limited to, Prevotella melaninogenica FutO, Clostridium bolteae FutP, Clostridium bolteae +13 FutP,

Lachnospiraceae sp. FutQ, Methanosphaerulapalustris FutR, Tannerellasp. FutS,

Bacteroides caccae FutU, Butyrivibrio FutV, Prevotella sp. FutW, Parabacteroides johnsonii FutX, Akkermansia muciniphilia FutY, Salmonella enterica FutZ, Bacteroides sp. FutZA. For example, the α(1,2) fucosyltransferases comprise the amino acid sequences comprising any one of the following: Prevotella melaninogenica FutO (SEQ ID NO: 10), Clostridium bolteae FutP (SEQ ID NO: 11), Clostridium bolteae +13 FutP (SEQ ID NO: 292),

Lachnospiraceae sp. FutQ (SEQ ID NO: 12), Methanosphaerula palustris FutR (SEQ ID NO: 13), Tannerella sp. FutS (SEQ ID NO: 14), Bacteroides caccae FutU (SEQ ID NO: 15), Butyrivibrio FutV (SEQ ID NO: 16), Prevotella sp. FutW (SEQ ID NO: 17), Parabacteroides johnsonii FutX (SEQ ID NO: 18), Akkermansia muciniphilia FutY (SEQ ID NO: 19), Salmonella enterica FutZ (SEQ ID NO: 20), and Bacteroides sp. FutZA (SEQ ID NO: 21), or a functional variant or fragment thereof. Other exemplary α(1,2) fucosyltransferases include any of the enzymes listed in Table 1, or functional variants or fragments thereof.

The present invention features a method for producing a fucosylated oligosaccharide in a bacterium by providing bacterium that express at least one exogenous lactose-utilizing α(1,2) fucosyltransferase. The amino acid sequence of the exogenous lactose-utilizing ct(l,2) fucosyltransferase is preferably at least 22% identical to H. pylori FutC or at least 25% identical to B. vulgatus FutN. In one aspect, the bacterium also expresses one or more exogenous lactose-utilizing α(1,3) fucosyltransferase enzymes and/or one or more exogenous lactose-utilizing α(1,4) fucosyltransferase enzymes. The combination of fucosyltransferases expressed in the production bacterium is dependent upon the desired fucosylated

oligosaccharide product. The method disclosed herein further includes retrieving the fucosylated oligosaccharide from said bacterium or from a culture supernatant of said bacterium.

Examples of suitable ct(l,3) fucosyltransferase enzymes include, but are not limited to Helicobacter pylori 26695 futA gene (GenBank Accession Number HV532291

(GI:365791177), incorporated herein by reference), H. hepaticus Hh0072, Kpylori 11639 FucT, and Epylori UA948 FucTa (e.g., GenBank Accession Number AF194963

(GI:28436396), incorporated herein by reference)(Rasko, D. A., Wang, G., Palcic, M. M. & Taylor, D. E. J Biol Chem 275, 4988-4994 (2000)). Examples of suitable α(1,4)

fucosyltransferase enzymes include, but are not limited to H. pylori UA948 FucTa (which has has relaxed acceptor specificity and is able to generate both α(1,3)- and α(1,4)-fucosyl linkages). An example of an enzyme possessing only α(1,4) fucosyltransferase activity is given by the FucT ΠΙ enzyme from Helicobacter pylori strain DMS6709 (e.g., GenBank Accession Number AY450598.1 (GI:40646733), incorporated herein by reference) (S.

Rabbani, V. Miksa, B. Wipf, B. Ernst, Glycobiology 15, 1076-83 (2005). ) The invention also features a nucleic acid construct or a vector comprising a nucleic acid enconding at least one a (1,2) fucosyltransferase or variant, or fragment thereof, as described herein. The vector can further include one or more regulatory elements, e.g., a heterologous promoter. By "heterologous" is meant that the control sequence and protein- encoding sequence originate from different bacterial strains. The regulatory elements can be operably linked to a gene encoding a protein, a gene construct encoding a fusion protein gene, or a series of genes linked in an operon in order to express the fusion protein. In yet another aspect, the invention comprises an isolated recombinant cell, e.g., a bacterial cell containing an aforementioned nucleic acid molecule or vector. The nucleic acid is optionally integrated into the genome of the host bacterium. In some embodiments, the nucleic acid construct also further comprises one or more α(1,3) fucosyltransferases and/or α(1,4) fucosyltransferases. Alternatively, the a (1,2) fucosyltransferase also exhibits α(1,3) fucosyltransferase and/or α(1,4) fucosyltransferase activity.

The bacterium utilized in the production methods described herein is genetically engineered to increase the efficiency and yield of fucosylated oligosaccharide products. For example, the host production bacterium is characterized as having a reduced level of β- galactosidase activity, a defective colanic acid synthesis pathway, an inactivated ATP- dependent intracellular protease, an inactivated lac A, or a combination thereof. In one embodiment, the bacterium is characterized as having a reduced level of β-galactosidase activity, a defective colanic acid synthesis pathway, an inactivated ATP-dependent intracellular protease, and an inactivated lac A.

As used herein, an "inactivated" or "inactivation of a" gene, encoded gene product (i.e„ polypeptide), or pathway refers to reducing or eliminating the expression (i.e. , transcription or translation), protein level (i.e., translation, rate of degradation), or enzymatic activity of the gene, gene product, or pathway. In the instance where a pathway is inactivated, preferably one enzyme or polypeptide in the pathway exhibits reduced or negligible activity. For example, the enzyme in the pathway is altered, deleted or mutated such that the product of the pathway is produced at low levels compared to a wild-type bacterium or an intact pathway. Alternatively, the product of the pathway is not produced. Inactivation of a gene is achieved by deletion or mutation of the gene or regulatory elements of the gene such that the gene is no longer transcribed or translated. Inactivation of a polypeptide can be achieved by deletion or mutation of the gene that encodes the gene product or mutation of the polypeptide to disrupt its activity. Inactivating mutations include additions, deletions or substitutions of one or more nucleotides or amino acids of a nucleic acid or amino acid sequence that results in the reduction or elimination of the expression or activity of the gene or polypeptide. In other embodiments, inactivation of a polypeptide is achieved through the addition of exogenous sequences (i.e. , tags) to the N or C-terminus of the polypeptide such that the activity of the polypeptide is reduced or eliminated (i.e., by steric hindrance).

A host bacterium suitable for the production systems described herein exhibits an enhanced or increased cytoplasmic or intracellular pool of lactose and/or GDP-fucose. For example, the bacterium is E. coli and endogenous E. coli metabolic pathways and genes are manipulated in ways that result in the generation of increased cytoplasmic concentrations of lactose and/or GDP-fucose, as compared to levels found in wild type E. coli. Preferably, the bacterium accumulates an increased intracellular lactose pool and an increased intracellular GDP-fucose pool. For example, the bacteria contain at least 10%, 20%, 50%, or 2X, 5X, 10X or more of the levels of intracellular lactose and/or intracellular GDP-fucose compared to a corresponding wild type bacteria that lacks the genetic modifications described herein.

Increased intracellular concentration of lactose in the host bacterium compared to wild-type bacterium is achieved by manipulation of genes and pathways involved in lactose import, export and catabolism. In particular, described herein are methods of increasing intracellular lactose levels in E. coli genetically engineered to produce a human milk oligosaccharide by simultaneous deletion of the endogenous β-galactosidase gene (lacZ) and the lactose operon repressor gene QacJ). During construction of this deletion, the laclq promoter is placed immediately upstream of (contiguous with) the lactose permease gene, lacY, i.e., the sequence of the laclq promoter is directly upstream and adjacent to the start of the sequence encoding the lacYgene, such that the lacY gene is under transcriptional regulation by the laclq promoter. The modified strain maintains its ability to transport lactose from the culture medium (via LacY), but is deleted for the wild-type chromosomal copy of the lacZ (encoding β-galactosidase) gene responsible for lactose catabolism. Thus, an intracellular lactose pool is created when the modified strain is cultured in the presence of exogenous lactose.

Another method for increasing the intracellular concentration of lactose in E. coli involves inactivation of the lacA gene. A inactivating mutation, null mutation, or deletion of lacA prevents the formation of mtracellular acetyl-lactose, which not only removes this molecule as a contaminant from subsequent purifications, but also eliminates Ecoli's ability to export excess lactose from its cytoplasm (Danchin A. Cells need safety valves. Bioessays 2009, Jul;31(7):769-73.), thus greatly facilitating purposeful manipulations of the Rcoli intracellular lactose pool.

The invention also provides methods for increasing intracellular levels of GDP- fucose in a bacterium by manipulating the organism's endogenous colanic acid biosynthesis pathway. This increase is achieved through a number of genetic modifications of endogenous E coli genes involved either directly in colanic acid precursor biosynthesis, or in overall control of the colanic acid synthetic regulon. Particularly preferred is inactivation of the genes or encoded polypeptides that act in the colanic acid synthesis pathway after the production of GDP-fucose (the donor substrate) and before the generation of colanic acid. Exemplary colanic acid synthesis genes include, but are not limited to: a weal gene, (e.g., GenBank Accession Number (amino acid) BAA15900 (GI: 1736749), incorporated herein by reference), a wcaA gene (e.g., GenBank Accession Number (amino acid) BAA15912.1 (GI:1736762), incorporated herein by reference), a wcaC gene (e.g., GenBank Accession Number (amino acid) BAE76574.1 (GI: 85675203), incorporated herein by reference), a wcaE gene (e.g., GenBank Accession Number (amino acid) BAE76572.1 (GI:85675201), incorporated herein by reference), a weal gene (e.g., GenBank Accession Number (amino acid) BAA15906.1 (GI:1736756), incorporated herein by reference), a wcaL gene (e.g., GenBank Accession Number (amino acid) BAAl 5898.1 (GI: 1736747), incorporated herein by reference), a wcaB gene (e.g., GenBank Accession Number (amino acid) BAA15911.1 (GI:1736761), incorporated herein by reference), a wcaF gene (e.g., GenBank Accession Number (amino acid) BAA15910.1 (GI:1736760), incorporated herein by reference), a wzxE gene (e.g., GenBank Accession Number (amino acid) BAE77506.1 (GI:85676256), incorporated herein by reference), a wzxC gene, (e.g., GenBank Accession Number (amino acid) BAAl 5899 (GI: 1736748), incorporated herein by reference), a wcaD gene, (e.g., GenBank Accession Number (amino acid) BAE76573 (GI:85675202), incorporated herein by reference), a wza gene (e.g., GenBank Accession Number (amino acid) BAE76576

(GI: 85675205), incorporated herein by reference), a wzb gene (e.g., GenBank Accession Number (amino acid) BAE76575 (GI:85675204), incorporated herein by reference), and a wzc gene (e.g., GenBank Accession Number (amino acid) BAA15913 (GI:1736763), incorporated herein by reference).

Preferably, a host bacterium, such as E. coli, is genetically engineered to produce a human milk oligosaccharide by the inactivation of the wcaJ gene, which encoding the UDP- glucose lipid carrier transferase. The inactivation of the wcaJ gene can be by deletion of the gene, a null mutation, or inactivating mutation of the wcaJ gene, such that the activity of the encoded wcaJ is reduced or eliminated compared to wild-type E coli. In a wcaJ null background, GDP-fucose accumulates in the E. coli cytoplasm.

Over-expression of a positive regulator protein, RcsA (e.g., GenBank Accession Number M58003 (GI: 1103316), incorporated herein by reference), in the colanic acid synthesis pathway results in an increase in intracellular GDP-fucose levels. Over-expression of an additional positive regulator of colanic acid biosynthesis, namely RcsB (e.g., GenBank Accession Number E04821 (GI:2173017), incorporated herein by reference), is also utilized, either instead of or in addition to over-expression of RcsA, to increase intracellular GDP- fucose levels.

Alternatively, colanic acid biosynthesis is increased following the introduction of a mutation into the E. coli Ion gene (e.g., GenBank Accession Number L20572 (GI:304907), incorporated herein by reference). Lon is an adenosine-5'-triphosphate (ATP)-dependant intracellular protease that is responsible for degrading RcsA, mentioned above as a positive transcriptional regulator of colanic acid biosynthesis in E. coli. In a lon null background, RcsA is stabilized, RcsA levels increase, the genes responsible for GDP-fucose synthesis in E coli are up-regulated, and intracellular GDP-fucose concentrations are enhanced.

Mutations in lon suitable for use with the methods presented herein include null mutations or insertions that disrupt the expression or function of lon.

A functional lactose permease gene is also present in the bacterium. The lactose permease gene is an endogenous lactose permease gene or an exogenous lactose permease gene. For example, the lactose permease gene comprises an E. coli lacY gene (e.g., GenBank Accession Number V00295 (GI:41897), incorporated herein by reference). Many bacteria possess the inherent ability to transport lactose from the growth medium into the cell, by utilizing a transport protein that is either a homolog of the E. coli lactose permease (e.g., as found in Bacillus licheniformis), or a transporter that is a member of the ubiquitous PTS sugar transport family (e.g., as found in Lactobacillus casei and Lactobacillus rhamnosus). For bacteria lacking an inherent ability to transport extracellular lactose into the cell cytoplasm, this ability is conferred by an exogenous lactose transporter gene (e.g., E. coli lacY) provided on recombinant DNA constructs, and supplied either on a plasmid expression vector or as exogenous genes integrated into the host chromosome.

As described herein, in some embodiments, the host bacterium preferably has a reduced level of β-galactosidase activity. In the embodiment in which the bacterium is characterized by the deletion of the endogenous β-galactosidase gene, an exogenous β- galactosidase gene is introduced to the bacterium. For example, a plasmid expressing an exogenous β-galactosidase gene is introduced to the bacterium, or recombined or integrated into the host genome. For example, the exogenous β-galactosidase gene is inserted into a gene that is inactivated in the host bacterium, such as the Ion gene.

The exogenous b-galactosidase gene is a functional b-galactosidase gene

characterized by a reduced or low leve of b-galactosidase activity compared to β- galactosidase activity in wild-type bacteria lacking any genetic manipulation. Exemplary β- galactosidase genes include E. coli lacZ and β-galactosidase genes from any of a number of other organisms (e.g., the lac4 gene of Kluyveromyces lactis (e.g., GenBank Accession Number M84410 (GI: 173304), incorporated herein by reference) that catalyzes the hydrolysis of b-galactosides into monosaccharides. The level of β-galactosidase activity in wild-type E. coli bacteria is, for example, 6,000 units. Thus, the reduced β-galactosidase activity level encompassed by engineered host bacterium of the present invention includes less than 6,000 units, less than 5,000 units, less than 4,000 units, less than 3,000 units, less than 2,000 units, less than 1,000 units, less than 900 units, less than 800 units, less than 700 units, less than 600 units, less than 500 units, less than 400 units, less than 300 units, less than 200 units, less than 100 units, or less than 50 units. Low, functional levels of β-galactosidase include β- galactosidase activity levels of between 0.05 and 1,000 units, e.g., between 0.05 and 750 units, between 0.05 and 500 units, between 0.05 and 400 units, between 0.05 and 300 units, between 0.05 and 200 units, between 0.05 and 100 units, between 0.05 and 50 units, between 0.05 and 10 units, between 0.05 and 5 units, between 0.05 and 4 units, between 0.05 and 3 units, or between 0.05 and 2 units of β-galactosidase activity. For unit definition and assays for determining β-galactosidase activity, see Miller JH, Laboratory CSH. Experiments in molecular genetics. Cold Spring Harbor Laboratory Cold Spring Harbor, NY; 1972;

(incorporated herein by reference). This low level of cytoplasmic β-galactosidase activity is not high enough to significantly diminish the intracellular lactose pool. The low level of β- galactosidase activity is very useful for the facile removal of undesired residual lactose at the end of fermentations.

Optionally, the bacterium has an inactivated thyA gene. Preferably, a mutation in a thyA gene in the host bacterium allows for the maintenance of plasmids that carry thyA as a selectable marker gene. Exemplary alternative selectable markers include antibiotic resistance genes such as BLA (beta-lactamase), or proBA genes (to complement a proAB host strain proline auxotropy) or pur A (to complement a pur A host strain adenine auxotrophy).

In one aspect, the E. coli bacterium comprises the genotyp - , and also

comprises any one of the exogenous α(1,2) fucosyltransferases described herein.

The bacterium comprising these characteristics is cultured in the presence of lactose. In some cases, the method further comprises culturing the bacterium in the presence of tryptophan and in the absence of thymidine. The fucosylated oligosaccharide is retrieved from the bacterium (i.e., a cell lysate) or from a culture supernatant of the bacterium.

The invention provides a purified fucosylated oligosaccharide produced by the methods described herein. The fucosylated oligosaccharide is purified for use in therapeutic or nutritional products, or the bacterium is used directly in such products. The fucosylated oligosaccharide produced by the engineered bacterium is 2'-fucosyllactose (2'-FL) or lactodifucotetraose (LDFT). The new alpha 1,2-fucosyltransferases are also useful to synthesize HMOS of larger molecular weight bearing alpha 1,2 fucose moieties, e.g., lacto- N-fucopentaose (LNF I) and lacto-N-difucohexaose (LDFH I). For example, to produce LDFT, the host bacterium is engineered to express an exogenous a (1,2) fucosyltransferase that also possesses a (1,3) fucosyltransferase activity, or an exogenous a (1,2)

fucosyltransferase and an exogenous a (1,3) fucosyltransferase. For the production of LNF I and LDFH I, the host bacterium is engineered to express an exogenous a (1,2)

fucosyltransferase that also possesses a (1,3) fucosyltransferase activity and/or a (1,4) fucosyltransferase activity, or an exogenous a (1,2) fucosyltransferase, an exogenous a (1,3) fucosyltransferas, and an exogenous a (1,4) fucosyltransferase.

A purified fucosylated oligosaccharide produced by the methods described above is also within the invention. The purified oligosaccharide (2'-FL) obtained at the end of the process is a white/slightly off-white, crystalline, sweet powder. For example, an engineered bacterium, bacterial culture supernatant, or bacterial cell lysate according to the invention comprises 2'-FL, LDFT, LNF I or LDFH I produced by the methods described herein, and does not substantially comprise a other fucosylated oligosaccharides prior to purification of the fucosylated oligosaccharide products from the cell, culture supernatant, or lysate. As a general matter, the fucosylated oligosaccharide produced by the methods contains a negligible amount of 3-FL in a 2'-FL-containing cell, cell lysate or culture, or supernatant, e.g., less than 1% of the level of 2'-FL or 0.5% of the level of 2'-FL. Moreover, the fucosylated oligosaccharide produced by the methods described herein also have a minimal amount of contaminating lactose, which can often be co-purified with the fucosylated oligosaccharide product, such as 2'FL. This reduction in contaminating lactose results from the reduced level of β-galactosidase activity present in the engineered host bacterium.

A purified oligosaccharide, e.g. _s 2'-FL, LDFT, LNF I, or LDFH I, is one that is at least 90%, 95%, 98%, 99%, or 100% (w/w) of the desired oligosaccharide by weight. Purity is assessed by any known method, e.g., thin layer chromatography or other chromatographic techniques known in the art. The invention includes a method of purifying a fucosylated oligosaccharide produced by the genetically engineered bacterium described above, which method comprises separating the desired fucosylated oligosaccharide (e.g., 2'-FL) from contaminants in a bacterial cell lysate or bacterial cell culture supernatant of the bacterium.

The oligosaccharides are purified and used in a number of products for consumption by humans as well as animals, such as companion animals (dogs, cats) as well as livestock (bovine, equine, ovine, caprine, or porcine animals, as well as poultry). For example, a pharmaceutical composition comprises purified 2'-FL and a pharmaceutically-acceptable excipient that is suitable for oral administration. Large quantities of 2'-FL are produced in bacterial hosts, e.g., anE. coli bacterium comprising an exogenous a (1,2) fucosyltransferase gene.

A method of producing a pharmaceutical composition comprising a purified human milk oligosaccharide (HMOS) is carried out by culturing the bacterium described above, purifying the HMOS produced by the bacterium, and combining the HMOS with an excipient or carrier to yield a dietary supplement for oral administration. These compositions are useful in methods of preventing or treating enteric and/or respiratory diseases in infants and adults. Accordingly, the compositions are administered to a subject suffering from or at risk of developing such a disease.

The invention also provides methods of identifying an a (1,2) fucosyltransferase gene capable of synthesizing fucosylated oligosaccharides in a host bacterium, i.e., 2'- fucosyllactose (2'-FL) in E. coli. The method of identifying novel lactose-utilizing, α(1,2)fucosyltransferase enzyme comprises the following steps:

1) performing a computational search of sequence databases to define a broad group of simple sequence homologs of any known, lactose-utilizing α(1,2)fucosyltransferase;

2) using the list from step (1), deriving a search profile containing common sequence and/or structural motifs shared by the members of the list; 3) searching sequence databases, using a derived search profile based on the common sequence or structural motif from step (2) as query, and identifying a candidate sequences, wherein a sequence homology to a reference lactose-utilizing α(l,2)fucosyltransferase is a predetermined percentage threshold;

4) compiling a list of candidate organisms, said organisms being characterized as expressing α(1,2)fucosyl- glycans in a naturally-occurring state;

5) selecting candidate sequences that are derived from candidate organisms to generate a list of candidate lactose-utilizing enzymes;

6) expressing the candidate lactose- utilizing enzyme in a host organism; and

7) testing for lactose- utilizing α(1,2)fucosyltransferase activity, wherein detection of the desired fucosylated oligosaccharide product in said organism indicates that the candidate sequence comprises a novel lactose- utilizing α(1,2)fucosyltransferase. In another embodiment, the search profile is generated from a multiple sequence alignment of the amino acid sequences of more than one enzyme with known α(1,2)fucosyltransferase activity. The database search can then be designed to refine and iteratively search for novel

α(1,2)fucosyltransferases with significant sequence similarlity to the multiple sequence alignment query.

The invention provides a method of treating, preventing, or reducing the risk of infection in a subject comprising administering to said subject a composition comprising a purified recombinant human milk oligosaccharide, wherein the HMOS binds to a palhogen and wherein the subject is infected with or at risk of infection with the pathogen. In one aspect, the infection is caused by a Norwalk-like virus or Campylobacter jejuni The subject is preferably a mammal in need of such treatment. The mammal is, e.g., any mammal, e.g., a human, a primate, a mouse, a rat, a dog, a cat, a cow, a horse, or a pig. In a preferred embodiment, the mammal is a human. For example, the compositions are formulated into animal feed (e.g., pellets, kibble, mash) or animal food supplements for companion animals, e.g., dogs or cats, as well as livestock or animals grown for food consumption, e.g., cattle, sheep, pigs, chickens, and goats. Preferably, the purified HMOS is formulated into a powder (e.g., infant formula powder or adult nutritional supplement powder, each of which is mixed with a liquid such as water or juice prior to consumption) or in the form of tablets, capsules or pastes or is incorporated as a component in dairy products such as milk, cream, cheese, yogurt or kefir, or as a component in any beverage, or combined in a preparation containing live microbial cultures intended to serve as probiotics, or in prebiotic preparations to enhance the growth of beneficial microorganisms either in vitro or in vivo. Polynucleotides, polypeptides, and oligosaccharides of the invention are purified and/or isolated. Purified defines a degree of sterility that is safe for administration to a human subject, e.g., lacking infectious or toxic agents. Specifically, as used herein, an "isolated" or "purified" nucleic acid molecule, polynucleotide, polypeptide, protein or oligosaccharide, is substantially free of other cellular material, or culture medium when produced by

recombinant techniques, or chemical precursors or other chemicals when chemically synthesized. For example, purified HMOS compositions are at least 60% by weight (dry weight) the compound of interest. Preferably, the preparation is at least 75%, more preferably at least 90%, and most preferably at least 99%, by weight the compound of interest. Purity is measured by any appropriate standard method, for example, by column chromatography, thin layer chromatography, or high-performance liquid chromatography (HPLC) analysis. For example, a "purified protein" refers to a protein that has been separated from other proteins, lipids, and nucleic acids with which it is naturally associated. Preferably, the protein constitutes at least 10, 20, 50, 70, 80, 90, 95, 99-100% by dry weight of the purified preparation.

Similarly, by "substantially pure" is meant an oligosaccharide that has been separated from the components that naturally accompany it. Typically, the oligosaccharide is substantially pure when it is at least 60%, 70%, 80%, 90%, 95%, or even 99%, by weight, free from the proteins and naturally-occurring organic molecules with which it is naturally associated.

By "isolated nucleic acid" is meant a nucleic acid that is free of the genes which, in the naturally-occurring genome of the organism from which the DNA of the invention is derived, flank the gene. The term covers, for example: (a) a DNA which is part of a naturally occurring genomic DNA molecule, but is not flanked by both of the nucleic acid sequences that flank that part of the molecule in the genome of the organism in which it naturally occurs; (b) a nucleic acid incorporated into a vector or into the genomic DNA of a prokaryote or eukaryote in a manner, such that the resulting molecule is not identical to any naturally occurring vector or genomic DNA; (c) a separate molecule such as a cDNA, a genomic fragment, a fragment produced by polymerase chain reaction (PCR), or a restriction fragment; and (d) a recombinant nucleotide sequence that is part of a hybrid gene, i.e., a gene encoding a fusion protein. Isolated nucleic acid molecules according to the present invention further include molecules produced synthetically, as well as any nucleic acids that have been altered chemically and/or that have modified backbones. A "heterologous promoter" is a promoter which is different from the promoter to which a gene or nucleic acid sequence is operably linked in nature.

The term "overexpress" or "overexpression" refers to a situation in which more factor is expressed by a genetically-altered cell than would be, under the same conditions, by a wild type cell. Similarly, if an unaltered cell does not express a factor that it is genetically altered to produce, the term "express" (as distinguished from "overexpress") is used indicating the wild type cell did not express the factor at all prior to genetic manipulation.

The terms "treating" and "treatment" as used herein refer to the administration of an agent or formulation to a clinically symptomatic individual afflicted with an adverse condition, disorder, or disease, so as to effect a reduction in severity and/or frequency of symptoms, eliminate the symptoms and/or their underlying cause, and/or facilitate improvement or remediation of damage. The terms "preventing" and "prevention" refer to the administration of an agent or composition to a clinically asymptomatic individual who is susceptible to a particular adverse condition, disorder, or disease, and thus relates to the prevention of the occurrence of symptoms and/or their underlying cause.

By the terms "effective amount" and "therapeutically effective amount" of a formulation or formulation component is meant a nontoxic but sufficient amount of the formulation or component to provide the desired effect.

The transitional term "comprising," which is synonymous with "including,"

"containing," or "characterized by," is inclusive or open-ended and does not exclude additional, unrecited elements or method steps. By contrast, the transitional phrase

"consisting of excludes any element, step, or ingredient not specified in the claim. The transitional phrase "consisting essentially of limits the scope of a claim to the specified materials or steps "and those that do not materially affect the basic and novel

characteristic(s)" of the claimed invention.

The host organism used to express the lactose-accepting fucosyltransferase gene is typically the enterobacterium Escherichia coli K12 (E, coli). E. coll K-12 is not considered a human or animal pathogen nor is it toxicogenic. E. coli K-l 2 is a standard production strain of bacteria and is noted for its safety due to its poor ability to colonize the colon and establish infections (see, e.g., epa.gov/oppt¾iotech/pubs/fra/fra004.htm). However, a variety of bacterial species may be used in the oligosaccharide biosynthesis methods, e.g., Erwinia herbicola {Pcmtoea agglomerate), Citrobacter freundii, Pantoea citrea, Pectobacterium carotovorum, or Xanthomonas campestris. Bacteria of the genus Bacillus may also be used, including Bacillus subtilis, Bacillus licheniformis, Bacillus coagulans, Bacillus thermophilus, Bacillus laterosporus, Bacillus megaterium, Bacillus mycoides, Bacillus pumilus, Bacillus lentus, Bacillus cereus, and Bacillus circulans. Similarly, bacteria of the genera

Lactobacillus and Lactococcus may be modified using the methods of this invention, including but not limited to Lactobacillus acidophilus, Lactobacillus salivarius,

Lactobacillus plantarum, Lactobacillus helveticus, Lactobacillus delbrueckH, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Lactobacillus crispatus, Lactobacillus gasseri, Lactobacillus casei, Lactobacillus reuteri, Lactobacillus jensenii, and Lactococcus lactis. Streptococcus themophiles and Proprionibacterium freudenreichii are also suitable bacterial species for the invention described herein. Also included as part of this invention are strains, modified as described here, from the genera Enterococcus (e.g., Enterococcus faecium and Enterococcus thermophiles), Bifidobacterium (e.g., Bifidobacterium longum, Bifidobacterium infantis, and Bifidobacterium bifidum), Sporolactobacillus spp., Micromomospora spp., Micrococcus spp., Rhodococcus spp., and Pseudomonas (e.g., Pseudomonas fluorescens and Pseudomonas aeruginosa). Bacteria comprising the characteristics described herein are cultured in the presence of lactose, and a fucosylated oligosaccharide is retrieved, either from the bacterium itself or from a culture supernatant of the bacterium. The fucosylated oligosaccharide is purified for use in therapeutic or nutritional products, or the bacteria are used directly in such products. A suitable production host bacterial strain is one that is not the same bacterial strain as the source bacterial strain from which the fucosyltransferase- encoding nucleic acid sequence was identified.

Other features and advantages of the invention will be apparent from the following description of the preferred embodiments thereof, and from the claims. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. All published foreign patents and patent applications cited herein are incorporated herein by reference. Genbank and NCBI submissions indicated by accession number cited herein are incorporated herein by reference. All other published references, documents, manuscripts and scientific literature cited herein are incorporated herein by reference. In the case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic illustration showing the synthetic pathway of the major neutral fucosyl-oligosaccharides found in human milk.

FIG. 2 is a schematic demonstrating metabolic pathways and the changes introduced into them to engineer 2'-fucosyllactose (2'-FL) synthesis in Escherichia coli ( E. coli).

Specifically, the lactose synthesis pathway and the GDP-fucose synthesis pathway are illustrated. In the GDP-fucose synthesis pathway: manA = phosphomannose isomerase (PMI), manB = phosphomannomutase (PMM), manC = mannose- 1 -phosphate

guanylyltransferase (GMP), gmd - GDP-mannose-4,6-dehydratase, fcl = GDP-fucose synthase (GFS), and bwcaJ = mutated UDP-glucose lipid carrier transferase.

FIG. 3A and FIG. 3B show the sequence identity and a multiple sequence alignment of 4 previously known lactose-utilizing α(1,2)-fucosyltransferase protein sequences. FIG. 3 A is a table showing the sequence identity between the 4 known lactose-utilizing α(1,2)- fucosyltransferases: H. pylori futC (SEQ ID NO: 1), H. mustelae FutL (SEQ ID NO: 2), Bacteroides vulgatus futN (SEQ ID NO: 3), and E. coli 0126 wbgL (SEQ ID NO: 4). FIG. 3B shows multiple sequence alignment of the 4 known α(1,2)-fucosyltransferases. The ovals highlight regions of particularly high sequence conservation between the four enzymes in the alignment.

FIG. 4 shows the sequence alignment of the 12 identified α(1,2)-fucosyltransferase syngenes identified, along with the 4 previously known lactose-utilizing α(1,2)- fucosyltransferase protein sequences. The 4 known lactose-utilizing α(1 ,2)- fucosyltransferases are boxed and include H. pylori futC (SEQ ID NO: 1), H. mustelae FutL (SEQ ID NO: 2), Bacteroides vulgatus futN (SEQ ID NO: 3), and E. coli 0126 wbgL (SEQ ID NO: 4). The 12 identified α(1,2)-fucosyltransferase are as follows: Prevotella

melaninogenica FutO (SEQ ID NO: 10), Clostridium bolteae +13 FutP (SEQ ID NO: 292), Lachnospiraceae sp. FutQ (SEQ ID NO: 12), Methanosphaerula palustris FutR (SEQ ID NO: 13), Tannerella sp. FutS (SEQ ID NO: 14), Bacteroides caccae FutU (SEQ ID NO: 15), Butyrivibrio FutV (SEQ ID NO: 16), Prevotella sp. FutW (SEQ ID NO: 17), Parabacteroides johnsonii FutX (SEQ ID NO: 18), Akkermansia muciniphilia FutY (SEQ ID NO: 19), Salmonella enterica FutZ (SEQ ID NO: 20), Bacteroides sp. FutZA (SEQ ID NO: 21). The sequence for Clostridium bolteae FutP (without the 13 additional amino acids in the N- terminus) (SEQ ID NO: 11) is also shown in the alignment. FIG. 5A and FIG. 5B are two pictures of gels showing the construction of the syngenes for each of the 12 novel α(1,2)-rucosyltransferases. FIG. 5 A shows post-Gibson assembly PCR. FIG. 5B shows gel-purified Rl/Xhol syngene fragments.

FIG. 6A and FIG. 6B are two photographs showing thin layer chromatograms of fucosylated oligosaccharide products produced in E. coli cultures using the 12 novel α(1,2)- fucosyltransferase syngenes. FIG. 6A shows fucosylated oligosaccharide products from 2μ1 of culture supernatant. FIG. 6B shows fucosylated oligosaccharide products from 0.2 OD ₆₀₀ cell equivalents of whole cell heat extracts.

FIG. 7 is a graph showing the growth curve of the host bacterium expressing plasmids containing the α(1,2) fucosyltransferase genes WbgL, FutN, FutO, FutQ, and FutX after tryptophan induction in the presence of lactose in the culture medium (i.e. lac + tip).

FIG. 8 is a photograph of a SDS-PAGE gel showing the proteins produced from host bacterium expressing α(1,2) fucosyltransferase genes WbgL, FutN, FutO, FutQ, and FutX after induction.

FIG. 9A and FIG. 9B are two photographs of thin layer chromatograms showing the production of fucosylated oligosaccharide products from in E coli cultures expressing select α(1,2)-fucosyltransferase syngenes WbgL, FutN, FutO, FutQ, and FutXst 7 hours or 24 hours after induction. FIG. 9A shows fucosylated oligosaccharide products from 2μ1 of culture supernatant. FIG. 9B shows fucosylated oligosaccharide products from 0.2 OD ₆₀₀ cell equivalents of whole cell heat extracts.

FIG. 10A and FIG. 1 OB are two photographs of thin layer chromatograms showing the fucosylated oligosaccharide products after two different 1.SL fermentation runs from E coli expressing FutN: FIG. 10A) 36B and FIG. 10B) 37A. The culture yield for run 36B was 33g/L while the yield for run 37A was 36.3 g/L.

FIG. 11 is a plasmid map of pG217 carrying the B. vulgatus FutN gene.

FIG. 12 is a schematic diagram showing the insertion of the Laclq promoter, the functional lacY gene, and the deletion of lacA.

FIG. 13 is a schematic diagram showing the deletion of the endogenous wcaJ gene using FRT recombination.

FIG. 14 is a schematic diagram of the E coli W3110 chromosome, showing the insertion of a DNA fragment carrying kanamycin resistance gene (derived from transposon Tn5) and wild-type lacZ into the Ion gene. DETAILED DESCRIPTION OF THE INVENTION

While some studies suggest that human milk glycans could be used as antimicrobial anti-adhesion agents, the difficulty and expense of producing adequate quantities of these agents of a quality suitable for human consumption has limited their full-scale testing and perceived utility. What has been needed is a suitable method for producing the appropriate glycans in sufficient quantities at reasonable cost. Prior to the invention described herein, there were attempts to use several distinct synthetic approaches for glycan synthesis. Some chemical approaches can synthesize oligosaccharides (Flowers, H. M. Methods Enzymol 50, 93-121 (1978); Seeberger, P. H. Chem Commun (Camb) 1115-1121 (2003)), but reactants for these methods are expensive and potentially toxic (Koeller, K. M. & Wong, C. H. Chem Rev 100, 4465-4494 (2000)). Enzymes expressed from engineered organisms (Albermann, C, Piepersberg, W. & Wehmeier, U. F. Carbohydr Res 334, 97-103 (2001); Bettler, E., Samain, E., Chazalet, V., Bosso, C, et al. Glycoconj J 16, 205-212 (1999); Johnson, K. F. Glycoconj J 16, 141-146 (1999); Palcic, M. M. Curr Opin Biotechnol 10, 616-624 (1999); Wymer.N. & Toone, E. J. Curr Opin Chem Biol 4, 110-119 (2000)) provide a precise and efficient synthesis (Palcic, M. M. Curr Opin Biotechnol 10, 616-624 (1999)); Crout, D. H. & Vic, G. Curr Opin Chem Biol 2, 98-111 (1998)), but the high cost of the reactants, especially the sugar nucleotides, limits their utility for low-cost, large-scale production. Microbes have been genetically engineered to express the glycosyltransferases needed to synthesize oligosaccharides from the bacteria's innate pool of nucleotide sugars (Endo, T., Koizumi, S., Tabata, K., Kakita, S. & Ozaki, A. Carbohydr Res 330, 439-443 (2001); Endo, T., Koizumi, S., Tabata, K. & Ozaki, A. Appl Microbiol Biotechnol 53, 257-261 (2000); Endo, T. & Koizumi, S. Curr Opin Struct Biol 10, 536-541 (2000); Endo, T., Koizumi, S., Tabata, K., Kakita, S. & Ozaki, A. Carbohydr Res 316, 179-183 (1999); Koizumi, S., Endo, T., Tabata, K. & Ozaki, A. Nat Biotechnol 16, 847-850 (1998)). However, prior to the invention described herein, there was a growing need to identify and characterize additional glycosyltransferases that are useful for the synthesis of HMOS in raetabolically engineered bacterial hosts.

Human Milk Glycans

Human milk contains a diverse and abundant set of neutral and acidic

oligosaccharides (Kunz, C, Rudloff, S., Baier, W., Klein, N., and Strobel, S. (2000). Annu Rev Nutr 20, 699-722; Bode, L. (2006). J Nutr 136, 2127-130). More than 130 different complex oligosaccharides have been identified in human milk, and their structural diversity and abundance is unique to humans. Although these molecules may not be utilized directly by infants for nutrition, they nevertheless serve critical roles in the establishment of a healthy gut microbiome (Marcobal, A., Barboza, M., Froehlich, J. W., Block, D. E., et al. J Agric Food Chem 58, 5334-5340 (2010)), in the prevention of disease (Newburg, D. S., Ruiz- Palacios, G. M. & Morrow, A. L. Annu Rev Nutr 25, 37-58 (2005)), and in immune function (Newburg, D. S. & Walker, W. A. Pediatr Res 61, 2-8 (2007)). Despite millions of years of exposure to human milk oligosaccharides (HMOS), pathogens have yet to develop ways to circumvent the ability of HMOS to prevent adhesion to target cells and to inhibit infection. The ability to utilize HMOS as pathogen adherence inhibitors promises to address the current crisis of burgeoning antibiotic resistance. Human milk oligosaccharides produced by biosynthesis represent the lead compounds of a novel class of therapeutics against some of the most intractable scourges of society.

One alternative strategy for efficient, industrial-scale synthesis of HMOS is the metabolic engineering of bacteria. This approach involves the construction of microbial strains overexpressing heterologous glycosyltransferases, membrane transporters for the import of precursor sugars into the bacterial cytosol, and possessing enhanced pools of regenerating nucleotide sugars for use as biosynthetic precursors (Dumon, C, Samain, E., and Priem, B. (2004). Biotechnol Prog 20, 412-19; Ruffing, A., and Chen, R.R. (2006). Microb Cell Fact 5, 25). A key aspect of this approach is the heterologous

glycosyltransferase selected for overexpression in the microbial host The choice of glycosyltransferase can significantly affect the final yield of the desired synthesized oligosaccharide, given that enzymes can vary greatly in terms of kinetics, substrate specificity, affinity for donor and acceptor molecules, stability and solubility. A few glycosyltransferases derived from different bacterial species have been identified and characterized in terms of their ability to catalyze the biosynthesis of HMOS in E. coli host strains (Dumon, C, Bosso, C, Utille, J.P., Heyraud, A., and Samain, E. (2006).

Chembiochem 7, 359-365; Dumon, C, Samain, E., and Priem, B. (2004). Biotechnol Prog 20, 412-19; Li, M., Liu, X.W., Shao, J., Shen, J., Jia, Q., Yi, W., Song, J.K., Woodward, R., Chow, C.S., and Wang, P.G. (2008). Biochemistry 47, 378-387). The identification of additional glycosyltransferases with faster kinetics, greater affinity for nucleotide sugar donors and/or acceptor molecules, or greater stability within the bacterial host significantly improves the yields of therapeutically useful HMOS. Prior to the invention described herein, chemical syntheses of HMOS were possible, but were limited by stereo-specificity issues, precursor availability, product impurities, and high overall cost (Flowers, H. M. Methods Enzymol 50, 93-121 (1978); Seeberger, P. H. Chem Commun (Camb) 1115-1121 (2003); Koeller, K. M. & Wong, C. H. Chem Rev 100, 4465-4494 (2000)). The invention overcomes the shortcomings of these previous attempts by providing new strategies to inexpensively manufacture large quantities of human milk oligosaccharides (HMOS) for use as dietary supplements. Advantages include efficient expression of the enzyme, improved stability and/or solubility of the fucosylated oligosaccharide product (2'-FL, LDFT, LNF I, and LDFH I) and reduced toxicity to the host organism. The present invention features novel α(1,2) FTs suitable for expression in production strains for increased efficacy and yield of fucosylated HMOS compared to α(1,2) FTs currently utilized in the field.

As described in detail below, E. coli (or other bacteria) is engineered to produce selected fucosylated oligosaccharides (Le., 2'-FL, LDFT, LDHF I, or LNF I) in commercially viable levels. For example, yields are >5 grams/liter in a bacterial fermentation process. In other embodiments, the yields are greater than 10 grams/liter, greater than 15 grams/liter, greater than 20 grams/liter, greater than 25 grams/liter, greater than 30 grams/liter, greater than 35 grams/liter, greater than 40 grams/liter, greater than 45 grams/liter, greater than 50 grams/liter, greater than 55 grams/liter, greater than 60 grams/liter, greater than 65 grams/liter, greater than 70 grams/liter, or greater than 75 grams/liter of fucosylated oligosaccharide products, such as 2'-FL, LDFT, LDHF I, and LNF I.

Role of Human milk plvcans in infectious disease

Human milk glycans, which comprise both unbound oligosaccharides and their glycoconjugates, play a significant role in the protection and development of the infant gastrointestinal (GI) tract. Neutral fucosylated oligosaccharides, including 2'-fucosyllactose (2'-FL), protect infants against several important pathogens. Milk oligosaccharides found in various mammals differ greatly, and the composition in humans is unique (Hamosh M., 2001 Pediatr Clin North Am, 48:69-86; Newburg D.S., 2001 Adv Exp Med Biol, 501:3-10).

Moreover, glycan levels in human milk change throughout lactation and also vary widely among individuals (Morrow A.L. et al., 2004 J Pediatr, 145:297-303; Chaturvedi P et al., 2001 Glycobiology, 11:365-372). Approximately 200 distinct human milk oligosaccharides have been identified and combinations of simple epitopes are responsible for this diversity (Newburg D.S., 1999 Curr Med Chem, 6:117-127; Ninonuevo M. et al., 2006 J Agric Food Chem, 54:7471-74801).

Human milk oligosaccharides are composed of 5 monosaccharides: D-glucose (Olc), D-galactose (Gal), N-acetylglucosamine (GlcNAc), L-fucose (Fuc), and sialic acid (N-acetyl neuraminic acid, Neu5Ac, NANA). Human milk oligosaccharides are usually divided into two groups according to their chemical structures: neutral compounds containing Glc, Gal, GlcNAc, and Fuc, linked to a lactose (Gaipi-4Glc) core, and acidic compounds including the same sugars, and often the same core structures, plus NANA (Charlwood J. et al., 1999 Anal Biochem, 273:261-277; Martin-Sosa et al., 2003 J Dairy Sci, 86:52-59; Parkkinen J. and Finne J., 1987 Methods Enzymol, 138:289-300; ShenZ. et al., 2001 J Chromatogr A, 921:315-321).

Approximately 70-80% of oligosaccharides in human milk are fucosylated, and their synthetic pathways are believed to proceed as shown in FIG. 1. A smaller proportion of the oligosaccharides are sialylated or both fucosylated and sialylated, but their synthetic pathways are not fully defined. Understanding of the acidic (sialylated) oligosaccharides is limited in part by the ability to measure these compounds. Sensitive and reproducible methods for the analysis of both neutral and acidic oligosaccharides have been designed. Human milk oligosaccharides as a class survive transit through the intestine of infants very efficiently, being essentially indigestible (Chaturvedi, P., Warren, C. D., Buescher, C. R., Pickering, L. K. & Newburg, D. S. Adv Exp Med Biol 501, 315-323 (2001)).

Human milk glvcans inhibit binding of enteropathogens to their receptors

Human milk glycans have structural homology to cell receptors for enteropathogens and function as receptor decoys. For example, pathogenic strains of Campylobacter bind specifically to glycans containing H-2, i.e., 2'-fucosyl--V-acetyUactosamine or 2'- fucosyllactose (2'FL); Campylobacter binding and infectivity are inhibited by 2'-FL and other glycans containing this H-2 epitope. Similarly, some diarrheagenic E. coli pathogens are strongly inhibited in vivo by human milk oligosaccharides containing 2-linked fucose moieties. Several major strains of human caliciviruses, especially the noroviruses, also bind to 2-linked fucosylated glycans, and this binding is inhibited by human milk 2-linked fucosylated glycans. Consumption of human milk that has high levels of these 2-linked fucosyloligosaccharides was associated with lower risk of norovirus, Campylobacter, ST of E. co/i-associated diarrhea, and moderate-to-severe diarrhea of all causes in a Mexican cohort of breastfeeding children (Newburg D.S. et al., 2004 Glycobiology, 14:253-263; Newburg D.S. et al., 1998 Lancet, 351:1160-1164). Several pathogens utilize sialylated glycans as their host receptors, such as influenza (Couceiro, J. N., Paulson, J. C. & Baum, L. G. Virus Res 29, 155-165 (1993)), parainfluenza (Amonsen, M., Smith, D. F., Cummings, R. D. & Air, G. M. J Virol 81, 8341-8345 (2007), and rotaviruses (Kuhlenschmidt, T. B., Hanafin, W. P., Gelberg, H. B. & Kuhlenschmidt, M. S. Adv Exp Med Biol 473, 309-317 (1999)). The sialyl-Lewis X epitope is used by Helicobacter pylori (Mahdavi, J., Sonden, B., Hurtig, M., Olfat, F. 0., et al. Science 297, 573-578 (2002)), Pseudomonas aeruginosa (Scharfman, A., Delmotte, P., Beau, J., Lamblin, G., et al. Glycoconj J 17, 735-740 (2000)), and some strains of noroviruses (Rydell, G. E., Nilsson, J., Rodriguez-Diaz, J., Ruvoen-Clouet, N., et al.

Glycobiology 19, 309-320 (2009)).

Identification of novel a(l.2) fucosvltransferases

The present invention provides novel α(1,2) fucosyltransferase enzymes. The present invention also provides nucleic acid constructs (i.e., aplasmid or vector) carrying the nucleic acid sequence of a novel α(1,2) fucosyltransferases for the expression of the novel α(1,2) fucosyltransferases in host bacterium. The present invention also provides methods for producing fucosylated oligosaccharides by expressing the novel α(1,2) fucosyltransferases in suitable host production bacterium, as further described herein.

Not all α(1 ,2)fucosyltrarisferases can utilize lactose as an acceptor substrate. An acceptor substrate includes, for example, a carbohydrate, an oligosaccharide, a protein or glycoprotein, a lipid or glycolipid, e.g., N-acetylglucosamine, N-acetyllactosamine, galactose, fucose, sialic acid, glucose, lactose, or any combination thereof. A preferred alpha (1,2) fucosyltransferase of the present invention utilizes GDP-fucose as a donor, and lactose is the acceptor for that donor.

A method of identifying novel α(1 ,2)fucosyltransferase enzymes capable of utilizing lactose as an acceptor was previously carried out (as described in PCT/US2013/051777, hereby incorporated by reference in its entirety) using the following steps: 1) perfonriing a computational search of sequence databases to define a broad group of simple sequence homologs of any known, lactose-utilizing α(1,2)fucosyltransferase; 2) using the list of homologs from step 1 to derive a search profile containing common sequence and/or structural motifs shared by the members of the broad group, e.g. by using computer programs such as MEME (Multiple Em for Motif Elicitation available at

http://meme.sdsc.edu/meme/cgi-bin/meme.cgi) or PSI-BLAST (Position-Specific Iterated BLAST available at ncbi.nlm.nih.gov/blast with additional information at

cnx.org/content/ml 1040/1 atest/); 3) searching sequence databases (e.g., using computer programs such as PSI-BLAST, or MAST (Motif Alignment Search Tool available at http://meme.sdsc.edu/meme/cgi-bin/mast.cgi);, using this derived search profile as query, and identifying "candidate sequences" whose simple sequence homology to the original lactose- accepting α(1,2)fucosyltransferase is 40% or less; 4) scanning the scientific literature and developing a list of "candidate organisms" known to express α(1,2)fucosyl-glycans; 5) selecting only those "candidate sequences" that are derived from "candidate organisms" to generate a list of "candidate lactose-utilizing enzymes"; and 6) expressing each "candidate lactose-utilizing enzyme" and testing for lactose-utilizing α(1,2)fucosyltransferase activity.

The MEME suite of sequence analysis tools (meme-sdsceduymeme/cgi-bin/memccgO can also be used as an alternative to PSI-BLAST. Sequence motifs are discovered using the program "MEME". These motifs can then be used to search sequence databases using the program "MAST". The BLAST and PSI-BLAST search algorithms are other well known alternatives.

To identify additional novel α(1,2)fucosyltransferases, a multiple sequence alignment query was generated using four previously identified lactose-utilizing

α(1,2)fucosyltransferase protein sequences: H. pylori futC (SEQ ID NO: 1), H. mustelae FutL (SEQ ID NO: 2), Bacteroides vulgatus futN (SEQ ID NO: 3), and E. coli 0126 wbgL (SEQ ID NO: 4). These sequence alignment and percentage of sequence identity is shown in FIG. 3. An iterative PSI-BLAST was performed, using the FASTA-formatted multiple sequence alignment as the query, and the NCBI PSI-BLAST program run on a local copy of NCBI BLAST+ version 2.2.29. An initial position-specific scoring matrix file (.pssm) was generated by PSI-BLAST, which the program then used to adjust the score of iterative homology search runs. The process is iterated to generate an even larger group of candidates, and the results of each run were used to further refine the matrix.

This PSI-BLAST search resulted in an initial 2515 hits. There were 787 hits with greater than 22% sequence identity to FutC. 396 hits were of greater than 275 amino acids in length. Additional analysis of the hits was performed, including sorting by percentage identity to FutC, comparing the sequences by BLAST to existing α(1,2) fucosyltransferase inventory (of known α(1,2) fucosyltransferases), and manual annotation of hit sequences to identify those originating from bacteria that naturally exist in the gastrointestinal tract. An annotated list of the novel α(1,2) fucosyltransferases identified by this screen are listed in Table 1. Table 1 provides the bacterial species from which the candidate enzyme is found, the GenBank Accession Number, GI Identification Number, amino acid sequence, and % sequence identity to FutC.

Of the identified hits, 12 novel α(1,2) fucosyltransferases were further analyzed for their functional capacity: Prevotella melaninogenica FutO, Clostridium bolteae FutP, Clostridium bolteae +13 FutP, Lachnospiraceae sp. FutQ, Methanosphaerula pediatries FutR, Tannerella sp. FutS, Bacteroides caccae FutU, Butyrivibrio FutV, Prevotellaa sp. FutW, Parabacteroides johnsonii FutX, Akkermansia muciniphilia FutY, Salmonella enterica FutZ, and Bacteroides sp. FutZA. For Clostridium bolteae FutP, the annotation named the wrong initiation methionine codon. Thus, the present invention includes FutP with an additional 13 amino acids at the N-terminus of the annotated FutP (derived in-frame from the natural upstream DNA sequence), which is designated herein as Clostridium bolteae +13 FutP. The sequence identity between the 12 novel α(l,2) fucosyltransferases identified and the 4 previously identified α(1,2) fucosyltransferases is shown in Table 2 below.

Table 2. Sequence Identity

Based on the amino acid sequences of the identified α(1,2) fucosyltransferases (i.e., in Table 1), syngenes can be readily designed and constructed by the skilled artisan using standard methods known in the art. For example, the syngenes include a ribosomal binding site, are codon-optimized for expression in a host bacterial production strain (i.e., E. coll), and have common 6-cutter restriction sites or sites recognized by endogenous restriction enzymes present in the host strain (i.e., EcoK restriction sites) removed to ease cloning and expression in the E. coli host strain. In a preferred embodiment, the syngenes are constructed with the following configuration: EcoRI site - T7gl0 RBS - α(1 ,2) FT syngene - Xhol site. The nucleic acid sequences of sample syngenes for the 12 identified α(1,2)

fucosyltransferases are shown in Table 3. (the initiating methionine ATG codon is bolded) Table 3. Nucleic acid sequences of 12 novel α(1,2) fucosyltransferase syngenes

CT/US2015/030823

In any of the methods described herein, the α(1,2) fiicosyltransferase genes or gene products may be variants or functional fragments thereof. A variant of any of genes or gene products disclosed herein may have 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% sequence identity to the nucleic acid or amino acid sequences described herein.

Variants as disclosed herein also include homolog, orthologs, or paralogs of the genes or gene products described herein that retain the same biological function as the genes or gene products specified herein. These variants can be used interchangeably with the genes recited in these methods. Such variants may demonstrate a percentage of homology or US 015/030823 identity, for example, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% identity conserved domains important for biological function, preferably in a functional domain, e.g. catalytic domain.

The term "% identity," in the context of two or more nucleic acid or polypeptide sequences, refer to two or more sequences or subsequences that are the same or have a specified percentage of amino acid residues or nucleotides that are the same, when compared and aligned for maximum correspondence, as measured using one of the following sequence comparison algorithms or by visual inspection. For example, % identity is relative to the entire length of the coding regions of the sequences being compared, or the length of a particular fragment or functional domain thereof.

For sequence comparison, typically one sequence acts as a reference sequence, to which test sequences are compared. When using a sequence comparison algorithm, test and reference sequences are input into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. The sequence

comparison algorithm then calculates the percent sequence identity for the test sequence(s) relative to the reference sequence, based on the designated program parameters.

Percent identity is determined using search algorithms such as BLAST and PSI- BLAST (Altschul et al., 1990, J Mol Biol 215:3, 403-410; Altschul et al., 1997, Nucleic

Acids Res 25:17, 3389-402). For the PSI-BLAST search, the following exemplary parameters are employed: (1 ) Expect threshold was 10; (2) Gap cost was Existence: 11 and Extension: 1 ; (3) The Matrix employed was BLOSUM62; (4) The filter for low complexity regions was "on".

Changes can be introduced by mutation into the nucleic acid sequence or amino acid sequence of any of the genes or gene products described herein, leading to changes in the amino acid sequence of the encoded protein or enayme, without altering the functional ability of the protein or enzyme. For example, nucleotide substitutions leading to amino acid substitutions at "non-essential" amino acid residues can be made in the sequence of any of sequences expressly disclosed herein. A "non-essential" amino acid residue is a residue at a position in the sequence that can be altered from the wild-type sequence of the polypeptide without altering the biological activity, whereas an "essential" amino acid residue is a residue at a position that is required for biological activity. For example, amino acid residues that are conserved among members of a family of proteins are not likely to be amenable to mutation. Other amino acid residues, however, (e.g. , those that are poorly conserved among members of the protein family) may not be as essential for activity and thus are more likely to be /030823 amenable to alteration. Thus, another aspect of the invention pertains to nucleic acid

molecules encoding the proteins or enzymes disclosed herein that contain changes in amino acid residues relative to the amino acid sequences disclosed herein that are not essential for activity (i.e., fucosyltransferase activity).

An isolated nucleic acid molecule encoding a protein essentially retaining the

functional capability compared to any of the genes described herein can be created by

introducing one or more nucleotide substitutions, additions or deletions into the

corresponding nucleotide sequence, such that one or more amino acid substitutions, additions or deletions are introduced into the encoded protein.

Mutations can be introduced into a nucleic acid sequence by standard techniques such that the encoded amino acid sequence is altered, such as site-directed mutagenesis and PCR- mediated mutagenesis. Preferably, conservative amino acid substitutions are made at one or more predicted non-essential amino acid residues. A "conservative amino acid substitution" is one in which the amino acid residue is replaced with an amino acid residue having a similar side chain. Families of amino acid residues having similar side chains have been defined in the art. Certain amino acids have side chains with more than one classifiable characteristic. These families include amino acids with basic side chains (e.g. , lysine, arginine, histidine), acidic side chains (e.g. , aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, tryptophan, cysteine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline,

phenylalanine, methionine, tyrosine, tryptophan), beta-branched side chains (e.g., threonine, valine, isoleucine) and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine). Thus, a predicted nonessential amino acid residue in a given polypeptide is replaced with another amino acid residue from the same side chain family. Alternatively, in another embodiment, mutations can be introduced randomly along all or part of a given coding sequence, such as by saturation mutagenesis, and the resultant mutants can be

screened for given polypeptide biological activity to identify mutants that retain activity.

Conversely, the invention also provides for variants with mutations that enhance or increase the endogenous biological activity. Following mutagenesis of the nucleic acid sequence, the encoded protein can be expressed by any recombinant technology known in the art and the activity of the protein can be determined. An increase, decrease, or elimination of a given biological activity of the variants disclosed herein can be readily measured by the ordinary person skilled in the art, i.e., by measuring the capability for mediating oligosaccharide modification, synthesis, or degradation (via detection of the products). The present invention also provides for functional fragments of the genes or gene products described herein. A fragment, in the case of these sequences and all others provided herein, is defined as a part of the whole that is less than the whole. Moreover, a fragment ranges in size from a single nucleotide or amino acid within a polynucleotide or polypeptide sequence to one fewer nucleotide or amino acid than the entire polynucleotide or polypeptide sequence. Finally, a fragment is defined as any portion of a complete polynucleotide or polypeptide sequence that is intermediate between the extremes defined above.

For example, fragments of any of the proteins or enzymes disclosed herein or encoded by any of the genes disclosed herein can be 10 to 20 amino acids, 10 to 30 amino acids, 10 to 40 amino acids, 10 to 50 amino acids, 10 to 60 amino acids, 10 to 70 amino acids, 10 to 80 amino acids, 10 to 90 amino acids, 10 to 100 amino acids, 50 to 100 amino acids, 75 to 125 amino acids, 100 to 150 amino acids, 150 to 200 amino acids, 200 to 250 amino acids, 250 to 300 amino acids, 300 to 350 amino acids, 350 to 400 amino acids, 400 to 450 amino acids, or 450 to 500 amino acids. The fragments encompassed in the present invention comprise fragments that retain functional fragments. As such, the fragments preferably retain the catalytic domains that are required or are important for functional activity. Fragments can be determined or generated by using the sequence information herein, and the fragments can be tested for functional activity using standard methods known in the art. For example, the encoded protein can be expressed by any recombinant technology known in the art and the activity of the protein can be determined. The biological function of said fragment can be measured by measuring ability to synthesize or modify a substrate oligosaccharide, or conversely, to catabolize an oligosaccharide substrate.

Within the context of the invention, ''functionally equivalent", as used herein, refers to a gene or the resulting encoded protein variant or fragment thereof capable of exhibiting a substantially similar activity as the wild-type fucosyltransferase. Specifically, the fucosyltransferase activity refers to the ability to transfer a fucose sugar to an acceptor substrate via an alpha-(l,2)-linkage. As used herein, "substantially similar activity" refers to an activity level within 5%, 10%, 20%, 30%, 40%, or 50% of the wild-type

fucosyltransferase.

To test for lactose-utilizing fucosylatransferase activity, the production of fucosylated oligossacharides (i.e., 2'-FL) is evaluated in a host organism that expresses the candidate enzyme (or syngene) and which contains both cytoplasmic GDP-fucose and lactose pools. The production of fucosylated oligosaccharides indicates that the candidate enzyme-encoding sequence functions as a lactose-utilizing α(1,2)fucosyltransferase. Engineering of E. coli to produce human milk oligosaccharide 2'-FL

Described herein is a gene screening approach, which was used to validate the novel a (1,2) fucosyltransferases (a (1,2) FTs) for the synthesis of fucosyl-linked oligosaccharides in metabolically engineered E. coli. Of particular interest are a (1,2) FTs that are capable of the synthesis of the HMOS 2'-fucosyllactose (2'-FL). 2'-FL is the most abundant fucosylated oligosaccharide present in human milk, and this oligosaccharide provides protection to newborn infants against infectious diarrhea caused by bacterial pathogens such as Campylobacter jejuni (Ruiz-Palacios, G.M., et al. (2003). J Biol Chem 278, 14112-120; Morrow, A.L. et al. (2004). J Pediatr 145, 297-303; Newburg, D.S. et al. (2004).

Glycobiology 14, 253-263). Other a (1,2) FTs of interest are those capable of synthesis of HMOS lactodifucotetraose (LDFT), laco-N-fucopentaose I (LNFI), or lacto-N-difucohexaose I (LDFH I).

The synthetic pathway of fucosyl oligosaccharides of human milk is illustrated in FIG. 1. Structurally, 2'-FL consists of a fucose molecule a 1 ,2 linked to the galactose portion of lactose (Fucal-2Galβl-4Glc). An a (1,2) FT from H. pylori strain 26695 termed FutC has been utilized to catalyze the synthesis of 2'-FL in metabolically engineered E coli (Drouillard, S. et al. (2006). Angew Chem Int Ed Engl 45, 1778-780).

Candidate α(1,2) FTs (i.e., syngenes) were cloned by standard molecular biological techniques into an expression plasmid. This plasmid utilizes the strong leftwards promoter of bacteriophage λ (termed PL) to direct expression of the candidate genes (Sanger, F. et al. (1982). J Mol Biol 162, 729-773). The promoter is controllable, e.g., a trp-cl construct is stably integrated the into the E.coli host's genome (at the ampC locus), and control is implemented by adding tryptophan to the growth media. Gradual induction of protein expression is accomplished using a temperature sensitive cl repressor. Another similar control strategy (temperature independent expression system) has been described

(Mieschendahl et al., 1986, Bio/Technology 4:802-808). The plasmid also carries the E. coli rcsA gene to up-regulate GDP-fucose synthesis, a critical precursor for the synthesis of fucosyl-linked oligosaccharides. In addition, the plasmid carries a β-lactamase (bid) gene for maintaining the plasmid in host strains by ampicillin selection (for convenience in the laboratory) and a native thy A (thymidylate synthase) gene as an alternative means of selection in thyA ^' hosts. Alternative selectable markers include the proBA genes to complement proline auxotrophy (Stein et al., (1984), J Bacteriol 158:2, 696-700 (1984) or purA to complement adenine auxotrophy (S. A. Wolfe, J. M. Smith, J Biol Chem 263, 19147-53 (1988)). To act as plasmid selectable markers each of these genes are first inactivated in the host cell chromosome, then wild type copies of the genes are provided on the plasmid. Alternatively a drug resistance gene may be used on the plasmid, e.g. beta-lactamase (this gene is already on the expression plasmid described above, thereby permitting selection with ampicillin). Ampicillin selection is well known in the art and described in standard manuals such as Maniatis et al., (1982) Molecular cloning, a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring, NY.

The nucleic acid sequence of such an expression plasmid, pEC2-(T7)FutX-rcsA-thyA (pG401) is provided below. The underlined sequence represents the FutX syngene, which can be readily replaced with any of the novel α(1 ,2) FTs described herein using standard recombinant DNA techniques.

The expression constructs were transformed into a host strain useful for the production of 2'-FL. Biosynthesis of 2'-FL requires the generation of an enhanced cellular pool of both lactose and GDP-fucose (FIG. 2). The wild-type Eschericia coli K12 prototrophic strain W3110 was selected as the parent background to test the ability of the candidates to catalyze 2'-FL production (Bachmann, B J. (1972). Bacteriol Rev 36, 525-557). The particular W3110 derivative employed was one that previously had been modified by the introduction (at the ampC locus) of a tryptophan-inducib repressor cassette, generating an E.coli strain known as GI724 (LaVallie, E.R. et al. (2000). Methods Enzymol 326, 322-340). Other features of GI724 include laclq and lacPL8 promoter mutations. Ecoli strain GI724 affords economical production of recombinant proteins from the phage λ P _L promoter following induction with low levels of exogenous tryptophan (LaVallie, E.R. et al. (1993). Biotechnology (N Y) 11, 187-193; Mieschendahl, et al. (1986). Bio/Technology 4, 802-08). Additional genetic alterations were made to this strain to promote the biosynthesis of 2'-FL. This was achieved in strain GI724 through several manipulations of the chromosome using λ Red recombineering (Court, D.L. et al. (2002). Annu Rev Genet 36, 361-388) and generalized PI phage transduction.

First, the ability of the E coli host strain to accumulate intracellular lactose was engineered by simultaneous deletion of the endogenous β-galactosidase gene (lacZ) and the lactose operon repressor gene (lacl). During construction of this deletion, the laclq promoter was placed immediately upstream of the lactose permease gene, lacY. The modified strain maintains its ability to transport lactose from the culture medium (via LacY), but is deleted for the wild-type copy of the lacZ(β-galactosidase) gene responsible for lactose catabolism. Therefore, an intracellular lactose pool is created when the modified strain is cultured in the presence of exogenous lactose. A schematic of the hromosomal construct is

shown in FIG. 12.

Genomic DNA sequence of the ch romosomal construct is set forth below

Second, the ability of the host E. coli strain to synthesize colanic acid, an extracellular capsular polysaccharide, was eliminated by the deletion of the wcaJ gene, encoding the UDP- glucose lipid carrier transferase (Stevenson, G. et al. (1996). J Bacteriol 178, 4885-893). In a wcaJ null background GDP-fucose accumulates in the E. coli cytoplasm (Dumon, C. et al. (2001). Glycoconj J 18, 465-474). A schematic of the chromosomal deletion of wcaJ is shown in FIG. 13.

The sequence of the chromosomal region of E. coli bearing the dwcaJ::FRI mutation is set forth below (SEQ ID NO: 289):

Third, the magnitude of the cytoplasmic GDP-fucose pool was enhanced by the introduction of a null mutation into the Ion gene. Lon is an ATP-dependant intracellular protease that is responsible for degrading RcsA, which is a positive transcriptional regulator of colanic acid biosynthesis in E coli (Gottesman, S. & Stout, V. Mol Microbiol 5, 1599- 1606 (1991)). In a Ion null background, RcsA is stabilized, RcsA levels increase, the genes responsible for GDP-fucose synthesis in E. coli are up-regulated, and intracellular GDP- fucose concentrations are enhanced. The Ion gene was almost entirely deleted and replaced by an inserted functional, wild-type, but promoter-less Ecoli lacZ+ gen ). λ Red recombineering was used to perform the construction. A schematic of the kan, lacZ insertion into the Ion locus is shown in FIG. 14.

Genomic DNA se uence surroundin the lacZ+ insertion into the Ion re ion in the

C /US2015/030823

Fourth, a thyA (thymidylate synthase) mutation was introduced into the strain by PI transduction. In the absence of exogenous thymidine, thyA strains are unable to make DNA and die. The defect can be complemented in trans by supplying a wild-type thyA gene on a multicopy plasmid (Belfort, M, Maley, G.F., and Maley, F. (1983). Proc Natl Acad Sci U S A 80, 1858-861). This complementation was used here as a means of plasmid maintenance.

An additional modification that is useful for increasing the cytoplasmic pool of free lactose (and hence the final yield of 2'-FL) is the incorporation of a lacA mutation. LacA is a lactose acetyltransferase that is only active when high levels of lactose accumulate in the E.

coli cytoplasm. High intracellular osmolarity (e.g., caused by a high intracellular lactose pool) can inhibit bacterial growth, and E. coli has evolved a mechanism for protecting itself from high intra cellular osmlarity caused by lactose by "tagging" excess intracellular lactose with an acetyl group using LacA, and then actively expelling the acetyl-lactose from the cell (Danchin, A. Bioessays 31, 769-773 (2009)). Production of acetyl-lactose in E. coli

engineered to produce 2'-FL or other human milk oligosaccharides is therefore undesirable: it reduces overall yield. Moreover, acetyl-lactose is a side product that complicates

oligosaccharide purification schemes. The incorporation of a lacA mutation resolves these problems. Sub-optimal production of fucosylated oligosaccharides occurs in strains lacking either or both of the mutations in the colanic acid pathway and the Ion protease. Diversion of lactose into a side product (acetyl-lactose) occurs in strains that do not contain the lacA mutation. A schematic of the lacA deletion and corresponding genomic sequence is provided above (SEQ ID NO: 288).

The strain used to test the different α(1,2) FT candidates incorporates all the above genetic modifications and has the following genotype:

The E. coli strains harboring the different α(1,2) FT candidate expression plasmids were analyzed. Strains were grown in selective media (lacking thymidine) to early exponential phase. Lactose was then added to a final concentration of 0.5%, and tryptophan (200 μΜ) was added to induce expression of each candidate α(1,2) FT from the P _L promoter. At the end of the induction period (~24 h) equivalent OD 600 units of each strain and the culture supernatant was harvested. Lysates were prepared and analyzed for the presence of 2'-FL by thin layer chromatography (TLC).

A map of plasmid pG217 is shown in FIG. 11, which carries the B. vulgatus FutN. The sequence of plasmid pG217 is set forth below (SEQ ID NO: 291):

Fucosylated oligosaccharides produced by metabolically engineered E. coli cells are purified from culture broth post-fermentation. An exemplary procedure comprises five steps. (1) Clarification: Fermentation broth is harvested and cells removed by sedimentation in a preparative centrifuge at 6000 x g for 30 min. Each bioreactor run yields about 5-7 L of partially clarified supernatant. (2) Product capture on coarse carbon : A column packed with coarse carbon (Calgon 12x40 TR) of ~1000 ml volume (dimension 5 cm diameter x 60 cm length) is equilibrated with 1 column volume (CV) of water and loaded with clarified culture supernatant at a flow rate of 40 ml/min. This column has a total capacity of about 120 g of sugar. Following loading and sugar capture, the column is washed with 1.5 CV of water, then eluted with 2.5 CV of 50% ethanol or 25% isopropanol (lower concentrations of ethanol at this step (25-30%) may be sufficient for product elution.) This solvent elution step releases about 95% of the total bound sugars on the column and a small portion of the color bodies. In this first step capture of the maximal amount of sugar is the primary objective. Resolution of contaminants is not an objective. (3) Evaporation: A volume of 2.5 L of ethanol or isopropanol eluate from the capture column is rotary-evaporated at 56 C° and a sugar syrup in water is generated. Alternative methods that could be used for this step include

lyophilization or spray-drying. (4) Flash chromatography on fine carbon and ion exchange media: A column (GE Healthcare HiScale50/40, 5x40cm, max pressure 20 bar) connected to a Biotage Isolera One FLASH Chromatography System is packed with 750 ml of a Darco Activated Carbon G60 (100-mesh): Celite 535 (coarse) 1:1 mixture (both column packings were obtained from Sigma). The column is equilibrated with 5 CV of water and loaded with sugar from step 3 (10-50 g, depending on the ratio of 2'-FL to contaminating lactose), using either a celite loading cartridge or direct injection. The column is connected to an evaporative light scattering (ELSD) detector to detect peaks of eluting sugars during the chromatography. A four-step gradient of isopropanol, ethanol or methanol is run in order to separate 2'-FL from monosaccharides (if present), lactose and color bodies. Fractions corresponding to sugar peaks are collected automatically in 120-ml bottles, pooled and directed to step 5. In certain purification runs from longer-than-normal fermentations, passage of the 2'-FL-containing fraction through anion-exchange and cation exchange columns can remove excess protein/DNA/caramel body contaminants. Resins tested successfully for this purpose are Dowex 22.

The gene screening approach described herein was successfully utilized to identify new α(1,2) FTs for the efficient biosynthesis of 2'-FL in metabolically engineered E coli host strains. The results of the screen are summarized in Table 1.

Production Host Strains

E. coli K-12 is a well-studied bacterium which has been the subject of extensive research in microbial physiology and genetics and commercially exploited for a variety of industrial uses. The natural habitat of the parent species, E. coll, is the large bowel of mammals. E. coli K-12 has a history of safe use, and its derivatives are used in a large number of industrial applications, including the production of chemicals and drugs for human administration and consumption. E. coli K-12 was originally isolated from a convalescent diphtheria patient in 1922. Because it lacks virulence characteristics, grows readily on common laboratory media, and has been used extensively for microbial physiology and genetics research, it has become the standard bacteriological strain used in microbiological research, teaching, and production of products for industry and medicine. E. coli K-12 is now considered an enfeebled organism as a result of being maintained in the laboratory environment for over 70 years. As a result, K-12 strains are unable to colonize the intestines of humans and other animals under normal conditions. Additional information on this well known strain is available at http://epa.gov/oppt/biotech/pubs/lra/fra004.htm. In addition to E. coli K12, other bacterial strains are used as production host strains, e.g., a variety of bacterial species may be used in the oligosaccharide biosynthesis methods, e.g., Erwinia herbicola (Pantoea agglomerans), Citrobacter freundii, Pantoea citrea, Pectobacterium carotovorum, or Xanthomonas campestris. Bacteria of the genus Bacillus may also be used, including Bacillus subtilis, Bacillus licheniformis, Bacillus coagulans, Bacillus thermophilus, Bacillus laterosporus, Bacillus megaterium, Bacillus my co ides, Bacillus pumilus, Bacillus lentus, Bacillus cereus, and Bacillus circulans. Similarly, bacteria of the genera Lactobacillus and Lactococcus may be modified using the methods of this invention, including but not limited to Lactobacillus acidophilus, Lactobacillus salivarius, Lactobacillus plantarum,

Lactobacillus Helvetians, Lactobacillus delbrueckii, Lactobacillus rhamnosus, Lactobacillus bulgaricus, Lactobacillus crispatus, Lactobacillus gasseri, Lactobacillus casei, Lactobacillus reuteri, Lactobacillus j'ensenii, and Lactococcus lactis. Streptococcus thermophiles and Proprionibacterium freudenreichii are also suitable bacterial species for the invention described herein. Also included as part of this invention are strains, modified as described here, from the genera Enterococcus (e.g., Enterococcus faecium and Enterococcus thermophiles), Bifidobacterium (e.g., Bifidobacterium longum, Bifidobacterium infantis, and Bifidobacterium bifidum), Sporolactobacillus spp., Micromomospora spp., Micrococcus spp., Rhodococcus spp., and Pseudomonas (e.g., Pseudomonas fluorescens and Pseudomonas aeruginosa).

Suitable host strains are amenable to genetic manipulation, e.g., they maintain expression constructs, accumulate precursors of the desired end product, e.g., they maintain pools of lactose and GDP-fucose, and accumulate endproduct, e.g., 2'-FL. Such strains grow well on defined minimal media that contains simple salts and generally a single carbon source. The strains engineered as described above to produce the desired fucosylated oligosaccharide(s) are grown in a minimal media. An exemplary minimal medium used in a bioreactor, minimal "FERM" medium, is detailed below.

Ferm (10 liters): Minimal medium comprising:

Water to 10 liters

DF204 antifoam (O.lml/L)

150 g glycerol (initial batch growth), followed by fed batch mode with a 90% glycerol-1% MgSO ₄ -1X trace elements feed, at various rates for various times.

A suitable production host strain is one that is not the same bacterial strain as the source bacterial strain from which the fucosyltransferase-encoding nucleic acid sequence was identified.

Bacteria comprising the characteristics described herein are cultured in the presence of lactose, and a fucosylated oligosaccharide is retrieved, either from the bacterium itself or from a culture supernatant of the bacterium. The fucosylated oligosaccharide is purified for use in therapeutic or nutritional products, or the bacteria are used directly in such products. EXAMPLES

Example 1: Identification of novel α(1.2) fucosyltransferases

To identify additional novel α(1,2)fucosyltransferases, a multiple sequence alignment query was generated using the alignment dgorithm of the CLCbio Main Workbench package, version 6.9 (CLCbio, 10 Rogers Street #101, Cambridge, Massachusetts 02142, USA) using four previously identified lactose-utilizing α(1,2)fucosyltransferase protein sequences: H. pylori futC (SEQ ID NO: 1), H. mustelae FutL (SEQ ID NO: 2), Bacteroides vulgatus futN (SEQ ID NO: 3), and E. coli 0126 wbgL (SEQ ID NO: 4). This sequence alignment and percentages of sequence identity between the four previously identified lactose-utilizing α(1,2)fucosyltransferase protein sequences is shown in FIG. 3. An iterative PSI-BLAST was performed, using the FASTA-formatted multiple sequence alignment as the query, and the NCBI PSI-BLAST program run on a local copy of NCBI BLAST+ version 2.2.29. An initial position-specific scoring matrix file (.pssm) was generated by PSI-BLAST, which was then used to adjust the score of iterative homology search runs. The process is iterated to generate an even larger group of candidates, and the results of each run were used to further refine the . matrix.

A ortion of the initial osition-s ecific scorin matrix file used is shown below:

5/030823

The command line of PSI-BLAST that was used is as follows:

psibIast-db<LOCAL NR database name> -max_taiget_seqs 2500-in_msa<MSA file in FAST format> -out <results output file> -outfint "7sskingdoms sscinames scomnames sseqid stitle evalue length pident" -out_pssm<PSSM file output> -out_ascii_pssm<PSSM (ascii) output -num iterations 6 -num_hreads 8

This PSI-BLAST search resulted in an initial 2515 hits. There were 787 hits with greater than 22% sequence identity to FutC. 396 hits were of greater than 275 amino acids in length. Additional analysis of the hits was performed, including sorting by percentage identity to FutC, comparing the sequences by BLAST to an existing α(1,2) fucosyltransferase inventory (of known α(1,2) fucosyltransferases, to eliminate known lactose-utilizing enzymes and duplicate hits), and manual annotation of hits to identify those originating from bacteria that naturally exist in the gastrointestinal tract. An annotated list of the novel α(1 ,2)

fucosyltransferases identified by this screen are listed in Table 1. Table 1 provides the bacterial species from which the enzyme is found, the GenBank Accession Number, GI

Identification Number, amino acid sequence, and % sequence identity to FutC.

Multiple sequence alignment of the 4 known α(1,2) FTs used for the PSI-BLAST query and 12 newly identified α(1,2) FTs is shown in FIG. 4.

Example 2: Validation of novel α(1.2) FTs

To test for lactose-utilizing fucosylatransferase activity, the production of fucosylated oligossacharides (i.e., 2'-FL) is evaluated in a host organism that expresses the candidate enzyme (i.e., syngene) and which contains both cytoplasmic GDP-fucose and lactose pools.

The production of fucosylated oligosaccharides indicates that the candidate enzyme-encoding sequence functions as a lactose-utilizing α(1,2)fucosyltransferase. Of the identified hits, 12 novel α(1,2) fucosyltransferases were further analyzed for their functional capacity to

produce 2'-fucosyllactose: Prevotella melaninogenica FutO, Clostridium bolteae FutP,

Clostridium bolteae +13 FutP, Lachnospiraceae sp. FutQ, Methanosphaerulapalustries

FutR, Tannerella sp. FutS, Bacteroides caccae FutU, Butyrivibrio FutV, Prevotellaa sp.

FutW, Parabacteroides johnsonii FutX, Akkermansia muciniphilia FutY, Salmonella enterica FutZ, and Bacteroides sp. FutZA.

Syngenes were constructed comprising the 12 novel α(1,2) FTs in the configuration as follows: EcoRI - T7gl0 RBS - syngene - Xhol. FIG. 5 A and FIG. 5B show the syngene fragments after PCR assembly and gel-purification. The candidate <x(l,2) FTs (i,e., syngenes) were cloned by standard molecular biological techniques into an exemplary expression plasmid pEC2-(T7)-Fut syngene-rcsA- thyA. This plasmid utilizes the strong leftwards promoter of bacteriophage λ (termed P _L ) to direct expression of the candidate genes (Sanger, F. et al. (1982). J Mol Biol 162, 729-773). The promoter is controllable, e.g., a trp-cl construct is stably integrated the into the E.coli host's genome (at the ampC locus), and control is implemented by adding tryptophan to the growth media. Gradual induction of protein expression is accomplished using a temperature sensitive cl repressor. Another similar control strategy (temperature independent expression system) has been described (Mieschendahl et al., 1986, Bio/Technology 4:802-808). The plasmid also carries the E. coli rcsA gene to up-regulate GDP-fucose synthesis, a critical precursor for the synthesis of fucosyl-linked oligosaccharides. In addition, the plasmid carries a β-lactamase (bla) gene for maintaining the plasmid in host strains by ampicillin selection (for convenience in the laboratory) and a native thyA (thymidylate synthase) gene as an alternative means of selection in thyA ^' hosts.

The expression constructs were transformed into a host strain useful for the production of 2'-FL. The host strain used to test the different α(1,2) FT candidates incorporates all the above genetic modifications described above and has the following genotype:

The E. coli strains harboring the different α(1,2) FT candidate expression plasmids were analyzed. Strains were grown in selective media (lacking mymidine) to early exponential phase. Lactose was then added to a final concentration of 0.5%, and tryptophan

(200 μΜ) was added to induce expression of each candidate α(1,2) FT from the P _L promoter.

At the end of the induction period (-24 h) the culture supernatants and cells were harvested.

Heat extracts were prepared from whole cells and the equivalent of 0.20D ₆₀₀ units of each strain analyzed for the presence of 2 ^' -FL by thin layer chromatography (TLC), along with 2μ1 of the corresponding clarified culture supernatant for each strain.

FIG. 6 shows the oligosaccharides produced by the α(1,2) FT-expressing bacteria, as determined by TLC analysis of the culture supernatant and extracts from the bacterial cells.

2'FL was produced by exogenous expression of WbgL (used as control), FutO, FutP, FutQ,

FutR, FutS, FutU, FutW, FutX, FutZ, and FutZA. /US2015/030823

Table 4 summarizes the iucosyltransferase activity for each candidate syngene as determined by the 2'FL synthesis screen described above. 11 of the 12 candidate α(1 ,2) FTs were found to have lactose-utilizing fucosyltransferase activity.

Table 4. 2'FL synthesis screen results

Further characterization of the bacterium expressing novel α(1,2) FTs FutO, FutQ, and FutX was performed. Specifically, proliferation rate and exogenous α(1,2) FT

expression was examined.

Expression plasmids containing fucosyltransferases WbgL (plasmid pG204), FutN (plasmid pG217), and novel α(1,2) FTs FutO (plasmid pG393), FutQ (plasmid pG395), and FutX (pG401) were introduced into host bacterial strains. For example, the host strains utilized has the following genotype:

Bacterial cultures expressing each exogenous fucosyltransferase were induced by addition of tryptophan (to induce expression of the exogenous fucosyltransferases) in the presence of lactose. Growth of the cultures was monitored by spectrophotometric readings at A600 at the following timepoints: 4 hours and 1 hour before induction, at the time of

induction (time 0), and 3 hours, 7 hours, and 24 hours after induction. The results are shown in FIG, 7, and indicate that expression of the exogenous fucosyltransferase did not prevent cell proliferation. Furthermore, the growth curve for the bacterial cultures expressing the novel α(1,2) fucosyltransferases FutO, FutQ, and FutX is similar to those expressing the known α(1,2)FT enzymes WbgL and FutN.

Protein expression was also assessed for the bacterial cultures expressing each

fucosyltransferase after induction. Cultures were induced as described previously, and protein lysates were prepared from the bacterial cultures at the time of induction (0 hours), 3 hours, 7 hours, and 24 hours after induction. The protein lysates were run on an SDS-PAGE gel and stained to examine the distribution of proteins at each time point. As shown in FIG. 8, induction at 7hours and 24 hours showed increases in a protein band at around 20-28 kDa for bacterial cultures expressing exogenous FutN, FutO, and FutX. These results indicate that induction results in significant expression of the exogenous fucosyltransferases.

Finally, additional TLC analysis to assess the efficiency and yield of 2'FL production in bacterial cultures expressing novel α(1,2) FTs FutO, FutQ, and FutX compared to known fucosyltransferases WbgL and FutN. Cultures were induced at 7 hours and 24 hours, and run out on TLC. FIG. 9A shows the level of 2'FL in the cell supernatant. The level of 2'FL found in the bacterial cells were also examined. As shown in FIG. 9B, 2'FL was produced in cell lysates from bacteria expressing the novel α(1 ,2) FTs FutO, FutQ, and FutX at 7 hours and 24 hours after induction.

Example 4: FutN exhibits increased efficiency for production of 2'FL

Fucosylated oligosaccharides produced by metabolically engineered E. coli cells to express B. vulgatus FutN was purified from culture broth post-fermentation.

Fermentation broth was harvested and cells were removed by sedimentation in a preparative centrifuge at 6000 x g for 30 min. Each bioreactor run yields about 5-7 L of partially clarified supernatant. A column packed with coarse carbon (Calgon 12x40 TR) of -1000 ml volume (dimension 5 cm diameter x 60 cm length) was equilibrated with 1 column volume (CV) of water and loaded with clarified culture supernatant at a flow rate of 40 ml/min. This column had a total capacity of about 120 g of sugar. Following loading and sugar capture, the column is washed with 1.5 CV of water, then was eluted with 2.5 CV of 50% ethanol or 25% isopropanol (lower concentrations of ethanol at this step (25-30%) may be sufficient for product elution.) This solvent elution step released about 95% of the total bound sugars on the column and a small portion of color bodies (caramelized sugars). A volume of 2.5 L of ethanol or isopropanol eluate from the capture column was rotary- evaporated at 56 C° and a sugar syrup in water was generated. A column (GE Healthcare HiScale50/40, 5x40cm, max pressure 20 bar) connected to a Biotage Isolera One FLASH Chromatography System was packed with 750 ml of a Darco Activated Carbon G60 (100- mesh): Celite 535 (coarse) 1:1 mixture (both column packings were obtained from Sigma). The column was equilibrated with 5 CV of water and loaded with sugar from step 3 (10-50 g, depending on the ratio of 2'-FL to contaminating lactose), using either a celite loading cartridge or direct injection. The column was connected to an evaporative light scattering (ELSD) detector to detect peaks of eluting sugars during the chromatography. A four-step gradient of isopropanol, ethanol or methanol was run in order to separate 2'-FL from monosaccharides (if present), lactose and color bodies. Fractions corresponding to sugar peaks were collected automatically in 120-ml bottles, pooled.

The results from two fermentation runs are shown in FIG. 10A and FIG. 10B. The cultures were grown for 136 (run 36B) or 112 hours (run 37A), and the levels of 2'-FL produced was analyzed by TLC analysis. As shown in both FIG. 1 OA and FIG. 10B, the 2'- fucosyllactose was produced at 40 hours of culture, and production continued to increase until the end point of the fermentation process. The yield of 2'-FL produced from run 36B was 33 grams per liter. The yield of 2'-FL produced from run 37A was 36.3 grams per liter. These results indicate that expression of exogenous FutN is suitable for high yield of 2'- fucosyllactose product.

OTHER EMBODIMENTS

While the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

The patent and scientific literature referred to herein establishes the knowledge that is available to those with skill in the art. All United States patents and published or unpublished United States patent applications cited herein are incorporated by reference. All published foreign patents and patent applications cited herein are hereby incorporated by reference. Genbank and NCBI submissions indicated by accession number cited herein are hereby incorporated by reference. All other published references, documents, manuscripts and scientific literature cited herein are hereby incorporated by reference.

While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.