FIELD OF THE INVENTION

The present invention relates to mutated and/or transformed microorganisms for the synthesis of various compounds. More specifically, the present invention discloses microorganisms mutated in the genes encoding for the regulators ArcA and IcIR. The latter mutations result in a significant upregulation of the genes that are part of the colanic acid operon. Hence, said microorganisms are useful for the synthesis of any compound being part of the colanic acid pathway such as GDP-fucose, GDP-mannose and colanic acid, and/or, can be further used - starting from GDP-fucose as a precursor - to synthesize fucosylated oligosaccharides or - starting from GDP-mannose as a precursor - to synthesize mannosylated oligosaccharides. In addition, mutations in the genes coding for the transcriptional regulators ArcA and IcIR lead to an acid resistance phenotype in the exponential growth phase allowing the synthesis of pH sensitive molecules and organic acids.

BACKGROUND OF THE INVENTION

The genes arcA encoding for the aerobic respiration control protein and icIR encoding the isocitrate lyase regulator are known to regulate the central carbon metabolism. ArcA is a global transcriptional regulator that regulates a wide variety of genes, while IcIR is a local transcriptional regulator that regulates the glyoxylate pathway. ArcA is known to regulate the central carbon metabolism in response to oxygen deprivation and has no connection with IcIR other than that it also regulates the glyoxylate pathway (24, 28, 29, 37, 38). In an earlier study the combined effect of AiclRAarcA mutant strains on the central carbon metabolism has been observed. Increased fluxes were shown in the tricarboxylic acid (TCA) cycle and glyoxylate pathway and an interesting and surprising phenotype appeared when both genes where knocked out, namely the double mutant strain formed biomass with a yield that approached the maximal theoretical yield (4, 39).

Some compounds, such as GDP-fucose, are in high demand. The latter compound is indeed a precursor of fucosylated oligosaccharides such as fucosyllactose, fucosyllactoNbiose and lewis X oligosaccharides, or, of fucosylated proteins. These sugars are components of human mother milk in which they have anti-inflammatory and prebiotic effects and/or have applications in therapeutics as nutraceutical, anti- inflammatory agent or prebiotic, in addition, fucosylated proteins find applications in the pharmaceutics (5, 8, 27). However, an efficient method to produce the latter high-value compounds is still needed.

In addition GDP-mannose is also an intermediate of the pathway towards GDP-fucose. Interrupting the pathway prematurely leads to the accumulation of this compound, which is a precursor of mannosylated oligosacharides. These oligosaccharides find for example applications in the treatment of gram-negative bacterial infections, in addition, GDP- mannose is important for the humanization of protein glycosylations, which is essential for the production of certain therapeutic proteins (18, 30). Mannosylated oligosaccharides and mannosylated glycoconjugates are also used for drug targeting, for instance mannosylated antivirals can specifically target the liver and kidneys (7).

The present invention provides microorganisms which are genetically changed in such a manner that they can efficiently produce the latter compounds.

Moreover, the synthesis of pH sensitive molecules, such as - but not limited to - glucosamine, and organic acids, such as - but not limited to - pyruvic acid, succinic acid, adipic, sialic acid, sialylated oligosaccharides... are preferably produced at low pH, either to stabilize the product or for downstream processing reasons (4, 12, 40). Therefore, strains that can grow at low pH are beneficial for these production processes. E. coli is an organism that can adapt easily to various conditions, for instance it can easily adapt to the harsh pH conditions in the stomach, which is about pH 2 (14). Nonetheless, E. coli does not seem to grow at these conditions, but induces its acid resistance mechanisms in the stationary phase (40). During this phase the cell does not multiply anymore and therefore hampers productivity. Up to now, no solution was found to this problem. However, in the present invention, a genetically engineered microorganism is provided that can induce acid resistance mechanisms in the exponential growth phase, which is the phase that is mostly used for production of organic acids and pH instable products.

BRIEF DESCRIPTION OF FIGURES

Figure 1 : Relative gene expression pattern of the wild type, the AiclR and AarcA mutant strain to the AarcAAiclR mutant strain of genes involved in colanic acid biosynthesis in batch fermentation conditions. The genes involved in colanic acid biosynthesis are presented in Figures 3 and 4. Figure 2: Gene expression pattern of the colanic acid operon of the wild type, the AiclR and AarcA mutant strain in chemostat fermentation conditions relative to the AarcAAiclR mutant strain.

Figure 3: The gene organisation of the colanic acid operon and an overview of the function of these genes:

Gene: Function: wza Component of capsular polysaccharide export

apparatus

wzb Tyrosine phosphatase wzc Tyrosine kinase wcaA Glycosyltransferase wcaB Acyltransferase wcaC Glycosyltransferase wcaD Colanic acid polymerase wcaE Glycosyltransferase wcaF Acyltransferase gmd GDP-mannose-4,6-dehydratase fcl GDP-fucose synthase gmm GDP-mannose hydrolase weal Glycosyltransferase cpsB Mannose-1 -phosphate guanylyltransferase cpsG Phosphomannomutase wcaJ UDP-glucose lipid carrier transferase wzxC Putative transporter wcaK Pyruvyltransferase wcaL Glycosyltransferase wcaM Predicted protein in colanic acid biosynthesis

Figure 4: The colanic acid biosynthesis pathway.

Figure 5: Regulatory network of the colanic acid operon. This network was constructed with Pathway tools v 13.0.

Figure 6: Effect of the AarcAAiclR mutations on the GDP-fucose biosynthesis route.

Figure 7: Overview of the genetic modifications needed to enhance fucosyllactose and fucosylated oligosaccharides production starting from glucose as a substrate.

Figure 8: Starting from sucrose, fucosylated sugar derivates such as fucosyllactose and more specifically 1 ,2-fucosyllactose are produced. The strain is modified to force the cell to produce frucose-6-phosphate which is an intermediate in the synthesis of GDP-fucose. Glucose or glucose-1 -phosphate (if the starting enzyme is either a sucrase or a sucrose phosphorylase) is then fed to the central carbon metabolism via the pentose phosphate pathway.

Figure 9: Overview of the genetic modifications needed to enhance fucosyllactose and fucosylated oligosaccharides production starting from glucose as a substrate in a split metabolism.

Figure 10: Detail of the pentose phosphate pathway flux in a strain in which the genes coding for phosphoglucose isomerase and phosphofructokinase are knocked out.

Figure 11 : Starting from sucrose, mannosylated sugar derivates are produced. The strain is modified to force the cell to produce frucose-6-phosphate which is an intermediate in the synthesis of GDP-fucose. Glucose or glucose-1 -phosphate (if the starting enzyme is either a sucrase or a sucrose phosphorylase) is then fed to the central carbon metabolism via the pentose phosphate pathway.

Figure 12: Gene expression pattern acid resistance related genes of the wild type, the AiclR and AarcA mutant strain in batch culturing conditions relative to the AarcAAiclR mutant strain.

Figure 13: LC MSMS analysis chromatograms of culture broth and a 2-fucosyllactose standard. A. LC MSMS analysis of the standard, B. LC MSMS analysis of a sample of the culture broth of a mutant strain expressing a H. pylori fucosyltransferase, C. LC MSMS analysis of a sample of the culture broth of a mutant strain expressing a H. pylori fucosyltransferase.

Figure 14: LC MSMS analysis mass spectrum from the chromatograms shown in Figure 13 of culture broth and a 2-fucosyllactose standard. A. Mass (m/z) of the standard, B. Mass (m/z) of the sample of the culturing broth of a mutant strain expressing a H. pylori fucosyltransferase, C. Mass (m/z) of the sample of the culturing broth of a mutant strain expressing a H. pylori fucosyltransferase.

Figure 15: The sequence of the artificial hybrid promoter as given by SEQ ID N° 6 (the combination of the native and an artificial promoter) that was cloned in front of the colanic acid operon.

DESCRIPTION OF INVENTION

The present invention provides microorganisms such as Enterobacteriaceae which are genetically changed in such a manner that they can efficiently produce compounds which are part of the colanic acid pathway. A particular compound of interest is GDP-fucose which is used as a precursor to synthesize fucosylated oligosaccharides. The latter have health-promoting effects as indicated above but there is no efficient production method available to produce said compounds.

The present invention thus provides for the usage of a mutated and/or transformed microorganism comprising a genetic change leading to a modified expression and/or activity of the transcriptional regulators the aerobic respiration control protein ArcA and the isocitrate lyase regulator IcIR to upregulate at least one of the genes of the colanic acid operon, wherein said operon comprises the genes cpsG, cpsB, gmd and fcl that code for a phosphomannomutase, a mannose-1 -phosphate guanylyltransferase, GDP-mannose 4,6-dehydratase and GDP-fucose synthase, respectively. The latter operon may also comprise the genes cpsG, cpsB, gmd, fcl and wza. In addition the expression of the gene rcsA is increased. This gene is a transcriptional regulator of the colanic acid operon. Enhanced expression of this gene increases transcription of the colanic acid operon (13, 36).

Hence the present invention relates to the usage of a mutated and/or transformed microorganism comprising a genetic change leading to a modified expression and/or activity of the transcriptional regulator, the aerobic respiration control protein, ArcA and the isocitrate lyase regulator IcIR to upregulate the transcriptional regulator of the colanic acid operon, rcsA, which in turn upregulates at least one of the genes of the colanic acid operon.

Hence, the present invention relates to a mutated and/or transformed microorganism such as -but not limited to Enterobacteriaceae such as an Escherichia coli (£. coli) strain comprising a genetic change leading to a modified expression of the transcriptional regulators: the aerobic respiration control protein ArcA and the isocitrate lyase regulator IcIR.

A mutated and/or transformed microorganism such as E. coli as used here can be obtained by any method known to the person skilled in the art, including but not limited to UV mutagenesis and chemical mutagenesis. A preferred manner to obtain the latter microorganism is by disrupting (knocking-out) the genes (arcA and icIR) encoding for the proteins ArcA and IcIR, or, by replacing the endogenous promoters of said genes by artificial promoters or replacing the endogenous ribosome binding site by an artificial ribosome binding site. The term 'artificial promoters' relates to heterologous or non-natural or in silico designed promoters with known expression strength, these promoters can be derived from libraries as described by Alper et al. (2005), Hammer et al. (2006), or De Mey et al. (2007) (3, 1 1 , 15). The term heterologous promoter refers to any promoter that does not naturally occur in front of the gene. The term 'artificial promoter' may also refer to promoters with DNA sequences that are combinations of the native (autologous) promoter sequence with parts of different (autologous or heterologous) promoter sequences as for example shown further in the examples. Sequences of such 'artificial promoters' can be found in databases such as for example partsregistry.org (6). The term 'artificial ribosome binding site' relates to heterologous or non-natural or in silico designed ribosome binding sites with known or measurable translation rates, these libraries can be derived from libraries or designed via algorithms as described by Salis et al (2009) (26). Hence, the present invention specifically relates to a mutated and/or transformed microorganism as indicated above wherein said genetic change is disrupting the genes encoding for ArcA and IcIR, or, reducing or eliminating the function of ArcA and IcIR via mutations in the coding sequence of the genes coding for ArcA and IcIR , or, is replacing the endogenous promoters of the genes encoding for ArcA and IcIR by artificial promoters; or, is replacing the endogenous ribosome binding site by an artificial ribosome binding site. It is further clear that the mutant and/or transformant according to the present invention may further comprise additional genetic changes in one or more other genes within its genome as is also described further. The term microorganism specifically relates to a bacterium, more specifically a bacterium belonging to the family of Enterobacteriaceae. The latter bacterium preferably relates to any strain belonging to the species Escherichia coli such as but not limited to Escherichia coii B, Escherichia coii C, Escherichia coli VSI, Escherichia coii K12, Escherichia coii Nissle. More specifically, the latter term relates to cultivated Escherichia coii strains - designated as £ coii K12 strains - which are well-adapted to the laboratory environment, and, unlike wild type strains, have lost their ability to thrive in the intestine. Well-known examples of the £ coii K12 strains are K12 Wild type, W31 10, MG1655, M182, MC1000, MC1060, MC1061 , MC4100, JM101 , NZN1 1 1 and AA200. Hence, the present invention specifically relates to a mutated and/or transformed Escherichia coii strain as indicated above wherein said £ coii strain is a K12 strain. More specifically, the present invention relates to a mutated and/or transformed Escherichia coii strain as indicated above wherein said K12 strain is £. coii MG1655.

The terms 'leading to a modified expression or activity' indicates that the above described mutations/transformations affects the transcription and/or translation and/or post- translational modification of said genes (arcA and icIR) into the transcriptional regulator proteins of the present invention (ArcA and IcIR) in such a way that the latter transcription has significantly decreased or has even been completely abolished compared to a wild type strain, which has not been mutated or transformed with regard to both particular genes of the present invention. Hence, the present invention relates to a mutated and/or transformed microorganism such as an Escherichia coii strain as indicated above wherein said modified expression is a decreased expression, and, to a mutated and/or transformed microorganism such as an Escherichia coii strain as indicated above wherein said decreased expression is an abolished expression.

The terms 'upregulating at least one of the genes of the colanic acid operon' indicates that the expression of at least 1 , 2, 3, 4,..., or all of the genes belonging to the colanic acid operon are significantly (= P > 0.05) upregulated when compared to the expression of said genes within a corresponding wild type microorganism which is cultivated under the same conditions as the mutated and/or transformed microorganism. The genes which belong to the colanic acid operon are wza, wzb, wzc, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, gmd, fcl, gmm, weal, cpsB, cpsG, wcaJ, wzxC, wcaK, wcaL and wcaM as indicated in Fig. 3 and/or as described in (35). Furthermore, the gene rcsA, coding for the transcriptional regulator of the colanic acid operon is upregulated (13, 36). More specifically the terms 'upregulating at least one of the genes of the colanic acid operon' or the transcriptional regulator of the colanic acid operon indicates that at least one of the genes of the colanic acid operon is 6 to 8 times upregulated in comparison to the expression of the genes of the colanic acid operon in the corresponding wild type microorganism. In addition the present invention relates to upregulating genes of the colanic acid operon as described above by replacing the native promoter by an 'artificial promoter'. More specifically, the present invention relates to a combination of the sequence of the native promoter with sequences of other artificial promoter sequences. The combination of the sequence of the native promoter with the sequence of other artificial promoter sequences is more specifically the replacement of the sigma factor binding site of the native promoter with a stronger sigma factor binding site. Sigma factors, such as but not limited to sigma70, sigmaS, sigma24,..., are described (41 ) , subunits of RNA polymerase that determine the affinity for promoter sequences and the transcription rate of genes. The present invention provides microorganisms which are genetically changed in such a manner that they can efficiently produce compounds which are part of the colanic acid pathway. The terms 'compounds which are part of the colanic acid pathway' refer to all compounds as indicated on Figure 4 starting from fructose-6-P and resulting in extracellular colanic acid. More specifically the latter terms refer to the compounds mannose-6-P, mannose-1 -P, GDP-mannose, GDP-4-dehydro-6deoxy-mannose, GDP-fucose and colanic acid. Hence the present invention specifically relates to the usage as indicated for the synthesis of colanic acid and/or for the synthesis of GDP-fucose. As GDP-fucose is a precursor for fucosylated oligosaccharides such as fucosyllactose, fucosyllactoNbiose and lewis X oligosaccharide or fucosylated proteins, and as these sugars have therapeutical, nutraceutical, anti-inflammatory and prebiotic effects, the present invention specifically relates to the usage as described above for the synthesis of fucosylated oligosaccharides. In other words, the present invention relates to a process for the synthesis of colanic acid and/or GDP-fucose and/or fucosylated oligosaccharides comprising genetically changing the transcriptional regulators the aerobic respiration control protein ArcA and the isocitrate lyase regulator IcIR to upregulate at least one of the genes of the colanic acid operon, wherein said operon comprises the genes cpsG, cpsB, gmd and fcl or genes cpsG, cpsB, gmd, fcl and wza. More specifically, the present invention relates to a process as described wherein the mutations for ArcA and IcIR are applied in combination with at least one mutation that enhances the production of fucosylated compounds. In order to efficiently produce fucosylated oligosaccharides (see Figures 1 , 2 and 5-10), the above described mutations in arcA and icIR can be applied in combination with other mutations which further enhance the production of fucosylated compounds. Some of these - non- limiting- other mutations are: a) the deletion of wcaJ from the colanic operon, stopping the initiation of the colanic acid biosynthesis and thus allowing the accumulation of GDP- fucose; b) the introduction of a fucosyltransferase to link fucose with different acceptor molecules such as lactose; c) for the accumulation of the precursor of the GDP-fucose biosynthetic pathway and additional to the deletion of wcaJ, at least one of the following colanic acid operon genes that do not code for GDP-fucose biosynthesis is knocked out: gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaJ, wcaK, wcaL, wzx, wza, wzb, wzc, and/or, wcaM; d) for the production of fucosyllactose, lacZ coding for β- galactosidase, is knocked out to avoid lactose degradation; e) to accumulate the precursor fructose and fructose-6-phosphate, a sucrose phosphorylase or invertase is introduced; f) because fructose-6-phosphate is easily degraded in the glycolysis, the glycolysis has to be interrupted in order to steer all fructose-6-phosphate in the direction of GDP-fucose and the genes pgi, pfkA and pfkB (coding for glucose-6-phosphate isomerase and phosphofructokinase A and B) are thus knocked out; g) reducing protein degradation by knocking out a protease coded by a gene such as Ion; h) By constitutively expressing a lactose permease, subpopulations are avoided in the production process which are common for lactose induced gene expression systems (19). In other words, the present invention relates to a process as described above for the synthesis of fucosylated oligosaccharides wherein said at least one mutation that enhance the production of fucosylated compounds is: the deletion of the wcaJ gene, and/or, knocking-out the colanic acid operon genes gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaJ, wcaK, wcaL, wzx, wza, wzb, wzc, and/or, wcaM, and/or, knocking-out lacZ, and/or, introducing a sucrose phosporylase or invertase, and/or, knocking out the genes pgi, pfkA and pfkB, and/or, knocking out the gene Ion, and/or introducing a fucosyltransferase, and/or a lactose permease. The term 'introducing a fucosyltransferase' relates to upregulating or heterologous expression of fucosyltransferases which are within, but not limited to the enzymes in enzyme classes classes EC2.4.1 .65, 2.4.1 .68, 2.4.1 .69, 2.4.1 .152, 2.4.1 .214, and/or 2.4.1.221 and/or the glycosyltransferase families GT1 , GT2, GT10 GT1 1 , GT23, GT37, GT65, GT68, and/or GT74 and/or originating from but not limited to Helicobacter pylori, Campylobacter jejuni, Dictyostellium discoideum, Mus musculus, Homo sapiens, ... and these fucosyltransferases catalyse the formation of a(1 ,2), a(1 ,3), a(1 ,4), or a(1 ,6) bonds on other sugars such as but not limited to galactose, lactose, lacto/Vbiose, lacto/Vtetraose, lactosamine, lacto/Vtetraose, sialyllactoses, disialyllactoses, or fucosylated proteins, or fucosylated fatty acids., or fucosylated aglycons such as, but not limited to, antivirals, antibiotics,... .

The present invention provides for the usage of a mutated and/or transformed microorganism comprising a genetic change leading to a modified expression and/or activity of the transcriptional regulators the aerobic respiration control protein ArcA and the isocitrate lyase regulator IcIR to upregulate at least one of the genes of the colanic acid operon, wherein said operon comprises the genes cpsG and cpsB, coding for phosphomannomutase and mannose-1 -phosphate guanylyltransferase, which are needed for the biosynthesis of GDP-mannose. As GDP-mannose is a precursor for mannosyllated oligosaccharides and mannosylated glycoconjugates. These oligosaccharides and glycoconjugates find for example applications in the treatment of gram-negative bacterial infections, in addition, GDP-mannose is important for the humanization of protein glycosylations, which is essential for the production of certain therapeutic proteins (18, 30). Mannosylated oligosaccharides and mannosylated glycoconjugates are also used for drug targeting, for instance mannosylated antivirals can specifically target the liver and kidneys (7). In order to efficiently produce mannosylated oligosaccharides (see Figures 1 , 2, 5, 6 and 1 1 ), the above described mutations in arcA and icIR can be applied in combination with other mutations which further enhance the production of mannosylated compounds. Some of these - non-limiting- other mutations are: a) the gene gmd of the colanic acid operon is deleted, and/or, b) wherein the gene gmm coding for GDP- mannose hydrolase is deleted, and/or, c) wherein the colanic acid operon genes that do not code for GDP-mannose biosynthesis reactions, the genes gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaJ, wcaK, wcaL, fcl, gmd, wzx, wza, wzb and/or, wcaM, are deleted, and/or, d) wherein a gene encoding for a sucrose phosphorylase or an invertase is introduced, and/or, e) wherein the the genes pgi, pfkA and pfkB, coding for phosphoglucose isomerase, phosphofructokinase A and phosphofructokinase B respectively, are deleted, and/or, f) knocking out the gene Ion encoding for a protease, and/or f) wherein a gene encoding for a mannosyltransferase is introduced. In other words, the present invention relates to a process as described above for the synthesis of colanic acid and/or GDP-fucose and/or fucosylated oligosaccharides for the synthesis of GDP-mannose and/or for the synthesis of mannosylated oligosaccharides. The present invention further relates to said process wherein the genes cpsG and cpsB of the colanic acid operon are upregulated and wherein: a) the gene gmd of the colanic acid operon is deleted, and/or, b) wherein the gene gmm is deleted, and/or c) wherein the colanic acid operon genes fcl, gmd, gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaJ, wcaK, wcaL, wzx, wza, wzb, wzc, and/or, wcaM are knocked out and/or, d) wherein a gene encoding for a sucrose phosphorylase or an invertase is introduced, and/or, e) wherein the the genes pgi, pfkA and pfkB are deleted, and/or, f) knocking out the gene Ion, and/or g) wherein a gene encoding for a mannosyltransferase is introduced. The term 'introducing a mannosyltransferase' relates to upregulating or heterologous expression of mannosyltransferases which are within, but not limited to the enzymes in enzyme classes EC 2.4.1.32, 2.4.1.B27, 2.4.1 .B44, 2.4.1 .48, 2.4.1 .54, 2.4.1 .57, 2.4.1 .83, 2.4.1 .109, 2.4.1.1 10, 2.4.1 .1 19, 2.4.1.130, 2.4.1.131 , 2.4.1 .132, 2.4.1.142, 2.4.1 .199, 2.4.1.217, 2.4.1.232, 2.4.1.246, 2.4.1 .251 , 2.4.1.252, 2.4.1 .257, 2.4.1.258, 2.4.1 .259, 2.4.1.260, 2.4.1.265, and/or 2.4.1.270 and/or the glycosyltransferase families GT1 , GT2, GT4, GT15, GT22, GT32, GT33, GT39, GT50 and/or GT58 and/or originating from but not limited to Helicobacter pylori, Campylobacter jejuni, Dictyostellium discoideum, Mus musculus, Homo sapiens, ... and these mannosyltransferases catalyse the formation of a(1 ,2), a(1 ,3), a(1 ,4), or a(1 ,6) bonds on other sugars such as but not limited to galactose, N- acetylglucosamine, Rhamnose, lactose, lacto/Vbiose, lacto/Vtetraose, lactosamine, lacto/Vtetraose, sialyllactoses, disialyllactoses, or mannosylated proteins, or mannosylated fatty acids, or mannosylated aglycons such as, but not limited to, antivirals, antibiotics,... .

The term 'heterologous expression' relates to the expression of genes that are not naturally present in the production host, genes which can be synthesized chemically or be picked up from their natural host via PCR, genes which can be codon optimized for the production host or in which point mutation can be added to enhance enzyme activity or expression. Expressing heterologous and/or native genes can either be done on the chromosome, artificial chromosomes or plasmids and transcription can be controlled via inducible, constitutive, native or artificial promoters and translation can be controlled via native or artificial ribosome binding sites.

Consequently, the present invention further relates to mutated and/or transformed organisms in which the regulators ArcA and IcIR as describe above, in combination with the genes encoding for the enzymes phosphoglucose isomerase and phosphofructokinase, are knocked out or are rendered less functional. More specifically, the present invention relates to the latter organisms wherein the enzyme phosphoglucose isomerase is encoded by the gene pgi and wherein the enzyme phosphofructokinase is encoded by the gene(s) pfkA and/or pfkB.

The terms 'genes which are rendered less-functional or non-functional' refer to the well- known technologies for a skilled person such as the usage of siRNA, RNAi, miRNA, asRNA, mutating genes, knocking-out genes, transposon mutagenesis, etc... which are used to change the genes in such a way that they are less able (i.e. statistically significantly 'less able' compared to a functional wild-type gene) or completely unable (such as knocked-out genes) to produce functional final products. The term '(gene) knock out' thus refers to a gene which is rendered non-functional. The term 'deleted gene' or 'gene deletion' also refers to a gene which is rendered non-functional.

The present invention further relates to a mutated and/or transformed organism as described in the latter paragraph wherein said organism is further transformed with a gene encoding for a sucrose phosphorylase.

The present invention also relates to a mutated and/or transformed organism as described above wherein, in addition, the activity and/or expression of the gene encoding for a lactose permease is made constitutive and/or increased. Said activity can be increased by over-expressing said gene and/or by transforming said organisms with a gene encoding for a lactose permease.

The present invention further relates to any mutated and/or transformed organism as described above wherein at least one of the following genes is knocked out or is rendered less functional:

a gene encoding for a beta-galactosidase, a gene encoding for a glucose-1 -phosphate adenylyltransferase, a gene encoding for a glucose-1 -phosphatase, a gene encoding for phosphogluconate dehydratase, a gene encoding for 2-keto-3-deoxygluconate-6- phosphate aldolase ,a gene encoding for a glucose-1 -phosphate uridyltransferase, a gene encoding for an UDP-glucose-4-epimerase, a gene encoding for an UDP- glucose:galactose-1 -phosphate uridyltransferase, a gene encoding for an UDP- galactopyranose mutase, a gene encoding for an UDP- galactose:(glucosyl)lipopolysaccharide-1 ,6-galactosyltransferase, a gene encoding for an UDP-galactosyltransferase, a gene encoding for an UDP-glucosyltransferase, a gene encoding for an UDP-glucuronate transferase, a gene encoding for an UDP-glucose lipid carrier transferase, a gene encoding for a GDP-mannose hydrolase, a gene encoding for an UDP-sugar hydrolase, a gene encoding for a mannose-6-phosphate isomerase, a gene encoding for an UDP-N-acetylglucosamine enoylpyruvoyl transferase, a gene encoding for an UDP-N-acetylglucosamine acetyltransferase, a gene encoding for an UDP-Nacetylglucosamine-2-epimerase, a gene encoding for an undecaprenyl-phosphate alfa-N-acetylglucosaminyl transferase, a gene encoding for a glucose-6-phosphate-1 - dehydrogenase, and/or, a gene encoding for a L-glutamine:D-fructose-6-phosphate aminotransferase, a gene encoding for a mannose-6-phosphate isomerase, a gene encoding for a sorbitol-6-phosphate dehydrogenase, a gene encoding for a mannitol-1 - phosphate 5-dehydrogenase, a gene encoding for a allulose-6-phosphate 3-epimerase, a gene encoding for an invertase, a gene encoding for a maltase, a gene encoding for a trehalase, a gene encoding for a sugar transporting phosphotransferase, a gene encoding for a protease, or a gene encoding for a hexokinase. The term 'at least one' indicated that at least 1 , but also 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 , 32 or all 33 genes is (are) knocked out or is (are) rendered less functional.

The present invention further relates also to the usage of a mutated and/or transformed microorganism such as an Escherichia coli strain comprising a genetic change leading to a modified expression of the transcriptional regulators the aerobic respiration control protein ArcA and the isocitrate lyase regulator IcIR to upregulate at least one of the following acid resistance related genes: ydeP, ydeO, hdeA, hdeD, gadB, gadC, gadE, gadX, gadW and/or sip (17, 22). These genes are normally expressed in stationary phase conditions; however, the present mutated and/or transformed microorganism is able to enhance the expression of these acid resistance related genes in the exponential growth phase. Hence, the present invention relates to the usage as described above for the synthesis of acids or pH sensitive molecules such as but not limited to glucosamine which is pH sensitive and should be produced at low pH (12). Organic acids, such as but not limited to pyruvic acid, succinic acid, adipic, sialic acid, sialylated oligosaccharides (e.g. sialyllactose, sialyl Lewis X sugars,... ), acetylated oligosaccharides (chitins, chitosans,... ), sulfonated oligosaccharides (heparans and heparosans)... are preferably produced at low pH for downstream processing purposes (4). In other words, the present invention relates to a process for the synthesis of acids, sialic acid, sialylated oligosaccharides or glucosamine comprising genetically changing the transcriptional regulators the aerobic respiration control protein ArcA and the isocitrate lyase regulator IcIR to upregulate at least one of the following acid resistance related genes: ydeP, ydeO, hdeA, hdeD, gadB, gadC, gadE, gadX, gadW and/or sip.

The present invention will now be illustrated by the following non-limiting examples.

EXAMPLES

A high throughput RT-qPCR screening of the microorganisms of the present invention has been setup with Biotrove OpenArray ^® technology. In this experiment the transcription of 1800 genes were measured in 4 strains (wild type, AarcA, AiclR, AarcA AiclR) in two conditions (chemostat and batch). The data was processed using a curve fitting toolbox in R (25, 34) and Quantile Normalization, the error on the data was calculated using Bayesian statistics (20, 21 , 31 ).

Material and methods

Strains and plasmids

Escherichia coli MG1655 [ ^" , F ^" , rp/7-1] was obtained from the Netherlands Culture Collection of Bacteria (NCCB). Escherichia coli BL21 (DE3) was obtained from Novagen. Escherichia coli MG1655 ackA-pta, poxB, pppc ppc-p37 (10), the single knock-outs E. coli MG1655 arcA and E. coli MG1655 icIR and the double knock-out E. coli MG1655 arcA, icIR were constructed in the Laboratory of Genetics and Microbiology (MICR) using the method of Datsenko & Wanner (9).

Media

The Luria Broth (LB) medium consisted of 1 % tryptone peptone (Difco, Erembodegem, Belgium), 0.5 % yeast extract (Difco) and 0.5 % sodium chloride (VWR, Leuven, Belgium). Shake flask medium contained 2 g/l NH ₄ CI, 5 g/l (NH ₄ ) ₂ S0 ₄ , 2.993 g/l KH ₂ P0 ₄ , 7.315 g/l K ₂ HP0 ₄ , 8.372 g/l MOPS, 0.5 g/l NaCI, 0.5 g/l MgS0 ₄ -7H ₂ 0, 16.5 g/l glucose-H ₂ 0, 1 ml/l vitamin solution, 100 μΙ/Ι molybdate solution, and 1 ml/l selenium solution. The medium was set to a pH of 7 with 1 M KOH.

Vitamin solution consisted of 3.6 g/l FeCI ₂ ^■ 4H ₂ 0, 5 g/l CaCI ₂ ^■ 2H ₂ 0, 1 .3 g/l MnCI ₂ ^■ 2H ₂ 0, 0.38 g/l CuCI ₂ ^■ 2H ₂ 0, 0.5 g/l CoCI ₂ ^■ 6H ₂ 0, 0.94 g/l ZnCI ₂ , 0.031 1 g/l H ₃ B0 ₄ , 0.4 g/l Na ₂ EDTA- 2H ₂ 0 and 1 .01 g/l thiamine ^■ HCI. The molybdate solution contained 0.967 g/l Na ₂ Mo0 ₄ ^■ 2H ₂ 0. The selenium solution contained 42 g/l Se0 ₂ .

The minimal medium for fermentations contained 6.75 g/l NH ₄ CI, 1 .25 g/l (NH ₄ ) ₂ S0 ₄ , 1 .15 g/l KH ₂ P0 ₄ , 0.5 g/l NaCI, 0.5 g/l MgS0 ₄ -7H ₂ 0, 16.5 g/l glucose-H ₂ 0, 1 ml/l vitamin solution, 100 μΙ/Ι molybdate solution, and 1 ml/l selenium solution with the same composition as described above.

Cultivation conditions

A preculture, from a single colony on a LB-plate, in 5 ml LB medium was incubated during 8 hours at 37 °C on an orbital shaker at 200 rpm. From this culture, 2 ml was transferred to 100 ml minimal medium in a 500 ml shake flask and incubated for 16 hours at 37 °C on an orbital shaker at 200 rpm. 4 % inoculum was used in a 2 I Biostat B Plus culture vessel with 1 .5 I working volume (Sartorius Stedim Biotech, Melsungen, Germany). The culture conditions were: 37 °C, stirring at 800 rpm, and a gas flow rate of 1 .5 l/min. Aerobic conditions were maintained by sparging with air, anaerobic conditions were obtained by flushing the culture with a mixture of 3 % C0 ₂ and 97 % of N ₂ . The pH was maintained at 7 with 0.5 M H ₂ S04 and 4 M KOH. The exhaust gas was cooled down to 4 °C by an exhaust cooler (Frigomix 1000, Sartorius Stedim Biotech, Melsungen, Germany). 10 % solution of silicone antifoaming agent (BDH 331512K, VWR Int Ltd., Poole, England) was added when foaming raised during the fermentation (approximately 10 μΙ). The off-gas was measured with an EL3020 off-gas analyser (ABB Automation GmbH, 60488 Frankfurt am Main, Germany). All data was logged with the Sartorius MFCS/win v3.0 system (Sartorius Stedim Biotech, Melsungen, Germany).

All strains were cultivated at least twice and the given standard deviations on yields and rates are based on at least 10 data points taken during the repeated experiments. Sampling methodology

The bioreactor contains in its interior a harvest pipe (BD Spinal Needle, 1.2x152 mm (BDMedical Systems, Franklin Lakes, NJ - USA) connected to a reactor port, linked outside to a Masterflex-14 tubing (Cole-Parmer, Antwerpen, Belgium) followed by a harvest port with a septum for sampling. The other side of this harvest port is connected back to the reactor vessel with a Masterflex-16 tubing. This system is referred to as rapid sampling loop. During sampling, reactor broth is pumped around in the sampling loop. It has been estimated that, at a flow rate of 150 ml/min, the reactor broth needs 0.04 s to reach the harvest port and 3.2 s to re-enter the reactor. At a p02 level of 50 %, there is around 3 mg/l of oxygen in the liquid at 37 °C. The p02 level should never drop below 20 % to avoid micro-aerobic conditions. Thus 1 .8 mg/l of oxygen may be consumed during transit through the harvesting loop. Assuming an oxygen uptake rate of 0.4 g oxygen/g biomass/h (the maximal oxygen uptake rate found at _max ), this gives for 5 g/l biomass, an oxygen uptake rate of 2 g/l/h or 0.56 mg/l/s, which multiplied by 3.2 s (residence time in the loop) gives 1 .8 mg/l oxygen consumption.

In order to quench the metabolism of cells during the sampling, reactor broth was sucked through the harvest port in a syringe filled with 62 g stainless steel beads pre-cooled at -20 °C, to cool down 5 ml broth immediately to 4 °C. Sampling was immediately followed by cold centrifugation (15000 g, 5 min, 4 °C). During the batch experiments, a sample for OD ₆ oonm and RT-qPCR measurements was taken using the rapid sampling loop and the cold stainless bead sampling method.

RT-qPCR

mRNA was extracted with the RNeasy kit (Qiagen,Venlo, The Netherlands). RNA quality and quantity was checked with a nanodrop ND-1000 spectrophotometer (Nanodrop technologies, Wilmingto, USA). The ratios 260:280 (nm) and 260:230 (nm) were between 1 .8 and 2 and at least 10Ong/μΙ was needed for further analysis. cDNA was synthesised with random primers with the RevertAid™ H minus first strand cDNA synthesis kit (Fermentas, St. Leon-Rot, Germany). Finally, the gene expression of 1800 genes was measured with the Biotrove OpenArray Real time PCR platform. The primers for the RT- PCR assay were designed with Primer design tools from the Primer database (23). The reaction mixture was composed as described in the Biotrove OpenArray™ Real-Time qPCR system users' manual. In short, a mastermix was made with 26.4μΙ LightCycler® DNA Master SYBR® Green I (Roche applied Science), 1.1 μΙ SYBR GREEN I (100x stock solution, Sigma S9430), 8.8μΙ glycerol (Sigma G5150), 5.3μΙ Pluronic® F68 (10% stock, Invitrogen), 2.64μΙ BSA (Sigma A7906), 26.4μΙ magnesium chloride (25mM stock solution, supplied in the LightCycler® kit of Roche applied Science ), 21 .1 μΙ HiDi™ formamide (Applied biosystems), and 94.66μΙ RNase free sterile water resulting in a 186.4μΙ mastermix, which is enough to load 1 OpenArray™. For 1 SubArray (each OpenArray is subdivided in 48 SubArrays on which 1 sample can be loaded) 1 .5μΙ sample (with a concentration of 10Ong/μΙ) was mixed with 3.5μΙ of mastermind, as a no template control, water was used as blanc. The sample-mastermix mixture was loaded in a Loader plate (MatriPlate™ 384-well black low volume polypropylene plate, Biotrove) in a RNase free hood. A full loader plate was loaded with an AutoLoader (Biotrove) and loader tips onto the OpenArrays. These OpenArrays were then submerged in OpenArray™ immersion fluid in an OpenArray™ Real-Time qPCR case. The case was sealed with Case sealing glue and incubated in the Case Sealing station, which polymerizes the glue with UV light.

Analytical methods

Cell density of the culture was frequently monitored by measuring optical density at 600 nm (Uvikom 922 spectrophotometer, BRS, Brussel, Belgium). Cell dry weight was obtained by centrifugation (15 min, 5000 g, GSA rotor, Sorvall RC-5B, Goffin Meyvis, Kapellen, Belgium) of 20 g reactor broth in pre-dried and weighted falcons. The pellets were subsequently washed once with 20 ml physiological solution (9 g/l NaCI) and dried at 70 °C to a constant weight. To be able to convert OD ₆ oonm measurements to biomass concentrations, a correlation curve of the OD ₆ oonm to the biomass concentration was made. The concentrations of glucose and organic acids were determined on a Varian Prostar HPLC system (Varian, Sint-Katelijne-Waver, Belgium), using an Aminex HPX-87H column (Bio-Rad, Eke, Belgium) heated at 65 °C, equipped with a 1 cm precolumn, using 5 mM H2S04 (0.6 ml/min) as mobile phase. A dual-wave UV-VIS (210 nm and 265 nm) detector (Varian Prostar 325) and a differential refractive index detector (Merck LaChrom L-7490, Merck, Leuven, Belgium) was used for peak detection. By dividing the absorptions of the peaks in both 265 and 210 nm, the peaks could be identified. The division results in a constant value, typical for a certain compound (formula of Beer- Lambert). Glucose, fructose, sucrose, fu cosy I lactose and glucose-1 -phosphate were measured by HPLC with a Hypercarb column and were detected with an MSMS detector (Antonio et al., 2007; Nielsen et al., 2006).

Genetic methods

All mutant strains were constructed via the methods described below.

Plasmids were maintained in the host E. coli DH5oc (F ^" , cp80d/acZAM15, A(lacZYA- argF)U169, deoR, recA , endA'\ , /?sc/R17(rk ^" , mk ⁺ ), phoA, supE44, K, i/?/ ^' -1 , gyrA96, ΓθΙΑλ ).

Plasmids. pKD46 (Red helper plasmid, Ampicillin resistance), pKD3 (contains an FRT- flanked chloramphenicol resistance (cat) gene), pKD4 (contains an FRT-flanked kanamycin resistance (kan) gene), and pCP20 (expresses FLP recombinase activity) plasmids were used for the mutant construction. The plasmid pBluescript (Fermentas, St. Leon-Rot, Germany) was used to construct the derivates of pKD3 and pKD4 with a promoter library, or with alleles carrying a point mutation.

Mutations. The mutations consisted in gene disruption (knock-out, KO). They were introduced using the concept of Datsenko and Wanner (9). The primers for the mutation strategies are described in Table 1.

Transformants carrying a Red helper plasmid were grown in 10 ml LB media with ampicillin (100 mg/l) and L-arabinose (10 mM) at 30 °C to an OD ₆₀ o _n m of 0.6. The cells were made electrocompetent by washing them with 50 ml of ice-cold water, a first time, and with 1 ml ice-cold water, a second time. Then, the cells were resuspended in 50 μΙ of ice-cold water. Electroporation was done with 50 μΙ of cells and 10-100 ng of linear double-stranded-DNA product by using a Gene Pulser™ (BioRad) (600 Ω, 25 FD, and 250 volts).

After electroporation, cells were added to 1 ml LB media incubated 1 h at 37 °C, and finally spread onto LB-agar containing 25 mg/l of chloramphenicol or 50 mg/l of kanamycin to select antibiotic resistant transformants. The selected mutants were verified by PCR with primers upstream and downstream of the modified region and were grown in LB-agar at 42 °C for the loss of the helper plasmid. The mutants were tested for ampicillin sensitivity.

Linear double-stranded-DNA. The linear ds-DNA amplicons were obtained by PCR using pKD3, pKD4 and their derivates as template. The primers used had a part of the sequence complementary to the template and another part complementary to the side on the chromosomal DNA where the recombination has to take place (

Table 1 ). For the KO, the region of homology was designed 50-nt upstream and 50-nt downstream of the start and stop codon of the gene of interest. For the Kl, the transcriptional starting point (+1 ) had to be respected. PCR products were PCR-purified, digested with Dpn\, repurified from an agarose gel, and suspended in elution buffer (5 mM Tris, pH 8.0).

Elimination of the antibiotic resistance gene. The selected mutants (chloramphenicol or kanamycin resistant) were transformed with pCP20 plasmid, which is an ampicillin and chloramphenicol resistant plasmid that shows temperature-sensitive replication and thermal induction of FLP synthesis. The ampicillin-resistant transformants were selected at 30 °C, after which a few were colony purified in LB at 42 °C and then tested for loss of all antibiotic resistance and of the FLP helper plasmid. The gene knock outs and knock ins are checked with control primers (Fw/Rv-gene-out). These primers are given in Table 1.

Table 1 : Primers used to create E. coli MG1655 arcA, E. coli MG1655 icIR and the double knock-out E. coli MG1655 arcA, icIR and all other genetic knock outs and knock ins

Primer name Sequence

lacZ

FW LacZ PI CATAATGGATTTCCTTACGCGAAATACGGGCAGACATGGCCTGCCCGGTTATTAgtgta ggctggagctgcttc (SEQ ID N° 7)

RV LacZ P2 GTATGTTGTGTGGAATTGTGAGCGGATAACAATTTCACACAGGAAACAGCTcatatgaa tatcctccttag (SEQ ID N° 8)

FW LacZ out GCGGTTGGAATAATAGCG (SEQ ID N° 9)

RV LacZ out CAGGTTTCCCGACTGGAAAG (SEQ ID N° 10)

glgC

FW-glgC-Pl Agaccgccggttttaagcagcgggaacatctctgaacatacatgtaaaacctgcagtgt aggctggagctgcttc (SEQ ID N° 11)

RV-glgC-P2 Gtctggcagggacctgcacacggattgtgtgtgttccagagatgataaaaaaggagtta gtccatatgaatatcctccttag (SEQ ID N° 12)

FW-glgC-out Gcgaatatcgggaaatgcagg (SEQ ID N° 13)

RV-glgC-out Cagagattgttttacctgctgg (SEQ ID N° 14)

agp FW agp PI CATATTTCTGTCACACTCTTTAGTGATTGATAACAAAAGAGGTGCCAGGAgtgtaggct ggagctgcttc (SEQ ID N° 15)

RV agp P2 TAAAAACGTTTAACCAGCGACTCCCCCGCTTCTCGCGGGGGAGTTTTCTGcatatgaat atcctccttag (SEQ ID N° 16)

FW agp out GCCACAGGTGCAATTATC (SEQ ID N° 17)

RV agp out CATTTTCGAAGTCGCCGGGTACG (SEQ ID N° 18)

pgi

Fw-pgi-Pl GGCGCTACAATCTTCCAAAGTCACAATTCTCAAAATCAGAAGAGTATTGCgtgtaggct ggagctgcttc (SEQ ID N° 19)

Rv-pgi-P2 GGTTGCCGGATGCGGCGTGAACGCCTTATCCGGCCTACATATCGACGATGcatatgaat atcctccttag (SEQ ID N° 20)

Fw pgi out ( GGCTCCTCCAACACCGTTAC (SEQ ID N° 21)

2)

Rv pgi out ( TACATATCGGCATCGACCTG (SEQ ID N° 22)

2)

pfkA

Fw-pfkA-out TACCGCCATTTGGCCTGAC (SEQ ID N° 23)

Rv-pfkA-out AAAGTGCGCTTTGTCCATGC (SEQ ID N° 24)

Fw-pfkA-Pl GACTTCCGGCAACAGATTTCATTTTGCATTCCAAAGTTCAGAGGTAGTCgtgtaggctg gagctgcttc (SEQ ID N° 25)

Rv-pfkA-P2 GCTTCTGTCATCGGTTTCAGGGTAAAGGAATCTGCCTTTTTCCGAAATCcatatgaata tcctccttag (SEQ ID N° 26)

pfkB

Fw-pfkB-out TAGCGTCCCTGGAAAGGTAAC (SEQ ID N° 27)

Rv-pfkB-out TCCCTCATCATCCGTCATAG (SEQ ID N° 28)

Fw-pfkB-Pl CACTTTCCGCTGATTCGGTGCCAGACTGAAATCAGCCTATAGGAGGAAATGgtgtaggc tggagctgcttc (SEQ ID N° 29)

Rv-pfkB-P2 GTTGCCGACAGGTTGGTGATGATTCCCCCAATGCTGGGGGAATGTTTTTGcatatgaat atcctccttag (SEQ ID N° 30)

arcA

FW-arcA-Pl Ggttgaaaaataaaaacggcgctaaaaagcgccgttttttttgacggtggtaaagccga gtgtaggctggagctgcttc (SEQ ID N° 31)

RV-arcA-P2 Ggtcagggacttttgtacttcctgtttcgatttagttggcaatttaggtagcaaaccat atgaatatcctccttag (SEQ ID N° 32)

FW-arcA-out Ctgccgaaaatgaaagccagta (SEQ ID N° 33) RV-arcA-out Ggaaagtgcatcaagaacgcaa (SEQ ID N° 34)

iclR

FW-iclR-Pl Ttgccactcaggtatgatgggcagaatattgcctctgcccgccagaaaaaggtgtaggc tggagctgcttc (SEQ ID N° 35)

RV-iclR-P2 Gttcaacattaactcatcggatcagttcagtaactattgcattagctaacaataaaaca tatgaatatcctccttag (SEQ ID N° 36)

FW-iclR-out Cggtggaatgagatcttgcga (SEQ ID N° 37)

RV-iclR-out Acttgctcccgacacgctca (SEQ ID N° 38)

FW iclR P8 TTGCCACTCAGGTATGATGGGCAGAATATTGCCTCTGCCCGCCAGAAAAAGccgcttac agacaagctgtg (SEQ ID N° 39)

RV iclR P9 GTTCAACATTAACTCATCGGATCAGTTCAGTAACTATTGCATTAGCTAACAATAAAAag ccatgacccgggaattac (SEQ ID N° 40)

Rv-iclR- CTATTGCATTAGCTAACAATAAAACTTTTTCTGGCGGGCAGAGG (SEQ ID N° 41) scarless KO

stap 2

Fw-iclR- CCTCTGCCCGCCAGAAAAAGTTTTATTGTTAGCTAATGCAATAGTTAC (SEQ ID N° scarless KO 42)

stap 2

wcaJ

Fw wcaJ out GCCAGCGCGATAATCACCAG (SEQ ID N° 43)

Rv wcaJ out TGCGCCTGAATGTGGAATC (SEQ ID N° 44)

Fw-wcaJ 2- TTTTGATATCGAACCAGACGCTCCATTCGCGGATGTACTCAAGGTCGAACgtgtaggct Pl ggagctgcttc (SEQ ID N° 45)

Rv-wcaJ 2- TCTATGGTGCAACGCTTTTCAGATATCACCATCATGTTTGCCGGACTATGcatatgaat P2 atcctccttag (SEQ ID N° 46)

fw wcaJ HI ' TCAATATGCCGCTTTGTTAACGAAACCTTTGAACACCGTCAGGAAAACGATTTTGATAT

CGAACCAGACG (SEQ ID N° 47)

Rv wcaJ H2 ' TGACAAATCTAAAAAAGCGCGAGCGAGCGAAAACCAATGCATCGTTAATCTCTATGGTG

CAACGCTTTTC (SEQ ID N° 48)

Fw wcaJ HI ' CGCTTTGTTAACGAAACCTTTGAACACCGTCAGGAAAACGATTTTGATATCGAACCAGA 2 CGCTCCATTCG (SEQ ID N° 49)

Ion

FW-lon-Pl CAGTCGTGTCATCTGATTACCTGGCGGAAATTAAACTAAGAGAGAGCTCTgtgtaggct ggagctgcttc (SEQ ID N° 50)

oMEMOlOO RV CGAATTAGCCTGCCAGCCCTGTTTTTATTAGTGCATTTTGCGCGAGGTCAcatatgaat -lon-P2 atcctccttag (SEQ ID N° 51) οΜΕΜΟΙΟΙ FW AGCGCAACAGGCATCTGGTG (SEQ ID N° 52)

-lon-out

OMEMO102 RV TATATCAGGCCAGCCATCCC (SEQ ID N° 53)

-lon-out

lacZYA:P22- lacY

Fw lacZYA c GCTGAACTTGTAGGCCTGATAAGCGCAGCGTATCAGGCAATTTTTATAATCTTCATTTA hi AATGGCGCGC (SEQ ID N° 54)

rv lacZYA c GCGCAACGCAATTAATGTGAGTTAGCTCACTCATTAGGCACCCCAGGCTTCGCCTACCT hi GTGACGGAAG (SEQ ID N° 55)

fw P221acY- GCGCAACGCAATTAATGTGAGTTAGCTCACTCATTAGGCACCCCAGGCTTGTGTAGGCT KI PI GGAGCTGCTTC (SEQ ID N° 56)

rv P221acY- GCTGAACTTGTAGGCCTGATAAGCGCAGCGTATCAGGCAATTTTTATAATCTTAAGCGA KI CTTCATTCACC (SEQ ID N° 57)

fw lacZYA H CGACGCTTGTTCCTGCGCTTTGTTCATGCCGGATGCGGCTAATGTAGATCGCTGAACTT 1 ' GTAGGCCTG (SEQ ID N° 58)

rv lacZYA H CATTAATGCAGCTGGCACGACAGGTTTCCCGACTGGAAAGCGGGCAGTGAGCGCAACGC 2 ' AATTAATGTG (SEQ ID N° 59)

pfkA:P22- BaSP

Fw-pfkA-Pl GACTTCCGGCAACAGATTTCATTTTGCATTCCAAAGTTCAGAGGTAGTCgtgtaggctg gagctgcttc (SEQ ID N° 60)

Rv-pfkA- GCTTCTGTCATCGGTTTCAGGGTAAAGGAATCTGCCTTTTTCCGAAATCaagcttgcat pCXP22 P2 gcctgcatcc (SEQ ID N° 61)

FW kan AGAGGCTATTCGGCTATGAC (SEQ ID N° 62)

Fw baSP seq CGCCATGTTGGAATGGGAGG (SEQ ID N° 63)

Fw pfkA HI TGATTGTTATACTATTTGCACATTCGTTGGATCACTTCGATGTGCAAGAAGACTTCCGG ext CAACAGA C (SEQ ID N° 64)

Rv pfkA H2 AATTGCAGAATTCATGTAGGCCTGATAAGCGAAGCGCATCAGGCATTTTTGCTTCTGTC ext ATCGGTTTCAG (SEQ ID N° 65)

Fw-pfkA-out TACCGCCATTTGGCCTGAC (SEQ ID N° 66)

Rv-pfkA-out AAAGTGCGCTTTGTCCATGC (SEQ ID N° 67)

adhE:P22- frk

Fw-adhE- ATCGGCATTGCCCAGAAGGGGCCGTTTATGTTGCCAGACAGCGCTACTGAgtgtaggct pCXP22-Pl ggagctgcttc (SEQ ID N° 68)

Rv-adhE- ATTCGAGCAGATGATTTACTAAAAAAGTTTAACATTATCAGGAGAGCATTaagcttgca pCXP22-P2 tgcctgcatcc (SEQ ID N° 69)

Fw-adhE-Hl ' AAGCCGTTATAGTGCCTCAGTTTAAGGATCGGTCAACTAATCCTTAACTGATCGGCATT

GCCCAGAAG (SEQ ID N° 70)

Rv-adhE-H2 ' TTGATTTTCATAGGTTAAGCAAATCATCACCGCACTGACTATACTCTCGTATTCGAGCA

GATGATT AC AAAAAAG (SEQ ID N° 71)

FW adhE out GCGTCAGGCAGTGTTGTATC (SEQ ID N° 72)

RV adhE out CTGGAAGTGACGCATTAGAG (SEQ ID N° 73)

ldhA:P14-

FT H.

pylori

FW ldhA out tgtcattacttacacatcccgc (SEQ ID N° 74)

RV ldhA out gcattcaatacgggtattgtgg (SEQ ID N° 75)

Fw-ldhA- CATTGGGGATTATCTGAATCAGCTCCCCTGGAATGCAGGGGAGCGGCAAGgtgtaggct pCXP22 PI ggagctgcttc (SEQ ID N° 76)

Rv-ldhA- TATTTTTAGTAGCTTAAATGTGATTCAACATCACTGGAGAAAGTCTTATGaagcttgca pCXP22 P2 tgcctgcatcc (SEQ ID N° 77)

Fw-ldhA-Hl ' CAATTACAGTTTCTGACTCAGGACTATTTTAAGAATAGAGGATGAAAGGTCATTGGGGA

TTATCTGAATCAG (SEQ ID N° 78)

Rv-ldhA-H2 ' GAATTTTTCAATATCGCCATAGCTTTCAATTAAATTTGAAATTTTGTAAAATATTTTTA

GTAGCTTAAATGTGATTCAAC (SEQ ID N° 79)

Fw-ldhA- TTCACCGCTAAAGCGGTTAC (SEQ ID N° 80)

long homol

Rv-ldhA- CGCGTAATGCGTGGGCTTTC (SEQ ID N° 81)

long homol

promCA:P14

pCXP14 SP F CCGGCATATGGTATAATAGGG (SEQ ID N° 82)

w

yegH rc pur ACGGCTTGCTGGCCATCA (SEQ ID N° 83)

e rv

fw P14- CGAATATAAGGTGACATTATGGTAATTGAATATTGGCTTTCCAATAATGCTACGGCCCC CA KI tetA AAGGTCCAA (SEQ ID N° 84)

rv P14- AATATTGTCAACCTAAAGAAACTCCTAAAAACCATATTGAATGACACTTATTGGCTTCA CA KI tetA GGGATGAGGCG (SEQ ID N° 85)

fw P14- TCCCGACTACGTGGACCTTG (SEQ ID N° 86)

CA KI overl

apA

rv P14- CATATGGTATAATAGGGAAATTTCCATGGCGGCCGCTCTAGAAGAAGCTTGGGATCCGT

CA KI overl CGACCTCGGCATTATTGGAAAGCCAATATTC (SEQ ID N° 87) apA

fw P14- GCCGCCATGGAAATTTCCCTATTATACCATATGCCGGCCAAGATGTCAAGAAACTTATA

CA KI overl GAATGAAGTAAGTGTCATTCAATATGG (SEQ ID N° 88)

apB

fw P14- AA ATTGTCAACC AAAGAAACTCC AAAAACCA ATTGAATGACACTTACTTCATTCT CA KI HI ATAAGTTTCTTGAC (SEQ ID N° 89)

rv P14- CGAATATAAGGTGACATTATGGTAATTGAATATTGGCTTTCCAATAATGCCGAGGTCGA CA KI H2 CGGATCCCAAGCTTC (SEQ ID N° 90)

Transformation. Plasmids were transformed in CaCI ₂ competent cells using the simplified procedure of Hanahan (16) or via electroporation as described above.

Calculation methods Introduction

Different experiments with different strains were performed. In total 8 different conditions were tested. There was variation in the genetic background (WT, icIR knock-out, arcA knock-out, and combined icIR-arcA knock-out) and the mode of fermentation (batch, and chemostat). Each experiment was repeated twice.

When running the samples through the BioTrove apparatus, a qPCR curve (fluorescences in function of cycle number) and a melt curve (fluorescences in function of the temperature) is obtained for each sample. Those data were exported from the BioTrove software and further analysed in R. The analysis was divided in two steps: first the qPCR curves were fitted and Ct values were calculated and in the second step the Ct values were converted to expression data.

Calculating the qPCR curves

The raw qPCR curve data were extracted from the BioTrove software and imported in R (1 ). The curves were fitted to a 5 parameter sigmoidal model, with the R package qPCR (25, 34). The maximum of the second derivative of those curves was used as Ct value. No normalisation was applied to the data prior to the curve fitting. However, outliers were removed. The detection of the outliers was done using the following procedure:

^■ Fit the model to the data.

^■ Calculate the residuals (defined as the measured fluorescences minus the model- calculated ones). ^■ Assuming the residuals are normally distributed, calculate the mean and standard deviation of the residuals.

^■ Using this mean and standard deviation, the 95 % interval is calculated.

^■ All data-points for which the residuals fall out of this 95 % interval are considered as outliers.

^■ The curve is refitted without the outliers.

^■ This is repeated until no outliers are detected anymore. Using this procedure, the data do not have to be normalised prior to fitting, neither must the first data-points be removed.

Many curves have to be fitted (1800 genes for one experiment). Therefore, it is undoable to manually check each curve and automated methods have to be applied to reject bad curves. For this different parameters are extracted from the curves: the cycle number value at which the maximum of the first derivative occurs (D1 ), the cycle number value at which the maximum of the second derivative occurs (D2), the minimal fluorescence (Fmin), and the maximal fluorescence (Fmax). Combining the values of those parameters, the validity of the curve and the extent of expression is assessed. How this is done is explained in the next section.

Filtering the data

For some gene-experiment combinations, no amplification is detected. This can be due to a variety of reasons:

^■ Expression is too low and 32 cycles (the number of cycles for all BioTrove arrays was set to 32) is not enough to detect the expression. In this case, the real Ct cannot be determined and is somewhere between 32 and infinity.

^■ No expression. In this case, the real Ct is infinite.

■ Technical failures: primers not suitable, wrong loading (it is very difficult to

uniformly load the BioTrove arrays, especially the holes at the sides of the array are frequently empty), etc. In this case the real Ct can vary between 0 and infinity.

Some genes are genuinely not expressed and setting their Ct value to something else than infinity is not correct. For genes that are expressed, but for which the expression value, due to technical failures or limitations, are not known, setting the Ct value to infinity is not correct. Furthermore, using arbitrary values that are outside the range of expression complicates the calculation routines and visualisation routines. Therefore it was opted to remove the gene-experiment combinations for which no correct expression data was detected. An obvious case of gene-experiment pairs for which no expression is detected, are those for which no curve could be fitted to the qPCR data. Less obvious cases are detailed below.

Typically for expressed genes, is that the fluorescence values cover a certain range. Data points for which this range was not high enough, were discarded, as they pointed to very poorly fitted curves and generally bad data. The minimal fluorescence range was set to 400 (thus Fmax - Fmin > 400).

In a good amplification curve, the first (D1 ) and second (D2) derivative are quite close to each other (see the documentation of the SOD function in the qpcR package(25)). Therefore, all data-points for which the difference between D1 and D2 is larger than an arbitrary value (7 was used) were discarded.

For each primer-pair, a qPCR experiment was performed without adding DNA. Only water was added. Normally no expression should be observed in those samples. However, amplification is detected in water for some primer-pairs. Genes for which the Ct value (as mentioned before, D2 was used) is more than the Ct value of water minus 5, are discarded, as it cannot be excluded that the fluorescence comes from the amplification of the primers and not the added DNA.

Normalising and calculating the contrasts

Prior to calculating the expression differences, the Ct values have to be normalised. As so many genes were measured (1800), quantile normalisation could be used (33). The 1800 genes measured, were divided over 3 types of arrays, each containing 600 genes. Quantile normalisation was done for each type of array separately. A table was constructed where the rows represent the different genes and the columns the different experiments (T1 , see Equations 1 ). Each column was sorted independently (T2) and the original position of the elements was saved. The values in this new table were replaced with the mean value over the different rows (T3). And finally this table was transformed so that the positions of the values corresponded again to the original positions (T4).

2 4 2 4 3 3 3 3

6 8 6 8 6 6 6 6

4 12 4 12 9 9 9 9 Equations 1 : Example of quantile normalisation

Differential expressions were calculated with the normalised data. This was done with the R package limma, which uses a Bayesian approach to calculate the statistical relevances of the differences (31 , 32). Limma was adapted to be able to cope with missing data: the original limma package discards all expression values from a gene over the different experiments, when one value in one experiment is not available. This hampers the analysis when one has many different conditions, as for each gene for which one of the experimental conditions produces no expression values, a different contrast matrix has to be generated omitting that experimental condition. Therefore the function for fitting the contrasts was adapted to drop data-points with missing data.

Differential expressions were calculated between Ct values and the mean Ct value for a certain gene. Thus, the higher the value, the lower the expression. For each gene, plots were generated showing those differences. However, in those plots, the Ct values were inversed, so that the higher the value, the higher the expression.

Example 1 : Effect of arcA and icIR gene deletions on the gene expression of the colanic acid biosynthesis

Figures 1 and 2 show the expression pattern of genes involved in colanic acid biosynthesis (35). Single arcA or icIR knock out mutations did not affect the expression of the operon in comparison of the wild type strain in batch and chemostat conditions. The double mutant strain, AarcAAiclR, however upregulates the genes of the colanic acid operon 6 to 8 times in comparison to the wild type and the single mutant strains in both chemostat and batch conditions. Both regulators have thus a surprisingly cooperative effect on the expression of this operon which is independent from the culturing condition that is applied. Looking at the regulatory network of this operon, no direct link could be found between both ArcA and IcIR and the transcription factor that controls the operon, RcsA (Figure 5). Only ArcA is connected with RcsA via 3 other transcription factors, which are all upregulated as well. However the AarcA single gene deletion mutant strain did not affect the transcription of the operon. Example 2: Effect of arcA and icIR gene deletions on the gene expression of the GDP-fucose biosynthesis genes

Figures 4 and 6 show the relationship of the colanic acid operon with GDP-fucose biosynthesis. In Figure 6 the upregulation of GDP-fucose biosynthesis specific genes is shown. These mutations thus enhance the biosynthesis of GDP-fucose, which is a precursor for fucosylated oligosaccharides such as fucosyllactose, fucosyllactoNbiose and lewis X oligosaccharide or fucosylated proteins. These sugars and proteins, as already indicated above, have applications in therapeutics as nutraceutical, as components in human mother milk in which they have anti-inflammatory and prebiotic effects (5, 8, 27).

Example 3: Enhancement of GDP-fucose and fucosylated oligosaccharide biosynthesis

The mutations AarcAAiclR applied in combination with other mutations enhance the production of fucosylated compounds. A first, 'other' genetic modification that enhances said production is the deletion of wcaJ from the colanic operon, stopping the initiation of the colanic acid biosynthesis and thus the accumulation of GDP-fucose. Further, a fucosyltransferase has to be introduced to link fucose with different acceptor molecules such as lactose. The metabolism is then engineered further to accumulate the precursor of the GDP-fucose biosynthetic pathway. These modifications are shown in Figure 7. Additional to wcaJ , the colanic acid operon genes that do not code for GDP-fucose biosynthesis reactions are knocked out, such as gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaK, wcaL and/or, wcaM. For the production of fucosyllactose, lacZ coding for β-galactosidase, is knocked out to avoid lactose degradation and the expression of lacY, coding for a lactose permease, is enhanced by means of a strong constitutive promoter.

Example 4: Enhancement of GDP-fucose and fucosylated oligosaccharide production via a split metabolism with sucrose as a substrate

To accumulate the GDP-fucose precursor fructose and fructose-6-phosphate, a sucrose phosphorylase or invertase is introduced. Because fructose-6-phosphate is easily degraded in the glycolysis, the glycolysis is interrupted in order to steer all fructose-6- phosphate in the direction of GDP-fucose. The genes pgi, pfkA and pfkB are thus knocked out, coding for glucose-6-phosphate isomerase and phosphofructokinase A and B. Finally a fucosyltransferase is introduced to link fucose to an acceptor molecule.

The growth rate of the wild type strain is somewhat affected when grown on sucrose after introduction of a sucrose phosphorylase (BaSP) (plasmid with sequence SEQ ID N° 2) (Table 2), however the introduction of pgi mutations and pfkA and pfkB double mutations led to significant reduction of growth rate, the latter was extremely low (0.02 h ^"1 ). The combination of all mutations {Apgi and ApfkA and ApfkB) led to the lowest growth rate, however, the growth rate on both sucrose and glucose was surprisingly similar to that of the pgi single mutant.

Table 2: specific growth rates of the glycolysis knock out strains on a minimal medium with glucose and sucrose

SEQ ID N° 2: Plasmid sequence with sucrose phosphorylase BaSP

AATTCGGAGGAAACAAAGATGGGGGGTTCTCATCATCATCATCATCATGGTATGGCTAGC ATGAAAAACAAGGTGCAGCTCATCACTTACGCCGACCGCCTTGGCGACGGCACCATCAAG TCGATGACCGACATTCTGCGCACCCGCTTCGACGGCGTGTACGACGGCGTTCACATCCTG CCGTTCTTCACCCCGTTCGACGGCGCCGACGCAGGCTTCGACCCGATCGACCACACCAAG GTCGACGAACGTCTCGGCAGCTGGGACGACGTCGCCGAACTCTCCAAGACCCACAACATC ATGGTCGACGCCATCGTCAACCACATGAGTTGGGAATCCAAGCAGTTCCAGGACGTGCTG GCCAAGGGCGAGGAGTCCGAATACTATCCGATGTTCCTCACCATGAGCTCCGTGTTCCCG AACGGCGCCACCGAAGAGGACCTGGCCGGCATCTACCGTCCGCGTCCGGGCCTGCCGTTC ACCCACTACAAGTTCGCCGGCAAGACCCGCCTCGTGTGGGTCAGCTTCACCCCGCAGCAG GTGGACATCGACACCGATTCCGACAAGGGTTGGGAATACCTCATGTCGATTTTCGACCAG ATGGCCGCCTCTCACGTCAGCTACATCCGCCTCGACGCCGTCGGCTATGGCGCCAAGGAA GCCGGCACCAGCTGCTTCATGACCCCGAAGACCTTCAAGCTGATCTCCCGTCTGCGTGAG GAAGGCGTCAAGCGCGGTCTGGAAATCCTCATCGAAGTGCACTCCTACTACAAGAAGCAG GTCGAAATCGCATCCAAGGTGGACCGCGTCTACGACTTCGCCCTGCCTCCGCTGCTGCTG CACGCGCTGAGCACCGGCCACGTCGAGCCCGTCGCCCACTGGACCGACATACGCCCGAAC AACGCCGTCACCGTGCTCGATACGCACGACGGCATCGGCGTGATCGACATCGGCTCCGAC CAGCTCGACCGCTCGCTCAAGGGTCTCGTGCCGGATGAGGACGTGGACAACCTCGTCAAC ACCATCCACGCCAACACCCACGGCGAATCCCAGGCAGCCACTGGCGCCGCCGCATCCAAT CTCGACCTCTACCAGGTCAACAGCACCTACTATTCGGCGCTCGGGTGCAACGACCAGCAC TACATCGCCGCCCGCGCGGTGCAGTTCTTCCTGCCGGGCGTGCCGCAAGTCTACTACGTC GGCGCGCTCGCCGGCAAGAACGACATGGAGCTGCTGCGTAAGACGAATAACGGCCGCGAC ATCAATCGCCATTACTACTCCACCGCGGAAATCGACGAGAACCTCAAGCGTCCGGTCGTC AAGGCCCTGAACGCGCTCGCCAAGTTCCGCAACGAGCTCGACGCGTTCGACGGCACGTTC TCGTACACCACCGATGACGACACGTCCATCAGCTTCACCTGGCGCGGCGAAACCAGCCAG GCCACGCTGACGTTCGAGCCGAAGCGCGGTCTCGGTGTGGACAACGCTACGCCGGTCGCC ATGTTGGAATGGGAGGATTCCGCGGGAGACCACCGTTCGGATGATCTGATCGCCAATCCG CCTGTCGTCGCCTGACTGCAGGTCGACCATATGGGAGAGCTCCCAACGCGTTGGATGCAG GCATGCAAGCTTGGCTGTTTTGGCGGATGAGAGAAGATTTTCAGCCTGATACAGATTAAA TCAGAACGCAGAAGCGGTCTGATAAAACAGAATTTGCCTGGCGGCAGTAGCGCGGTGGTC CCACCTGACCCCATGCCGAACTCAGAAGTGAAACGCCGTAGCGCCGATGGTAGTGTGGGG TCTCCCCATGCGAGAGTAGGGAACTGCCAGGCATCAAATAAAACGAAAGGCTCAGTCGAA AGACTGGGCCTTTCGTTTTATCTGTTGTTTGTCGGTGAACGCTCTCCTGAGTAGGACAAA TCCGCCGGGAGCGGATTTGAACGTTGCGAAGCAACGGCCCGGAGGGTGGCGGGCAGGACG CCCGCCATAAACTGCCAGGCATCAAATTAAGCAGAAGGCCATCCTGACGGATGGCCTTTT TGCGTTTCTACAAACTCTTTTTGTTTATTTTTCTAAATACATTCAAATATGTATCCGCTC ATGAGACAATAACCCTGATAAATGCTTCAATAATATTGAAAAAGGAAGAGTATGAGTATT CAACATTTCCGTGTCGCCCTTATTCCCTTTTTTGCGGCATTTTGCCTTCCTGTTTTTGCT CACCCAGAAACGCTGGTGAAAGTAAAAGATGCTGAAGATCAGTTGGGTGCACGAGTGGGT TACATCGAACTGGATCTCAACAGCGGTAAGATCCTTGAGAGTTTTCGCCCCGAAGAACGT TTTCCAATGATGAGCACTTTTAAAGTTCTGCTATGTGGCGCGGTATTATCCCGTGTTGAC GCCGGGCAAGAGCAACTCGGTCGCCGCATACACTATTCTCAGAATGACTTGGTTGAGTAC TCACCAGTCACAGAAAAGCATCTTACGGATGGCATGACAGTAAGAGAATTATGCAGTGCT GCCATAACCATGAGTGATAACACTGCGGCCAACTTACTTCTGACAACGATCGGAGGACCG AAGGAGCTAACCGCTTTTTTGCACAACATGGGGGATCATGTAACTCGCCTTGATCGTTGG GAACCGGAGCTGAATGAAGCCATACCAAACGACGAGCGTGACACCACGATGCCTACAGCA ATGGCAACAACGTTGCGCAAACTATTAACTGGCGAACTACTTACTCTAGCTTCCCGGCAA CAATTAATAGACTGGATGGAGGCGGATAAAGTTGCAGGACCACTTCTGCGCTCGGCCCTT CCGGCTGGCTGGTTTATTGCTGATAAATCTGGAGCCGGTGAGCGTGGGTCTCGCGGTATC ATTGCAGCACTGGGGCCAGATGGTAAGCCCTCCCGTATCGTAGTTATCTACACGACGGGG AGTCAGGCAACTATGGATGAACGAAATAGACAGATCGCTGAGATAGGTGCCTCACTGATT AAGCATTGGTAACTGTCAGACCAAGTTTACTCATATATACTTTAGATTGATTTAAAACTT CATTTTTAATTTAAAAGGATCTAGGTGAAGATCCTTTTTGATAATCTCATGACCAAAATC CCTTAACGTGAGTTTTCGTTCCACTGAGCGTCAGACCCCGTAGAAAAGATCAAAGGATCT TCTTGAGATCCTTTTTTTCTGCGCGTAATCTGCTGCTTGCAAACAAAAAAACCACCGCTA CCAGCGGTGGTTTGTTTGCCGGATCAAGAGCTACCAACTCTTTTTCCGAAGGTAACTGGC TTCAGCAGAGCGCAGATACCAAATACTGTCCTTCTAGTGTAGCCGTAGTTAGGCCACCAC

TTCAAGAACTCTGTAGCACCGCCTACATACCTCGCTCTGCTAATCCTGTTACCAGTG GCT GCTGCCAGTGGCGATAAGTCGTGTCTTACCGGGTTGGACTCAAGACGATAGTTACCGGAT AAGGCGCAGCGGTCGGGCTGAACGGGGGGTTCGTGCACACAGCCCAGCTTGGAGCGAACG ACCTACACCGAACTGAGATACCTACAGCGTGAGCTATGAGAAAGCGCCACGCTTCCCGAA GGGAGAAAGGCGGACAGGTATCCGGTAAGCGGCAGGGTCGGAACAGGAGAGCGCACGAGG GAGCTTCCAGGGGGAAACGCCTGGTATCTTTATAGTCCTGTCGGGTTTCGCCACCTCTGA CTTGAGCGTCGATTTTTGTGATGCTCGTCAGGGGGGCGGAGCCTATGGAAAAACGCCAGC AACGCGGCCTTTTTACGGTTCCTGGCCTTTTGCTGGCCTTTTGCTCACATGTTCTTTCCT GCGTTATCCCCTGATTCTGTGGATAACCGTATTACCGCCTTTGAGTGAGCTGATACCGCT CGCCGCAGCCGAACGACCGAGCGCAGCGAGTCAGTGAGCGAGGAAGCGGAAGAGCGCCTG ATGCGGTATTTTCTCCTTACGCATCTGTGCGGTATTTCACACCGCATATGGTGCACTCTC AGTACAATCTGCTCTGATGCCGCATAGTTAAGCCAGTATACACTCCGCTATCGCTACGTG ACTGGGTCATGGCTGCGCCCCGACACCCGCCAACACCCGCTGACGCGCCCTGACGGGCTT GTCTGCTCCCGGCATCCGCTTACAGACAAGCTGTGACCGTCTCCGGGAGAGCTCGATATC CCGGGCGGCCGCTTCATTTATAAATTTCTTGACATTTTGGAATAGATGTGATATAATGTG TACATATCCATGGCGGCCGCTCTAGAAGAAGCTTGGGATCCGTCGACCTCG

The flux redirections and mutations for GDP-fucose and fucosyllated oligosaccharide biosynthesis in a split metabolism are shown in Figure 8, both for a strain expressing a heterologous invertase and sucrose phosphorylase. Additional to wcaJ, the colanic acid operon genes that do not code for GDP-fucose biosynthesis reactions are knocked out, such as gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaK, wcaL and/or, wcaM. For the production of fucosyllactose, lacZ, coding for β-galactosidase, is knocked out to avoid lactose degradation and the expression of lacY, coding for a lactose permease, is enhanced by means of a strong constitutive promoter.

Example 5: Enhancement of GDP-fucose and fucosylated oligosaccharide production via a split metabolism with glucose as substrate

When the genes pgi, pfkA, and pfkB are knocked out, carbon, taken up as glucose can only be metabolised via the pentose phosphate pathway. Due to the biochemical properties of this pathway, fructose-6-phosphate is formed (Figures 9 and 10). To form biomass glyceraldehyde-3-phosphate has to be formed, which is formed by the transketolase reactions coded by tktA and tktB in E. coli. This Glyceraldehyde-3- phosphate is formed together with fructose-6-phosphate from xylulose-5-phosphate and erythrose-5-phosphate. The latter is in turn formed together with fructose-6-phosphate from glyceraldehyde-3-phosphate and sedoheptulose-7-phosphate via transaldolase reactions coded by talA and talB. To balance all of these reactions together the flux has to be distributed between xylulose-5-phosphate and ribose-5-phosphate, as such that from 1 mole glucose, 2/3 mole of xylulose-5-phosphate and 1/3 mole ribose-5-phosphate is formed. To drive these equilibrium reactions, fructose-6-phosphate is pulled out of the pentose phosphate pathway by the GDP-fucose and fucosyllacted oligosaccharide biosynthesis pathway. Additional to wcaJ, the colanic acid operon genes that do not code for GDP-fucose biosynthesis reactions are knocked out, such as gmm, wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaK, wcaL and/or, wcaM. For the production of fucosyllactose, lacZ coding for β-galactosidase, is knocked out to avoid lactose degradation and the expression of lacY, coding for a lactose permease, is enhanced by means of a strong constitutive promoter.

Example 6: Fermentative 2-fucosyllactose production with a fucosyltransferase originating from Helicobacter pylori

The mutant strain in which the genes lacZ, glgC, agp, pfkA, pfkB, pgi, arcA, icIR, wcaJ are knocked out and lacY was expressed via constitutive expression to ensure expression under all culturing conditions, was transformed further with a fucosyltransferase originating from Helicobacter pylori and a sucrose phosphorylase originating from Bifidobacterium adolescentis, which were also constitutively expressed. The constitutive promoters originate from the promoter library described by De Mey et al. 2007. This strain was cultured in a medium as described in the materials and methods, however with 30 g/l of sucrose and 50 g/l of lactose. This resulted in the formation of 2-fucosyllactose as shown in Figure 13 and 14.

Example 7: Fermentative fucosyllactose production with a fucosyltransferase originating from Dictyostellium discoideum

The mutant strain in which the genes lacZ, glgC, agp, pfkA, pfkB, pgi, arcA, icIR, wcaJ are knocked out and lacY was expressed via constitutive expression to ensure expression under all culturing conditions, was transformed further with a fucosyltransferase originating from Dictyostellium discoideum and a sucrose phosphorylase originating from Bifidobacterium adolescentis, which were also expressed constitutively. The constitutive promoters originate from the promoter library described by De Mey et al. 2007. This strain was cultured in a medium as described in the materials and methods, however with 30 g/l of sucrose and 50 g/l of lactose. This resulted in the formation of 2-fucosyllactose as shown in Figure 13 and 14. Example 8: Enhancement of GDP-mannose and mannosylated oligosaccharide production via a split metabolism with sucrose as substrate

To accumulate the GDP-mannose precursors fructose and fructose-6-phosphate, a sucrose phosphorylase or invertase is introduced. Because fructose-6-phosphate is easily degraded in the glycolysis, the glycolysis is interrupted in order to steer all fructose-6- phosphate in the direction of GDP-fucose. The genes pgi, pfkA and pfkB are thus knocked out, coding for glucose-6-phosphate isomerase and phosphofructokinase A and B. Finally a mannosyltransferase is introduced to link mannose to an acceptor molecule. To avoid GDP-mannose degradation the genes gmm and gmd have to be knocked out in the colanic acid operon. In addition, the genes that do not code for GDP-mannose biosynthesis reactions are knocked out, such as wcaA, wcaB, wcaC, wcaD, wcaE, wcaF, weal, wcaJ, wcaK, wcaL and/or, wcaM.

Example 9: Upregulation of acid resistance related genes

Similar to the colanic acid operon upregulation, acid resistance related genes are also upregulated in a AarcAAiclR double mutant strain in comparison to the wild type strain and the single mutant strains. These genes make a strain more resistant to low pH, which is beneficial for the production of acids (4) or the production of glucosamine (12) which is not stable at neutral and high pH. Figure 12 presents the gene expression pattern of these acid resistance related genes and indicates up to 8 fold expression increase in the double mutant strain.

Example 10: Fed batch production of 2-fucosyllactose

A mutant strain was constructed via the genetic engineering methodologies described above with the following genotype:

AlacZYA::P22-lacYAglgCAagpApgiApfkA-P22- baSPApfkBAarcAAiclR::slAwcaJAIonAadhE-P14-frk + pCXP14-FT_H. pylori (a vector with sequence SEQ ID N°1 ). The promoter P22 and P14 originate from the promoter library constructed by De Mey et al (1 1 ) and was cloned similar to the methodology described by Aerts et al (2). "::sl" marks a scarless gene deletion, thus without a FRT site that remains in the chromosome.

This strain was cultured in a bioreactor as described above in materials and methods, in the mineral medium with 30 g/l of sucrose and 50 g/l of lactose. After the batch phase the bioreactor was fed with 500 g/l of sucrose, 50 g/l lactose and 1 g/l of magnesium sulphate heptahydrate. This led to the accumulation of 27.5 g/l of fucosyllactose in the supernatant. SEQID N° 1 : pCXP14-FT H. pylori

CGCGTTGGATGCAGGCATGCAAGCTTGGCTGTTTTGGCGGATGAGAGAAGATTTTCAGCC TGATACAGATTAAATCAGAACGCAGAAGCGGTCTGATAAAACAGAATTTGCCTGGCGGCA GTAGCGCGGTGGTCCCACCTGACCCCATGCCGAACTCAGAAGTGAAACGCCGTAGCGCCG ATGGTAGTGTGGGGTCTCCCCATGCGAGAGTAGGGAACTGCCAGGCATCAAATAAAACGA AAGGCTCAGTCGAAAGACTGGGCCTTTCGTTTTATCTGTTGTTTGTCGGTGAACGCTCTC CTGAGTAGGACAAATCCGCCGGGAGCGGATTTGAACGTTGCGAAGCAACGGCCCGGAGGG TGGCGGGCAGGACGCCCGCCATAAACTGCCAGGCATCAAATTAAGCAGAAGGCCATCCTG ACGGATGGCCTTTTTGCGTTTCTACAAACTCTTTTTGTTTATTTTTCTAAATACATTCAA ATATGTATCCGCTCATGAGACAATAACCCTGATAAATGCTTCAATAATATTGAAAAAGGA AGAGTATGAGTATTCAACATTTCCGTGTCGCCCTTATTCCCTTTTTTGCGGCATTTTGCC TTCCTGTTTTTGCTCACCCAGAAACGCTGGTGAAAGTAAAAGATGCTGAAGATCAGTTGG GTGCACGAGTGGGTTACATCGAACTGGATCTCAACAGCGGTAAGATCCTTGAGAGTTTTC GCCCCGAAGAACGTTTTCCAATGATGAGCACTTTTAAAGTTCTGCTATGTGGCGCGGTAT TATCCCGTGTTGACGCCGGGCAAGAGCAACTCGGTCGCCGCATACACTATTCTCAGAATG ACTTGGTTGAGTACTCACCAGTCACAGAAAAGCATCTTACGGATGGCATGACAGTAAGAG AATTATGCAGTGCTGCCATAACCATGAGTGATAACACTGCGGCCAACTTACTTCTGACAA CGATCGGAGGACCGAAGGAGCTAACCGCTTTTTTGCACAACATGGGGGATCATGTAACTC GCCTTGATCGTTGGGAACCGGAGCTGAATGAAGCCATACCAAACGACGAGCGTGACACCA CGATGCCTACAGCAATGGCAACAACGTTGCGCAAACTATTAACTGGCGAACTACTTACTC TAGCTTCCCGGCAACAATTAATAGACTGGATGGAGGCGGATAAAGTTGCAGGACCACTTC TGCGCTCGGCCCTTCCGGCTGGCTGGTTTATTGCTGATAAATCTGGAGCCGGTGAGCGTG GGTCTCGCGGTATCATTGCAGCACTGGGGCCAGATGGTAAGCCCTCCCGTATCGTAGTTA TCTACACGACGGGGAGTCAGGCAACTATGGATGAACGAAATAGACAGATCGCTGAGATAG GTGCCTCACTGATTAAGCATTGGTAACTGTCAGACCAAGTTTACTCATATATACTTTAGA TTGATTTAAAACTTCATTTTTAATTTAAAAGGATCTAGGTGAAGATCCTTTTTGATAATC TCATGACCAAAATCCCTTAACGTGAGTTTTCGTTCCACTGAGCGTCAGACCCCGTAGAAA AGATCAAAGGATCTTCTTGAGATCCTTTTTTTCTGCGCGTAATCTGCTGCTTGCAAACAA AAAAACCACCGCTACCAGCGGTGGTTTGTTTGCCGGATCAAGAGCTACCAACTCTTTTTC CGAAGGTAACTGGCTTCAGCAGAGCGCAGATACCAAATACTGTCCTTCTAGTGTAGCCGT AGTTAGGCCACCACTTCAAGAACTCTGTAGCACCGCCTACATACCTCGCTCTGCTAATCC TGTTACCAGTGGCTGCTGCCAGTGGCGATAAGTCGTGTCTTACCGGGTTGGACTCAAGAC GATAGTTACCGGATAAGGCGCAGCGGTCGGGCTGAACGGGGGGTTCGTGCACACAGCCCA GCTTGGAGCGAACGACCTACACCGAACTGAGATACCTACAGCGTGAGCTATGAGAAAGCG CCACGCTTCCCGAAGGGAGAAAGGCGGACAGGTATCCGGTAAGCGGCAGGGTCGGAACAG GAGAGCGCACGAGGGAGCTTCCAGGGGGAAACGCCTGGTATCTTTATAGTCCTGTCGGGT TTCGCCACCTCTGACTTGAGCGTCGATTTTTGTGATGCTCGTCAGGGGGGCGGAGCCTAT GGAAAAACGCCAGCAACGCGGCCTTTTTACGGTTCCTGGCCTTTTGCTGGCCTTTTGCTC ACATGTTCTTTCCTGCGTTATCCCCTGATTCTGTGGATAACCGTATTACCGCCTTTGAGT GAGCTGATACCGCTCGCCGCAGCCGAACGACCGAGCGCAGCGAGTCAGTGAGCGAGGAAG CGGAAGAGCGCCTGATGCGGTATTTTCTCCTTACGCATCTGTGCGGTATTTCACACCGCA TATGGTGCACTCTCAGTACAATCTGCTCTGATGCCGCATAGTTAAGCCAGTATACACTCC GCTATCGCTACGTGACTGGGTCATGGCTGCGCCCCGACACCCGCCAACACCCGCTGACGC GCCCTGACGGGCTTGTCTGCTCCCGGCATCCGCTTACAGACAAGCTGTGACCGTCTCCGG GAGAGCTCGATATCCCGGGCGGCCGCCTTCATTCTATAAGTTTCTTGACATCTTGGCCGG CATATGGTATAATAGGGAAATTTCCATGGCGGCCGCTCTAGAAGAAGCTTGGGATCCGTC GACCTCGAATTCGGAGGAAACAAAGATGGCCTTTAAAGTTGTTCAGATTTGTGGTGGTCT GGGCAAT C AGAT GT T T C AGT AT GC AT T T GCAAAAAGCCT GC AGAAAC AT AGC AAT AC ACC GGTTCTGCTGGATATTACCAGCTTTGATTGGAGCAATCGTAAAATGCAGCTGGAACTGTT TCCGATTGATCTGCCGTATGCAAGCGAAAAAGAAATTGCAATTGCCAAAATGCAGCATCT GCCGAAACTGGTTCGTAATGTTCTGAAATGCATGGGTTTTGATCGTGTGAGCCAAGAAAT CGTGTTTGAATATGAACCGAAACTGCTGAAAACCAGCCGTCTGACCTATTTTTATGGCTA TTTTCAGGATCCGCGTTATTTTGATGCAATTAGTCCGCTGATCAAACAGACCTTTACCCT GCCTCCGCCTCCGGAAAATGGTAATAACAAAAAAAAAGAAGAAGAGTATCATCGTAAACT GGCACTGATTCTGGCAGCAAAAAATAGCGTGTTTGTGCATATTCGTCGCGGTGATTATGT TGGTATTGGTTGTCAGCTGGGCATCGATTATCAGAAAAAAGCACTGGAATACATGGCAAA ACGTGTTCCGAATATGGAACTGTTTGTGTTTTGCGAGGACCTGGAATTTACCCAGAATCT GGATCTGGGCTATCCGTTTATGGATATGACCACCCGTGATAAAGAGGAAGAGGCATATTG GGATATGCTGCTGATGCAGAGCTGTAAACATGGTATTATTGCCAACAGCACCTATAGTTG GTGGGCAGCATATCTGATTAATAACCCGGAAAAAATCATTATTGGTCCGAAACATTGGCT GTTTGGCCATGAAAACATCCTGTGTAAAGAATGGGTGAAAATCGAAAGCCACTTTGAAGT GAAAAGCCAGAAATATAATGCCTAATAAGAGCTCCCAA

Example 11 : Fed batch production of 2-fucosyllactose with a hybrid colanic acid promoter

A hybrid colanic acid promoter was constructed based on the genome information and the sequences from the promoter library described by De Mey et al (1 1 ).

AlacZYA::P22-lacYAglgCAagpApgiApfkA::P22-BaSPApfkB AarcAAiclR:sl AwcaJ ΔΙοη AadhE-P14-frk AldhA::P14- FT_H. pylori ApromCA:P14

This strain was cultured in a bioreactor as described above in materials and methods, in the mineral medium with 30 g/l of sucrose and 20 g/l of lactose. After the batch phase the bioreactor was fed with 500 g/l of sucrose, 20 g/l lactose and 1 g/l of magnesium sulphate heptahydrate. This led to the accumulation of 26 g/l of fucosyllactose in the supernatant with nearly stoichiometric conversion of lactose. Increasing the lactose feed concentrations leads further to increased final fucosyllactose titers and stoichiometric lactose conversion.

REFERENCES:

1 . 2006. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. R-Development Core Team, Vienna, Austria.

2. Aerts, D., T. Verhaeghe, M. De Mey, T. Desmet, and W. Soetaert. 2010. A

constitutive expression system for high throughput screening. Engineering in Life

Sciences 10:DOI: 10.1002/elsc.201000065.

3. Alper, H., C. Fischer, E. Nevoigt, and G. Stephanopoulos. 2005. Tuning genetic control through promoter engineering. Proceedings of the national academy of sciences of the United States of America 102:12678-12683.

4. Beauprez, J. 2010. Metabolic modelling and engineering of Escherichia coli for succinate production. PhD. Ghent University, Ghent.

5. Bode, L. 2006. Recent Advances on Structure, Metabolism, and Function of

Human Milk Oligosaccharides. The Journal of Nutrition 136:2127-2130.

6. Canton, B., A. Labno, and D. Endy. 2008. Refinement and standardization of

synthetic biological parts and devices. Nat Biotech 26:787-793.

7. Cavallaro, G., L. Maniscalco, P. Caliceti, S. Salmaso, A. Semenzato, and G.

Giammona. 2004. Glycosilated Macromolecular Conjugates of Antiviral Drugs with a Polyaspartamide. Journal of Drug Targeting 12:593-605.

8. Coppa, G. V., L. Zampini, T. Galeazzi, and O. Gabrielli. 2006. Prebiotics in human milk: a review. Digestive and Liver Disease 38:S291 -S294.

9. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proceedings of the national academy of sciences of the United States of America 97:6640-6645.

10. De Mey, M., G. J. Lequeux, J. J. Beauprez, J. Maertens, E. Van Horen, W. K.

Soetaert, P. A. Vanrolleghem, and E. J. Vandamme. 2007. Comparison of different strategies to reduce acetate formation in Escherichia coli. Biotechnology Progress.

1 1 . De Mey, M., J. Maertens, G. J. Lequeux, W. K. Soetaert, and E. J. Vandamme.

2007. Construction and model-based analysis of a promoter library for £. coli: an indispensable tool for metabolic engineering. BMC Biotechnology 7:34-48.

12. Deng, M.-D., D. K. Severson, A. D. Grund, S. L. Wassink, R. P. Burlingame, A.

Berry, J. A. Running, C. A. Kunesh, L. Song, T. A. Jerrell, and R. A. Rosson. 2005. Metabolic engineering of Escherichia coli for industrial production of glucosamine and N-acetylglucosamine. Metabolic Engineering 7:201 -214.

13. Ebel, W., and J. E. Trempy. 1999. Escherichia coli RcsA, a Positive Activator of Colanic Acid Capsular Polysaccharide Synthesis, Functions To Activate Its Own

Expression. Journal of bacteriology 181 :577-584.

14. Foster, J. W. 2004. Escherichia coli acid resistance: Tales of an amateur

acidophile. Nature Reviews Microbiology 2:898-907. Hammer, K., I. Mijakovic, and P. R. Jensen. 2006. Synthetic promoter libraries - tuning of gene expression. TRENDS in Biotechnology 24:53-55.

Hanahan, D., J. Jessee, and F. R. Bloom. 1991 . Plasmid transformation of Escherichia coii and other bacteria. Methods in Enzymology 204:63-1 13.

Hommais, F., E. Krin, J.-Y. Coppee, C. Lacroix, E. Yeramian, A. Danchin, and P. Bertin. 2004. GadE (YhiE): a novel activator involved in the response to acid environment in Escherichia coii. Microbiology 150:61 -72.

Jigami, Y. 2008. Yeast Glycobiology and Its Application. Bioscience,

Biotechnology, and Biochemistry 72:637-648.

Keasling, J. D. 1999. Gene-expression tools for the metabolic engineering of bacteria. Trends in Biotechnology 17:452-460.

Lequeux, G. 2008. Metabolic modelling and analysis for the optimization of Escherichia coii as a production host. Ghent University, Ghent.

Lequeux, G., L. Johansson, J. Maertens, P. Vanrolleghem, and G. Liden. 2005. Metabolic flux analysis of C- and P-limited shikimic acid producing E. coii. Journal of Biotechnology 1 18:S121 -S121.

Masuda, N., and G. M. Church. 2003. Regulatory network of acid resistance genes in Escherichia coii. Molecular Microbiology 48:699-712.

Pattyn, F., F. Speleman, A. De Paepe, and J. Vandesompele. 2003. RTPrimerDB: The Real-Time PCR primer and probe database. Nucleic Acids Research 31 :122- 123.

Perrenoud, A., and U. Sauer. 2005. Impact of global transcriptional regulation by ArcA, ArcB, Cra, Crp, Cya, Fnr, and MIc on glucose catabolism in Escherichia coii. Journal of Bacteriology 187:3171 -3179.

Ritz, C, and A.-N. Spiess. 2008. qpcR: an R package for sigmoidal model selection in quantitative real-time polymerase chain reaction analysis.

Bioinformatics 24:1549-1551 .

Salis, H. M., E. A. Mirsky, and C. A. Voigt. 2009. Automated design of synthetic ribosome binding sites to control protein expression. Nat Biotech 27:946-950.

Seed, B., and J. Holgersson. 1999. Fucosyltransferase genes and uses thereof patent US5858752.

Shalel-Levanon, S., K. Y. San, and G. N. Bennett. 2005. Effect of ArcA and FNR on the expression of genes related to the oxygen regulation and glycolysis pathway in Escherichia coii under growth conditions. Biotechnology and

Bioengineering 92:147-159.

Shalel-Levanon, S., K. Y. San, and G. N. Bennett. 2005. Effect of oxygen on the Escherichia coii ArcA and FNR regulation systems and metabolic responses. Biotechnology and Bioengineering 89:556-564.

Smith, S., A. D. Elbein, and Y. T. Pan. 1999. Inhibition of Pathogenic Bacterial Binding by High Mannose Oligosaccharides patent US5939279.

Smyth, G. K. 2005. limma: Linear models for microarray data. , p. 397-420. In R. Gentleman, V. Carey, W. Huber, R. A. Irizarry, and S. Dudoit (ed.), Bioinformatics and Computational Biology Solutions Using R and Bioconductor. Springer.

Smyth, G. K. 2004. Linear Models and Empirical Bayes Methods for Assessing Differential Expression in Microarray Experiments. Statistical Applications in Genetics and Molecular Biology 3:3. Smyth, G. K., and T. Speed. 2003. Normalization of cDNA microarray data.

Methods 31 :265-273.

Spiess, A.-N., C. Feig, and C. Ritz. 2008. Highly accurate sigmoidal fitting of realtime PCR data by introducing a parameter for asymmetry. BMC Bioinformatics 9:221 .

Stevenson, G., K. Andrianopoulos, M. Hobbs, and P. R. Reeves. 1996.

Organization of the Escherichia coii K-12 gene cluster responsible for production of the extracellular polysaccharide colanic acid. Journal of Bacteriology 178:4885- 4893.

Stout, V., A. Torres-Cabassa, M. R. Maurizi, D. Gutnick, and S. Gottesman. 1991.

RcsA, an unstable positive regulator of capsular polysaccharide synthesis. Journal of bacteriology 173:1738-1747.

Sunnarborg, A., D. Klumpp, T. Chung, and D. C. Laporte. 1990. Regulation of the glyoxylate bypass operon: cloning and characterization of IcIR. Journal of

Bacteriology 172:2642-2649.

Waegeman, H. J., J. Beauprez, H. Moens, J. Maertens, M. De Mey, M. R.

Foulquie-Moreno, J. J. Heijnen, D. Charlier, and W. Soetaert. 2010. Effect of icIR and arcA knockouts on biomass formation and metabolic fluxes in Escherichia coii K12 and its implications on understanding the metabolism of Escherichia coii BL21 (DE3). BMC microbiology 1 1 :1 -17.

Waegeman, H. J., J. Beauprez, H. Moens, J. Maertens, M. De Mey, M. R.

Foulquie-Moreno, J. J. Heijnen, D. Charlier, and W. Soetaert. 201 1 . Effect of icIR and arcA knockouts on biomass formation and metabolic fluxes in Escherichia coii K12 and its implications on understanding the metabolism of Escherichia coii BL21 (DE3). BMC microbiology 1 1 :1 -17.

Warnecke, T., and R. T. Gill. 2005. Organic acid toxicity, tolerance, and production in Escherichia coii biorefining applications. Microbial Cell Factories 4.