Skip to main content

Engineering Escherichia coli for increased Und-P availability leads to material improvements in glycan expression technology

Abstract

Background

Bacterial surface glycans are assembled by glycosyltransferases (GTs) that transfer sugar monomers to long-chained lipid carriers. Most bacteria employ the 55-carbon chain undecaprenyl phosphate (Und-P) to scaffold glycan assembly. The amount of Und-P available for glycan synthesis is thought to be limited by the rate of Und-P synthesis and by competition for Und-P between phosphoglycosyl transferases (PGTs) and GTs that prime glycan assembly (which we collectively refer to as PGT/GTs). While decreasing Und-P availability disrupts glycan synthesis and promotes cell death, less is known about the effects of increased Und-P availability.

Results

To determine if cells can maintain higher Und-P levels, we first reduced intracellular competition for Und-P by deleting all known non-essential PGT/GTs in the Gram-negative bacterium Escherichia coli (hereafter called ΔPGT/GT cells). We then increased the rate of Und-P synthesis in ΔPGT/GT cells by overexpressing the Und-P(P) synthase uppS from a plasmid (puppS). Und-P quantitation revealed that ΔPGT/GT/puppS cells can be induced to maintain 3-fold more Und-P than wild type cells. Next, we determined how increasing Und-P availability affects glycan expression. Interestingly, increasing Und-P availability increased endogenous and recombinant glycan expression. In particular, ΔPGT/GT/puppS cells could be induced to express 7-fold more capsule from Streptococcus pneumoniae serotype 4 than traditional E. coli cells used to express recombinant glycans.

Conclusions

We demonstrate that the biotechnology standard bacterium E. coli can be engineered to maintain higher levels of Und-P. The results also strongly suggest that Und-P pathways can be engineered to increase the expression of potentially any Und-P-dependent polymer. Given that many bacterial glycans are central to the production of vaccines, diagnostics, and therapeutics, increasing Und-P availability should be a foremost consideration when designing bacterial glycan expression systems.

Background

Enveloped bacteria express a wide array of polysaccharides at the cell surface that confer morphology, protect against environmental insults, and resist killing by immune systems [1]. Most of these polysaccharides are linked to proteins or lipids and are referred to as glycoconjugates, a diverse class of molecules that include glycoproteins, lipopolysaccharides, capsular polysaccharides, lipoarabinomannans, peptidoglycan (PG), glycosylated teichoic acids, and many other clinically important surface structures [2]. The precursors that form these layers are assembled on an essential lipid carrier known as undecaprenyl phosphate (Und-P), which is also referred to as bactoprenyl phosphate [3]. Und-P is a 55-carbon isoprenoid that is first synthesized as a diphosphate (Und-PP) on the inner face of the cytoplasmic membrane by the undecaprenyl pyrophosphate synthase (UppS) via the methylerythritol phosphate (MEP) pathway [4, 5]. Und-PP is also generated on the outer face of the cytoplasmic membrane when it is released during glycan polymerization. Und-PP is then dephosphorylated by Und-PP pyrophosphatases to Und-P. The integral membrane pyrophosphatase BacA and PAP2 family proteins dephosphorylate Und-PP [6,7,8] and several lines of evidence indicate that this activity occurs on the outer face of the cytoplasmic membrane [8,9,10,11]. Since Und-PP is synthesized on the inner face of the cytoplasmic membrane, enzymes possessing Und-PP pyrophosphatase activity would appear to be required on both sides of the cytoplasmic membrane. However, a cytoplasmic Und-PP phosphatase has yet to be discovered.

Once Und-P is active, phosphoglycosyl transferases (PGTs) catalyze the transfer of phosphosugars from nucleoside diphosphate sugar donors to Und-P to form Und-PP-linked sugar monomers [12], which are progressively built into oligosaccharide building blocks by other GTs. GTs also catalyze the transfer of sugars to Und-P to form Und-P-linked sugar monomers that serve as donors to other glycans [13, 14]. Collectively, we refer to these priming enzymes as PGT/GTs. Since most bacteria express multiple Und-P-dependent glycans, cells usually encode several PGT/GTs. For example, the Gram-negative bacterium Escherichia coli, which is the workhorse of the biotechnology industry, harbors five PGT/GTs [Enzyme, product]: MraY, PG; WecA, O-antigen and enterobacterial common antigen (ECA); WcaJ, colanic acid capsule; ArnC, aminoarabinose modification of phosphate groups on Lipid A; GtrB, glucose modification of O-antigen. We note that most lab strains of E. coli do not produce O-antigen due to an insertion in wbbL (K-12 derivatives) or wbbD (B derivatives) [15, 16]. Of these PGT/GTs, only MraY is essential, so that mutants lacking individual PGT/GTs are readily obtained [17]. Recently, we discovered that deleting wecA increases the free pool of Und-P in E. coli [18]. This finding prompted us to determine if other Und-P pathways could be modified to increase the pool of Und-P.

The present results show that E. coli can be induced to maintain at least three times more Und-P than wild type cells by overexpressing the Und-PP synthase uppS in cells lacking all non-essential PGT/GTs. That Und-P levels can rise to such a degree is notable given that an overabundance of polyprenyl phosphates like Und-P is thought to destabilize phospholipid membranes [19]. Since increasing Und-P levels increases production of Und-PP-linked intermediates [18, 20, 21], we reasoned that similar effects might occur for the production of finished glycans. Indeed, increasing cellular Und-P levels increased both endogenous and heterologous glycan expression in E. coli. In particular, we found that recombinant Streptococcus pneumoniae capsular polysaccharide expression (important for the production of bioconjugate pneumococcal vaccines [22,23,24]) dramatically rises in cells containing more Und-P. Since Und-P is a universal carrier lipid, the results suggest similar increases may occur for potentially any Und-P-dependent polymer. Thus, increasing Und-P availability should be considered a prime driver for systems used to express bacterial glycans. Such processes (termed Glycan Expression Technology) are currently underpowered by low glycan yields [25,26,27,28,29].

Results

The free pool of Und-P increases in a Und-P pathway-minimized strain

Previously, we constructed a mutant lacking WecA in the E. coli strain MG1655 and found these cells to contain more Und-P [18]. Since deleting wecA prevents formation of ECA in this strain background, we sought to determine if similar effects would occur for other Und-P-using pathways. To this end, we systematically disrupted all non-essential Und-P-dependent pathways in E. coli. However, since MG1655 cells do not produce O-antigen due to insertion sequences in wbbL [15], we engineered pathway mutants in MG1655 wbbL + cells (hereafter referred to as wild type [WT] cells) [30]. Pathway mutations were introduced at the point of initiation to prevent the accumulation of Und-PP-linked intermediates, which lower Und-P levels [18]. Thus, we individually deleted wecA (ECA and O-antigen), wcaJ (colanic acid), gtrB (O-antigen glucosylation), and arnC (modification of lipid A with aminoarabinose). Since WecA initiates ECA and O-antigen synthesis in MG1655 derivatives, we specifically prevented ECA expression (but not O-antigen) by deleting the wecB epimerase, which is required to produce the second building block in ECA synthesis (i.e., UDP-N-acetylmannosaminuronate) [31]. Importantly, WecA+WecB cells are not expected to sequester Und-P in Und-PP-linked ECA intermediates (i.e., they will not lower Und-P levels). Similarly, the wbbL::IS5 insertion mutation prevents the formation of the second building block in O-antigen synthesis (i.e., dTDP-L-rhamnose) without sequestering Und-P [32].

To ensure our mutations did not inadvertently lower Und-P levels, we first examined our pathway mutants for changes in morphology (i.e., PG synthesis), which is highly responsive to reductions in Und-P availability [32,33,34]. As expected, pathway mutants produced rod-shaped cells (Fig. S1A) that were similar in size to wild type cells when examined by flow cytometry (Fig. S1B). We note that cells lacking O-antigen (i.e., wbbL::IS5 and ΔwecA) grew slightly larger, confirming its role as a mechanical element within the cell envelope [35]. In short, our cell morphology results indicated that Und-P levels were not reduced in our pathway mutants.

To determine if Und-P levels had increased in our pathway mutants, we measured Und-P levels directly by liquid chromatography – mass spectrometry (LC-MS). We note that wherever Und-P levels are discussed, they only include the free pool of Und-P and not those Und-P molecules linked to glycan intermediates. Intracellular quantitation of Und-P levels revealed that WT cells harbor on average 123,000 molecules of Und-P per cell (Fig. S1C and Table S1), which was consistent with previously measured values [36]. To our surprise, similar levels of Und-P were also obtained for mutants lacking individual pathways and WecA, which is ECA and O-antigen minus. Since multiple Und-P-utilizing pathways were still active in our pathway mutants, and since Und-P pathways compete for a common pool of Und-P [32, 33, 37], the observed results suggested that Und-P was being redirected in our strain backgrounds. Therefore, we engineered an Und-P pathway-minimized strain of E. coli lacking WecA, WcaJ, GtrB, and ArnC, which we refer to as ΔPGT/GT cells. Morphological analysis revealed that ΔPGT/GT cells produce rod-shaped cells that are somewhat larger than WT (Fig. 1A and B), consistent with the absence of O-antigen. Further analysis revealed that ΔPGT/GT cells grow identical to wild type cells in rich media (Fig. 1C). Interestingly, Und-P quantitation revealed that ΔPGT/GT cells harbor on average 301,000 molecules of Und-P per cell, which represents a 145% increase in the free pool of Und-P when compared to WT cells (Fig. 1D and Table S1, compare WT to ΔPGT/GT cells). These results indicate that ~ 175,000 molecules of Und-P are (collectively) employed by WecA, WcaJ, GtrB, and ArnC when cultured under standard laboratory conditions. In summary, our results demonstrate that E. coli can be engineered to express fewer Und-P-utilizing pathways. Our results also demonstrate that E. coli cells can tolerate increases in the free pool of Und-P.

Fig. 1
figure 1

Maximizing Und-P levels in a Und-P pathway-minimized strain of E. coli. (A and E) Micrographs of cells with the indicated genotypes. Cells were grown in TB (panel A) and TB containing 500 µM IPTG (panel E) at 37 °C until the culture reached an OD600 of 0.4 to 0.6. The cells were then photographed by phase contrast microscopy. Bar, 3 μm. (B and F) Flow cytometry data from live cells in panels A and E. Histograms of the forward scatter area from 100,000 cells are shown. The mean cell size is shown in arbitrary units (AU). (C and G) Growth curves for cells with the indicated genotypes cultured at 37 °C in TB (panel C) or TB containing 500 µM IPTG (panel G). Error bars show +/- standard deviation of the means. (D and H) Und-P levels from cells grown in panel A (after 3.5 h) and panel E (after 24 h). Und-P levels were normalized by dividing Und-P measurements by the mean CFU/ml. Absolute Und-P values are detailed in Table S1. Additional Und-P pathway mutants are shown in Fig. S1. Error bars show +/- standard error of the means. Significance was determined by using an unpaired t-test followed by Welch’s correction. *p < 0.05. Morphological data are representative of two independent experiments. Growth and Und-P measurements are representative of two independent experiments performed in triplicate. The E. coli strains shown are MAJ330 (WT), MAJ557 (ΔPGT/GT), MAJ1385 (ΔPGT/GT/vector), and MAJ1386 (ΔPGT/GT/puppS).

Increasing Und-P in a Und-P-pathway minimized strain

At this point, we sought to determine if ΔPGT/GT cells could be induced to maintain even higher levels of Und-P. When E. coli is cultured under aerobic growth, the Und-PP synthase UppS competes with IspB for the isoprenoid precursors IPP (isopentenyl pyrophosphate) and FPP (farnesyl pyrophosphate) [38, 39]. Since Und-PP synthesis is limited by substrate competition, we reasoned that overexpressing uppS would promote Und-P availability in ΔPGT/GT cells, presumably by limiting the flux of isoprenoid precursors to IspB. Therefore, we transformed ΔPGT/GT cells with a plasmid expressing an IPTG-inducible copy of uppS (puppS). As can be seen in Fig. 1, uppS overexpression had no effect on the shape (Fig. 1E and F) or growth (Fig. 1G) of ΔPGT/GT cells. Subsequent Und-P quantitation revealed that ΔPGT/GT/puppS cells could be induced to contain on average 387,000 molecules of Und-P per cell (Fig. 1H and Table S1), which represents a 29% and 215% increase in the free pool of Und-P when compared to ΔPGT/GT and WT cells, respectively. Interestingly, ΔPGT/GT/vector cells contained on average 100,000 molecules of Und-P per cell, far below the expected 300,000 molecules per cell we observed in ΔPGT/GT cells. Although we cannot fully explain this result, we are currently investigating the effect of plasmid suppression on Und-P levels. In any event, our results suggest that UppS activity bottlenecks Und-P availability and that increasing flux through Und-P pathways is an effective way to increase Und-P levels in different strain backgrounds.

Increasing Und-P availability increases endogenous glycan expression

Several studies have shown that uppS overexpression (increases Und-P levels) increases Und-PP-linked intermediate formation in E. coli [18, 20, 21]. However, to our knowledge, no study has clearly linked increased Und-P to the production of finished glycans. Therefore, to determine if increasing Und-P levels increases production of finished glycans, we measured the effect of uppS overexpression on ECA (O14 antigen) surface expression in E. coli wbbL::IS5 (MG1655) cells. We note that wbbL::IS5 cells express two forms of ECA at the cell surface, (1) LPS-linked ECA (due to the loss of O-antigen expression [40]) and (2) phosphoglyceride-linked ECA (a third cyclic form is also expressed in the periplasm) [41]. Interestingly, dot blot analysis revealed that wbbL::IS5/puppS cells produce 75% more signal at the cell surface than wbbL::IS5/vector cells (Fig. 2A and B); cells lacking WecA did not produce an appreciable amount of signal, even when induced to overexpress uppS (Fig. 2A and B). Western blot analysis revealed similar increases in ECA expression, with the strongest effects observed for ECA chains less than 50 kDa (Fig. 2C, compare wbbL::IS5/vector to wbbL::IS5/puppS). Bands corresponding to longer ECA chains (> 50 kDa) were also stronger in cells overexpressing uppS (Fig. 2C). We also observed a non-specific band ~ 30 kDa whose expression varied in our strains (Fig. 2C). We are currently investigating the source of this band. In short, our results indicate that increasing Und-P availability increases endogenous glycan expression in bacteria.

Fig. 2
figure 2

Increasing Und-P levels promotes enterobacterial common antigen surface expression in E. coli. (A) Dot blot showing that increasing Und-P levels by overexpressing uppS promotes surface expression of ECA in E. coli. Cells were grown at 37 °C in LB media containing 500 µM IPTG (0 µM IPTG for baseline [BL]). Cells lacking WecA do not produce ECA (negative control). (B) Densitometry was performed from dot blots to quantitate surface expression of ECA. Signal values are given as a percent of the average wbbL::IS5/vector (wbbL::IS5 is E. coli MG1655) signal. All data are representative of three independent experiments. Error bars show +/- standard error of the means. Significance was determined by using an ordinary one-way ANOVA test with Dunnett’s multiple correction. Asterisks above sample bars denote significance relative to wbbL::IS5/vector. *p < 0.05, **** <0.0001. (C) Western blotting was used to examine the effect of increasing Und-P levels on ECA chain length from cells grown in panel A. *, nonspecific band. MreB, which has a predicted molecular mass of 37 kDa, served as the loading control. The E. coli strains shown are MAJ286 (wbbL::IS5/vector), MAJ981 (ΔwecA/vector), MAJ1354 (wbbL::IS5/puppS), and MAJ1677 (ΔwecA/puppS).

Increasing Und-P availability increases recombinant glycan expression

Since increasing Und-P availability increased endogenous glycan expression (Fig. 2), we next sought to determine if similar effects would occur for E. coli cells expressing a non-native recombinant glycan. For these studies, we examined the effect of increasing Und-P availability in E. coli cells expressing the capsular polysaccharide from Streptococcus pneumoniae serotype 4 (SP4); SP4 was expressed from an IPTG-inducible promoter on plasmid pB4 [29]. To do this, we first examined the effect of increasing Und-P on SP4 expression in E. coli W3110. To our surprise, Western blot analysis revealed that uppS overexpression produced little effect on SP4 expression in W3110 (Fig. S2). However, since N-acetylgalactosamine (GalNAc) availability limits recombinant SP4 expression [29], we overexpressed uppS in E. coli Falcon cells, a derivative of W3110 that expresses an integrated copy of the gne epimerase from Campylobacter jejuni at the wecA locus (ΔwecA-wzzE[gne]) [29, 42]. Gne is a bifunctional UDP-GlcNAc/Glc 4-epimerase that increases GalNAc availability by converting UDP-GlcNAc to UDP-GalNAc [42]. Subsequent analysis of this strain by Western blot revealed that E. coli Falcon cells express moderately more SP4 than W3110 cells (Fig. S2, compare W3110/pB4/vector to Falcon/pB4/vector) and the addition of uppS overexpression further increased SP4 expression in this strain background (Fig. S2, compare Falcon/pB4/vector to Falcon/pB4/puppS).

Since Gne activity appeared to be limiting in our strain backgrounds, we switched to overexpressing gne from a plasmid (pgne) in E. coli CLM37 cells, a derivative of W3110 lacking WecA [43]. Subsequent analysis by Western blot revealed that gne overexpression increased SP4 expression in CLM37 cells, particularly for SP4 chains < 30 kDa (Fig. 3A and C). Quantitation of SP4 levels by ELISA showed that gne overexpression in CLM37/pB4/vector cells increased SP4 expression 522% after 5 h induction and 1687% after 25 h when compared to the typical output of our previous expression efforts (Fig. 3B and D; Table 1, compare W3110/pB4/vector to CLM37/pB4/pgne/vector cells). Notably, the addition of uppS overexpression in CLM37/pB4/pgne cells further increased SP4 expression in this strain background; CLM37/pB4/pgne/puppS cells expressed significantly more SP4 chains that were also longer on average after 5- and 25-hours induction (Fig. 3A and C, compare CLM37/pB4/pgne/vector to CLM37/pB4/pgne/puppS). Quantitation of SP4 by ELISA showed that uppS overexpression in CLM37/pB4/pgne cells increased SP4 expression 568% after 5 h induction and 40% after 25 h (Fig. 3B and D; Table 1). We next investigated SP4 expression in our Und-P pathway minimized strain. Western blot and ELISA analysis revealed that gne overexpression in ΔPGT/GT cells mirrored results obtained for CLM37 cells (Fig. 3, compare CLM37/pB4/pgne/vector to ΔPGT/GT/pB4/pgne/vector). Similar trends were also observed for the addition of puppS into ΔPGT/GT/pB4/pgne cells, noting that SP4 expression was highest in this strain background after 5- and 25-hours induction (Fig. 3; Table 1, compare CLM37/pB4/pgne/puppS to ΔPGT/GT/pB4/pgne/puppS).

Fig. 3
figure 3

Increasing Und-P levels promotes recombinantly expressed S. pneumoniae capsular polysaccharide in E. coli. (A and C) Western blots showing that increasing Und-P levels enhances production of S. pneumoniae capsule serotype 4 in E. coli cells harboring plasmid pB4 (contains capsule loci). Cells with the indicated genotypes were grown at 28 °C in 2YP media for 5 or 25 h. Lysed, whole cell samples were then separated by SDS-PAGE on a 4–12% bis-tris gel and detected using anti-serotype CPS primary and anti-rabbit fluorescent secondary antibody. Results are representative of three independent experiments. (B and D) ELISA. Capsular polysaccharide quantitated from cells grown in panels A and C. All experiments represent three biological replicates, with each sample probed in duplicate. Values are expressed as means of ng capsular polysaccharide serotype 4 production per µl of culture adjusted to OD600 = 10. Error bars show standard errors of the means. Significance was determined using an ordinary one-way ANOVA with Dunnett’s multiple correction. Asterisks above sample bars denote significance relative to CLM37/pB4/pgne/vector. *p < 0.05, **<0.005, ***<0.001, **** <0.0001. Absolute capsular polysaccharide values are detailed in Table 1. The E. coli strains shown are EJK1 (W3110/pB/puppS), EJK2 (W3110/pB4/vector), EJK3 (W3110/pB4/puppS), EJK7 (CLM37/pB4/pgne/vector), EJK8 (CLM37/pB4/pgne/puppS), EJK11 (ΔPGT/GT/pB4/pgne/vector), and EJK12 (ΔPGT/GT/pB4/pgne/puppS).

Altogether, these data confirm that increasing the free pool of Und-P serves to increase glycan expression. Since polysaccharides like SP4 are used as antigens in polysaccharide-based vaccines [44] and because glycan length is correlated with immunogenicity [45], our findings have important implications for the expression and production of glycans in E. coli cells that may be used for a variety of therapeutic purposes.

Table 1 Quantification of SP4 production by sandwich ELISA

Increasing Und-P availability increases glycan expression in bacteria used to make glycoproteins

E. coli strains are used to produce glycoprotein vaccine candidates by a recombinant process termed Protein Glycan Coupling Technology (PGCT). PGCT initiates with the assembly of sugar monomers (like GalNAc) onto Und-P and culminates with the transfer of oligosaccharides onto immunogenic carrier proteins (in the periplasmic space) by enzymes possessing oligosaccharyltransferase activity like Campylobacter jejuni PglB [46]. Since Und-P availability is central to PGCT, we reasoned that increasing Und-P could be used to improve recombinant glycan expression in PGCT. As a first step to address this question, we expressed SP4 in E. coli Hobby cells [29], a derivative of Falcon deleted for the O-antigen WaaL ligase, which prevents SP4 from being transferred onto lipid A-core. Curiously, SP4 expression was poor in Hobby cells and uppS overexpression produced no observable effect (Fig. S2). Since plasmid pB4 does not contain the wzDE chain regulators from the S. pneumoniae capsule locus [28], this suggested that reintroducing wzDE may produce an effect similar to gne and uppS co-overexpression. Thus, uppS was overexpressed in E. coli Sparrowhawk cells, a derivative of Hobby cells that expresses wzDE from the wzzB locus (ΔwzzB[wzD-wzE]) [29]. As shown in the Western blot in Fig. 4A, increasing Und-P strongly increased SP4 expression in Sparrowhawk cells after 7.5 h induction, with the strongest effects observed for SP4 chains less than 50 kDa. In retrospect, these results suggest that Und-PP-linked SP4 intermediates were accumulating (but not polymerizing efficiently) in Falcon and Hobby cells overexpressing uppS (Fig. S2). We also examined SP4 expression in Sparrowhawk cells after overnight induction. However, increasing Und-P did not appear to improve SP4 expression at this time point (Fig. 4B). Collectively, these data confirm that increasing Und-P availability can be used to increase recombinant glycan expression, especially over shorter periods of time. Current efforts are focused on engineering Und-P pathways to improve other aspects of PGCT systems.

Fig. 4
figure 4

Increasing Und-P levels increases glycan expression in E. coli cells engineered for enhanced expression of S. pneumoniae capsular polysaccharides. Western blots showing that increasing Und-P levels by overexpressing uppS enhances expression of S. pneumoniae capsule serotype 4 (expressed from pB4) in E. coli Sparrowhawk cells over shorter periods of time. Cells were grown at 28 °C in 2YP media containing 500 µM IPTG (0 µM IPTG for baseline [BL]) for 7.5 h (panel A) or 23 h (panel B). Lysed, whole cell samples were then separated by SDS-PAGE on a 4–12% bis-tris gel and detected using anti-serotype CPS primary and anti-rabbit fluorescent secondary antibody. The E. coli strains shown are EJK13 (pB4), EJK14 (pB4/vector), and EJK15 (pB4/puppS).

Discussion

Many of the sugars that coat and protect bacteria are produced from oligosaccharide precursors that are built on an essential lipid carrier known as undecaprenyl phosphate (Und-P). While most research has focused on establishing the role of Und-P in scaffolding glycan synthesis, as well as understanding the physiological consequences of producing too little Und-P, less is known about the effects of increasing Und-P on glycan assembly. The present results demonstrate that increasing Und-P levels potentiates expression of native and foreign glycans. Since Und-P is a universal lipid carrier, we argue that similar results will occur for potentially any Und-P-dependent polymer expressed in E. coli cells (i.e., a rising tide lifts all boats) (Fig. 5). While increasing Und-P levels was well-tolerated in E. coli, whether similar effects occur in other bacteria will need to be determined. In summary, Und-P availability forms a critical bottleneck that can be exploited to improve glycan expression in the biotechnology benchmark bacterium E. coli.

Fig. 5
figure 5

Model depicting the effect of increasing Und-P on glycan expression. (A) Und-P synthesis. Und-P is synthesized via the methylerythritol phosphate (MEP) pathway. Und-P is first synthesized as Und-PP by the UppS synthase. Integral membrane phosphatases (BacA/PAP2 family proteins) then dephosphorylate Und-PP to Und-P (the level of which is visually indicated by the black wavy line in the bucket icon). Und-P is then distributed to phosphoglycosyl transferases and glycosyltransferases (PGT/GTs) that prime glycan assembly in non-essential and essential (peptidoglycan) pathways. In E. coli K-strains, PGT/GTs transfer precursors of peptidoglycan (MraY), O-antigen (WecA), enterobacterial common antigen (WecA), and colanic acid (WcaJ) to Und-P. Und-P is also required to glycosylate lipid A with aminoarabinose (ArnC) and O-antigen with glucose (GtrB). Since Und-P limits glycan assembly (gear icon), the glycosylation potential of any bacterium is limited by Und-P availability. (B) Simultaneously deleting all non-essential PGT/GTs and overexpressing uppS increases the amount of Und-P available for recombinant glycan expression. Our findings indicate that increasing Und-P levels increases both glycan yield and chain length

The connection between Und-P and glycan chain length

Bacterial polysaccharides like SP4 trigger antibody production by cross-linking B cell receptors on the surface of naïve B cells [47]. For this reason, the ability of plain polysaccharide vaccines to elicit effective immune responses is highly correlated with glycan chain length (i.e., longer glycan chains elicit stronger immune responses) [45]. To that end, there are likely many factors that influence glycan chain length. For example, in Wzx/Wzy-dependent systems (like SP4), Wzx translocase activity limits Und-PP-linked intermediate availability and, consequently, Wzy polymerase activity [48, 49]. Similarly, the amount of Wzy polymerases [50] and the activity of chain-length regulatory Wzz proteins [51] are known to affect glycan chain length. Our results also show that the availability of chain length regulators (WzDE, Fig. 4), as well as sugars (GalNAc, Fig. 3A and C), play important roles in determining SP4 chain length. In addition, cells that contain more Und-P also produce more glycans that are also increased in length (Fig. 3A and C). More work will be needed to determine the molecular basis for this connection, including determining the impact of Und-P levels on Wzx/Wzy activity, as well as growth dynamics and membrane composition. Since Und-P is a universal lipid carrier, the connection between Und-P and glycan chain length could be leveraged in pathogens like S. pneumoniae, whose capsular polysaccharide forms the protective basis of the pneumococcal glycoconjugate vaccine [44].

Revisiting BacA and Und-PP dephosphorylation

Und-P is central to glycan expression technology, yet most strategies to produce glycans in bacteria do not address the limiting nature of Und-P. Why? The answer to this question is likely rooted in the discovery of BacA, which dephosphorylates Und-PP to Und-P. BacA was originally identified in a screen for genes whose overexpression confers resistance to bacitracin [52], an antibiotic that binds the pyrophosphate group on Und-PP [53]. Since BacA also binds Und-PP, an excess of BacA is expected to prevent Und-PP sequestration by outcompeting bacitracin for Und-PP. However, the prevailing hypothesis at the time of its discovery was that BacA mediated resistance by increasing the free pool of Und-P. While bacA overexpression is expected to increase the rate of Und-PP dephosphorylation, no study has shown that increasing BacA activity increases the free pool of Und-P. On the contrary, bacA overexpression does not reverse the effects of a low Und-P state [33, 54] nor does it improve glycosylation efficiency [27]. Thus, early assumptions about bacA led to negative results that otherwise discouraged research into Und-P by glycan expression systems. For this reason, conclusions based on bacA overexpression should be reevaluated, ideally by overexpressing uppS or earlier steps in Und-P(P) synthesis.

Defining how much Und-P is too much

Polyprenyl phosphates like Und-P are prone to form micelles and disrupt membrane architecture when present in high concentrations [19]. For this reason, there is probably a limit to the amount of free (i.e., unbound) Und-P that cells can stably maintain. However, in addition to its free form, Und-P can also exist in a bound form as a Und-P- or Und-PP-linked intermediate. Interestingly, several recent studies have shown that cells can accommodate large quantities of Und-PP-linked intermediates [18, 20, 21]. Thus, one way to maximize the pool of Und-P without crossing the threshold for free Und-P is to steadily increase the pool of Und-P while simultaneously synthesizing more Und-PP-linked intermediates. That said, we note that the accumulation of Und-PP-linked intermediates can be deleterious (due to Und-P sequestration) [33, 55, 56], even in systems engineered to contain more Und-P [57].

Reallocating Und-P from essential pathways

Presumably, the only pathways utilizing Und-P in ΔPGT/GT/pB4 derivatives are SP4 and cell wall synthesis. In a situation where the free pool of Und-P reaches total occupancy in this strain background, SP4 production is maxed out. At this point, one way to liberate additional Und-P for SP4 production is to reduce the flux of Und-P to cell wall synthesis. Since the cell wall is essential, this brings up an interesting question: how much cell wall synthesis is necessary? The answer to this question is rooted in the amount of cell wall material required to maintain growth and integrity. Based on studies using E. coli strains auxotrophic for diaminopimelic acid (i.e., an essential component for cell wall synthesis), a two-fold reduction in the amount of cell wall per unit of cell surface area is compatible with normal growth and morphology in this bacterium [58, 59]. Thus, bacteria like E. coli produce an excess of cell wall material, presumably as a fail-safe against disruptions in cell wall synthesis. This observation suggests that cell wall synthesis can be reduced to increase Und-P levels in ΔPGT/GT cells.

Conclusions

Bacteria assemble glycans on an essential lipid carrier known as undecaprenyl phosphate (Und-P). Normally, the amount of Und-P available for glycan expression in bacteria is limited by competition between different Und-P-utilizing pathways and by the overall rate of Und-P synthesis. Here, we show that decreasing competition for Und-P and increasing Und-P synthesis promote Und-P availability and, consequently, Und-P-dependent polymer formation. Collectively, our findings show that Und-P levels should be a prime consideration for glycoengineering, which has far-reaching implications. The rational engineering of bacteria to produce greater quantities of Und-P will increase the yield of glycan expression, thereby making glycan-based therapies more affordable.

Materials and methods

General procedures

All strains, plasmids, and primers are listed in Tables S2, S3, and S4, respectively. Cells were cultured in LB miller broth (1% tryptone, 0.5% yeast extract, and 1% NaCl), 2YP broth (2% yeast extract and 4% peptone), or modified TB media (2.4% yeast extract, 2.0% tryptone, 0.5% glycerol, 100 mM phosphate pH 7.4, and 2 mM MgCl2). Plates contained 1.5% agar. As required, antibiotics were used at the following concentrations: 100 µg ml− 1 ampicillin, 50 µg ml− 1 kanamycin, 20 µg ml− 1 tetracycline, and 80 µg ml− 1 spectinomycin.

Strain construction

E. coli MG1655 wbbL + is the parent strain for this study and expresses O-antigen, which obscures the P1 binding receptor (i.e., LPS core) [60]. Thus, all gene deletions were constructed by using lambda Red recombination. Kanamycin and chloramphenicol resistance markers were excised by using FLP recombinase produced on pCP20 [61]. All gene deletions were verified by PCR. Sequences to generate deletion and check primers were obtained from the Ecocyc database [62]. PCR fragments and plasmids were purified by using kits obtained from Qiagen.

Morphological analyses of Und-P pathway mutants

Overnight cultures were diluted 1:2,000 in TB medium (Und-P pathway mutants) or TB medium containing ampicillin and 500 µM IPTG (ΔPGT/GT derivatives) and grown at 37 °C to an optical density at 600 (OD600) of 0.4–0.6 (approximately 10 doublings). Live cells were then spotted onto 1% agarose pads and imaged by phase-contrast microscopy by using an XM10 monochrome camera coupled to an Olympus BX60 microscope. Live cells for flow cytometry were prepared by pelleting 1 mL of cells (above), washing twice in phosphate-buffered saline (PBS), and diluting to an OD600 ~ 0.05. Cells were then analyzed by using the forward scatter detector in a BD LSRFortessa Flow Cytometer at the UAMS Flow Cytometry Core Facility.

Preparing cell lysates for Und-P quantification

Overnight cultures were diluted 1:2,000 in 10 mL TB (Und-P pathway mutants) or TB containing 500 µM IPTG (ΔPGT/GT derivatives) and incubated at 37 °C for 3.5 h or 24 h, respectively. Cells were then pelleted, washed with 0.9% NaCl, and resuspended in 0.7 ml of water. In parallel, cells were plated on LB medium to obtain the number of colony forming units per ml (cfu ml− 1). Cells were then lysed by transferring the suspension to glass tubes containing 3 ml of a 2:1 methanol:chloroform mixture and incubating at room temperature for 20 min. Next, glass tubes were placed in a CentriVap and centrifuged without vacuum for 20 min, after which the soluble supernatant was transferred to new glass tubes and placed at -80 °C. Once a slurry formed, the glass tubes were placed back in the CentriVap with vacuum and dried to completion. The crude cell lysate was stored in 0.2 ml of n-propanol and 0.1% ammonium hydroxide (1:3) at − 20 °C for up to one week.

LC-MS conditions for Und-P quantitation

Cell lysates (see above) were centrifuged at 10,000 x g briefly and then 5 µL were injected for LC-MS analysis. Samples were analyzed on an Agilent 1260 Infinity II system equipped with a single quadrupole electrospray ionization (ESI) MS detector. A high-pH stable Waters Xbridge Peptide BEH C18 column (4.6 × 50 mm with 2.5 µM particle size) was used. The mobile phases used included 0.1% Ammonium Hydroxide (A) and n-propanol (B). A gradient was used to evaluate cell lysates containing Und-P starting at 15% B, which was increased to 95% over 10.0 min at 1 mL min− 1, then held at 95% B for 2 min, then decreased to 15% B for an additional 2 min. A 4 min post-run at 15% B was run in-between injections. The ESI capillary voltage was applied to 4,000 V with a drying gas temperature of 350 °C at 12 L min− 1. Scanning ion mode was set to detect the [M-1 H]-1 ion species of Und-P (845.7 m/z) with a peak width of 0.1 min. A fragmentor voltage of 240 V was found to be optimal for Und-P detection.

Und-P standard curve generation and quantitation analysis

A two-step chemoenzymatic approach was employed to prepare Und-P as described previously [63]. To quantify the concentration of the sample, total phosphorous content was determined and compared to a standard curve [64, 65]. Und-P (100 µM) was supplemented to an aliquot of wild type cell lysate and the AUCTotal was determined by using the ChemStation integration tool (Agilent Technologies). Endogenous background Und-P, AUCEndo, was subtracted from AUCTotal to yield AUCAdjusted to reduce the influence of the sample matrix effects. Injections of 25, 50, 100, 500, and 1,000 pmol of Und-P were used to generate the LC-MS response curve for quantitation. For cell lysate quantitation, the slope and y-intercept from the Und-P standard were used to calculate the quantity of Und-P based on AUC. Und-P levels were normalized by dividing Und-P measurements by the mean cfu ml− 1.

ECA dot blot

Cells were cultured in LB medium containing ampicillin and 500 µM IPTG (no IPTG was added to puppS baseline [BL]) at 37 °C overnight. Total cell material was matched by centrifuging the equivalent of 1 ml of culture at an OD600 = 1.0 and resuspending the pellet in 20 µl of PBS. 3 µl aliquots were then applied to a nitrocellulose membrane and allowed to dry for 20 min. Membranes were blocked with 5% skimmed milk in PBS for 1 h. Rabbit anti-ECA serum (Statens Serum Institut) was then applied at 1:1,000 in PBS for 3 h. The primary antibody was detected by using the goat anti-rabbit IgG conjugated to Alexa Fluor 488 (Thermo Fisher Scientific) secondary antibody and was applied at 1:4,000 for 1 h. Blots were imaged by using a ChemiDoc MP imaging system (Bio-Rad Laboratories). Signals corresponding to surface ECA were quantified by using ImageJ.

ECA immunoblot

Cells were cultured in LB media containing 500 µM IPTG (0 µM IPTG was added to puppS baseline [BL]) at 37 °C overnight. Total cell material was matched by centrifuging the equivalent of 1 ml of culture at an OD600 = 1.0 and resuspending the pellet in 25 µL BugBuster® Protein Extraction Reagent (MilliporeSigma) containing 1:100 benzonase. Cell suspensions were incubated at room temperature for 10 min, mixed with loading buffer (LI-COR Biosciences), and boiled for 5 min. 10 µL aliquots were then separated on 12% Mini-PROTEAN® TGX™ (Bio-Rad Laboratories) gels. ECA polymers were transferred onto nitrocellulose membranes using a Trans Blot Turbo system (Bio-Rad Laboratories) set at 25 V (1.3 amps) for 7 min. Membranes were dried at 37 °C and then blocked with 5% skimmed milk in PBS for 1 h. Membranes were then washed three times with PBS (0.2% Tween 20). Rabbit anti-ECA antibody (a gift from Angela Mitchell) was then applied at 1:24,000 in PBS (0.2% Tween 20) for 1 h. The membranes were then washed three times with PBS (0.2% Tween 20). The primary antibody was detected by using the goat anti-rabbit IgG conjugated to IRDye680 (LI-COR Biosciences) secondary antibody at 1:10,000 for 1 h. Membranes were washed a further three times in PBS (0.2% Tween 20) before signal detection with an Odyssey LI-COR detection system (LI-COR Biosciences). Similar procedures were used to detect MreB; the primary rabbit anti-MreB antibody was applied at 1:2,000 for 1 h and the secondary goat anti-rabbit IgG conjugated to AF488 (Thermo Fisher Scientific) antibody at 1:4,000 for 1 h.

Recombinant capsule expression and immunoblot

Overnight cultures were diluted to an OD600 of 0.03 in 10 ml of 2YP medium containing 500 µM of IPTG and cultured at 28 °C. Cells from 1.5 ml of culture grown from various times were withdrawn and pelleted. Pellets were then resuspended to an OD600 = 10.0 in PBS, mixed with 1 mg ml− 1 lysozyme and 40 U ml− 1 benzonase, and boiled for 10 min. Lysed samples were mixed with NuPAGE LDS sample buffer (Invitrogen) and separated on Bolt 4–12% Bis-Tris 1 mm gels in MOPS buffer (Invitrogen) at 100 V for 2 h 25 min in an ice bath. Capsular polysaccharides were transferred onto nitrocellulose membranes using an iBlot 2 dry transfer system (Thermo Fisher) set at 20 V for 1 min, 23 V for 4 min, and 25 V for 2 min. Membranes were blocked with 2% skimmed milk in PBS for 1 h. S. pneumoniae Serotype 4 (SP4) rabbit anti-capsule antibody (Statens Serum Institut) was then applied at 1:1,000 in PBS containing 2% skimmed milk and 0.1% Tween 20 for 1 h. The membranes were then washed three times with PBS (0.1% Tween 20). The primary antibody was detected by using the goat anti-rabbit IgG conjugated to IRDye800 (LI-COR Biosciences) secondary antibody at 1:10,000 for 45 min. Membranes were washed a further three times in PBS (0.1% Tween 20) and once with PBS before signal detection with an Odyssey LI-COR detection system (LI-COR Biosciences).

Indirect enzyme-linked immunosorbent assay (ELISA)

Dilutions of boiled E. coli lysates were used to coat a MaxiSorp microtiter plate (Nunc) overnight at 4 °C. A standard curve was generated by using dilutions of purified type 4 pneumococcal polysaccharide (Statens Serum Institut). Wells were then washed with PBS containing 0.05% Tween 20 four times, blocked with PBS containing 5% skimmed milk for 1.5 h, and washed twice again. SP4 rabbit anti-capsule antibody (Statens Serum Institut, Denmark) was then applied at 1:1,000 in PBS containing 2% skimmed milk for 1 h. The wells were then washed four times with PBS 0.05% Tween 20. The primary antibody was detected by using the goat anti-rabbit IgG HRP (Abcam) secondary antibody at 1:20,000 in PBS containing 2% milk for 45 min. After washing four times with PBS 0.05% Tween 20, tetramethylbenzidine (eBioscience) was added, and the reaction was stopped with 2M H2SO4 (sulfuric acid). Indirect ELISA detection was performed using a SpectrMax iD5 microplate reader (Molecular Devices) at an absorbance of 450 nm.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Silhavy TJ, Kahne D, Walker S. The bacterial cell envelope. Cold Spring Harb Perspect Biol. 2010;2:a000414.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Tytgat HL, Lebeer S. The sweet tooth of bacteria: common themes in bacterial glycoconjugates. Microbiol Mol Biol Rev. 2014;78:372–417.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Manat G, Roure S, Auger R, Bouhss A, Barreteau H, Mengin-Lecreulx D, Touze T. Deciphering the metabolism of undecaprenyl-phosphate: the bacterial cell-wall unit carrier at the membrane frontier. Microb Drug Resist. 2014;20:199–214.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Apfel CM, Takacs B, Fountoulakis M, Stieger M, Keck W. Use of genomics to identify bacterial undecaprenyl pyrophosphate synthetase: cloning, expression, and characterization of the essential uppS gene. J Bacteriol. 1999;181:483–92.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Kato J, Fujisaki S, Nakajima K, Nishimura Y, Sato M, Nakano A. The Escherichia coli homologue of yeast RER2, a key enzyme of dolichol synthesis, is essential for carrier lipid formation in bacterial cell wall synthesis. J Bacteriol. 1999;181:2733–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. El Ghachi M, Bouhss A, Blanot D, Mengin-Lecreulx D. The bacA gene of Escherichia coli encodes an undecaprenyl pyrophosphate phosphatase activity. J Biol Chem. 2004;279:30106–13.

    Article  PubMed  Google Scholar 

  7. El Ghachi M, Derbise A, Bouhss A, Mengin-Lecreulx D. Identification of multiple genes encoding membrane proteins with undecaprenyl pyrophosphate phosphatase (UppP) activity in Escherichia coli. J Biol Chem. 2005;280:18689–95.

    Article  PubMed  Google Scholar 

  8. Manat G, El Ghachi M, Auger R, Baouche K, Olatunji S, Kerff F, Touze T, Mengin-Lecreulx D, Bouhss A. Membrane topology and biochemical characterization of the Escherichia coli BacA Undecaprenyl-Pyrophosphate phosphatase. PLoS ONE. 2015;10:e0142870.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Touze T, Blanot D, Mengin-Lecreulx D. Substrate specificity and membrane topology of Escherichia coli PgpB, an undecaprenyl pyrophosphate phosphatase. J Biol Chem. 2008;283:16573–83.

    Article  CAS  PubMed  Google Scholar 

  10. El Ghachi M, Howe N, Huang CY, Olieric V, Warshamanage R, Touze T, Weichert D, Stansfeld PJ, Wang M, Kerff F, Caffrey M. Crystal structure of undecaprenyl-pyrophosphate phosphatase and its role in peptidoglycan biosynthesis. Nat Commun. 2018;9:1078.

    Article  PubMed  PubMed Central  ADS  Google Scholar 

  11. Fan J, Jiang D, Zhao Y, Liu J, Zhang XC. Crystal structure of lipid phosphatase Escherichia coli phosphatidylglycerophosphate phosphatase B. Proc Natl Acad Sci USA. 2014;111:7636–40.

    Article  CAS  PubMed  PubMed Central  ADS  Google Scholar 

  12. Lukose V, Walvoort MTC, Imperiali B. Bacterial phosphoglycosyl transferases: initiators of glycan biosynthesis at the membrane interface. Glycobiology. 2017;27:820–33.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Guan S, Bastin DA, Verma NK. Functional analysis of the O antigen glucosylation gene cluster of Shigella flexneri bacteriophage SfX. Microbiology. 1999;145(Pt 5):1263–73.

    Article  CAS  PubMed  Google Scholar 

  14. Breazeale SD, Ribeiro AA, McClerren AL, Raetz CR. A formyltransferase required for Polymyxin Resistance in Escherichia coli and the modification of lipid A with 4-Amino-4-deoxy-L-arabinose. Identification and function of UDP-4-deoxy-4-formamido-L-arabinose. J Biol Chem. 2005;280:14154–67.

    Article  CAS  PubMed  Google Scholar 

  15. Liu D, Reeves PR. Escherichia coli K12 regains its O antigen. Microbiol (Reading). 1994;140(Pt 1):49–57.

    Article  CAS  Google Scholar 

  16. Jeong H, Barbe V, Lee CH, Vallenet D, Yu DS, Choi SH, Couloux A, Lee SW, Yoon SH, Cattolico L, et al. Genome sequences of Escherichia coli B strains REL606 and BL21(DE3). J Mol Biol. 2009;394:644–52.

    Article  CAS  PubMed  Google Scholar 

  17. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol. 2006;2:20060008.

    Article  Google Scholar 

  18. Eade CR, Wallen TW, Gates CE, Oliverio CL, Scarbrough BA, Reid AJ, Jorgenson MA, Young KD, Troutman JM. Making the Enterobacterial Common Antigen Glycan and Measuring its substrate sequestration. ACS Chem Biol. 2021;16:691–700.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Hartley MD, Imperiali B. At the membrane frontier: a prospectus on the remarkable evolutionary conservation of polyprenols and polyprenyl-phosphates. Arch Biochem Biophys. 2012;517:83–97.

    Article  CAS  PubMed  Google Scholar 

  20. Sham LT, Butler EK, Lebar MD, Kahne D, Bernhardt TG, Ruiz N. Bacterial cell wall. MurJ is the flippase of lipid-linked precursors for peptidoglycan biogenesis. Science. 2014;345:220–2.

    Article  CAS  PubMed  PubMed Central  ADS  Google Scholar 

  21. Reid AJ, Eade CR, Jones KJ, Jorgenson MA, Troutman JM. Tracking colanic acid repeat unit formation from stepwise biosynthesis inactivation in Escherichia coli. Biochemistry. 2021;60:2221–30.

    Article  CAS  PubMed  Google Scholar 

  22. Herbert JA, Kay EJ, Faustini SE, Richter A, Abouelhadid S, Cuccui J, Wren B, Mitchell TJ. Production and efficacy of a low-cost recombinant pneumococcal protein polysaccharide conjugate vaccine. Vaccine. 2018;36:3809–19.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Reglinski M, Ercoli G, Plumptre C, Kay E, Petersen FC, Paton JC, Wren BW, Brown JS. A recombinant conjugated pneumococcal vaccine that protects against murine infections with a similar efficacy to Prevnar-13. Npj Vaccines. 2018;3:53.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Harding CM, Nasr MA, Scott NE, Goyette-Desjardins G, Nothaft H, Mayer AE, Chavez SM, Huynh JP, Kinsella RL, Szymanski CM, et al. A platform for glycoengineering a polyvalent pneumococcal bioconjugate vaccine using E. Coli as a host. Nat Commun. 2019;10:891.

    Article  PubMed  PubMed Central  ADS  Google Scholar 

  25. Pandhal J, Ow SY, Noirel J, Wright PC. Improving N-glycosylation efficiency in Escherichia coli using shotgun proteomics, metabolic network analysis, and selective reaction monitoring. Biotechnol Bioeng. 2011;108:902–12.

    Article  CAS  PubMed  Google Scholar 

  26. Pandhal J, Woodruff LBA, Jaffe S, Desai P, Ow SY, Noirel J, Gill RT, Wright PC. Inverse metabolic Engineering to improve Escherichia coli as an N-Glycosylation host. Biotechnol Bioeng. 2013;110:2482–93.

    Article  CAS  PubMed  Google Scholar 

  27. Pandhal J, Desai P, Walpole C, Doroudi L, Malyshev D, Wright PC. Systematic metabolic engineering for improvement of glycosylation efficiency in Escherichia coli. Biochem Biophys Res Commun. 2012;419:472–6.

    Article  CAS  PubMed  Google Scholar 

  28. Kay EJ, Yates LE, Terra VS, Cuccui J, Wren BW. Recombinant expression of Streptococcus pneumoniae capsular polysaccharides in Escherichia coli. Open Biol. 2016;6:150243.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Kay EJ, Mauri M, Willcocks SJ, Scott TA, Cuccui J, Wren BW. Engineering a suite of E. Coli strains for enhanced expression of bacterial polysaccharides and glycoconjugate vaccines. Microb Cell Fact. 2022;21:66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Rendueles O, Beloin C, Latour-Lambert P, Ghigo JM. A new biofilm-associated colicin with increased efficiency against biofilm bacteria. ISME J. 2014;8:1275–88.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Sala RF, Morgan PM, Tanner ME. Enzymatic formation and release of a stable glycal intermediate: the mechanism of the reaction catalyzed by UDP-N-acetylglucosamine 2-epimerase. J Am Chem Soc. 1996;118:3033–4.

    Article  CAS  Google Scholar 

  32. Jorgenson MA, Young KD. Interrupting biosynthesis of O Antigen or the Lipopolysaccharide Core produces morphological defects in Escherichia coli by sequestering undecaprenyl phosphate. J Bacteriol. 2016;198:3070–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Jorgenson MA, Kannan S, Laubacher ME, Young KD. Dead-end intermediates in the enterobacterial common antigen pathway induce morphological defects in Escherichia coli by competing for undecaprenyl phosphate. Mol Microbiol. 2016;100:1–14.

    Article  CAS  PubMed  Google Scholar 

  34. Jorgenson MA, MacCain WJ, Meberg BM, Kannan S, Bryant JC, Young KD. Simultaneously inhibiting undecaprenyl phosphate production and peptidoglycan synthases promotes rapid lysis in Escherichia coli. Mol Microbiol. 2019;112:233–48.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Rojas ER, Billings G, Odermatt PD, Auer GK, Zhu L, Miguel A, Chang F, Weibel DB, Theriot JA, Huang KC. The outer membrane is an essential load-bearing element in Gram-negative bacteria. Nature. 2018;559:617–21.

    Article  CAS  PubMed  PubMed Central  ADS  Google Scholar 

  36. Barreteau H, Magnet S, El Ghachi M, Touze T, Arthur M, Mengin-Lecreulx D, Blanot D. Quantitative high-performance liquid chromatography analysis of the pool levels of undecaprenyl phosphate and its derivatives in bacterial membranes. J Chromatogr B Analyt Technol Biomed Life Sci. 2009;877:213–20.

    Article  CAS  PubMed  Google Scholar 

  37. Maczuga N, Tran ENH, Qin J, Morona R. Interdependence of Shigella flexneri O Antigen and Enterobacterial Common Antigen Biosynthetic Pathways. J Bacteriol. 2022;204:e0054621.

    Article  PubMed  Google Scholar 

  38. Fujisaki S, Nishino T, Katsuki H. Isoprenoid synthesis in Escherichia coli. Separation and partial purification of four enzymes involved in the synthesis. J Biochem. 1986;99:1327–37.

    Article  CAS  PubMed  Google Scholar 

  39. Asai K, Fujisaki S, Nishimura Y, Nishino T, Okada K, Nakagawa T, Kawamukai M, Matsuda H. The identification of Escherichia coli ispB (CEL) gene encoding the octaprenyl diphosphate synthase. Biochem Biophys Res Commun. 1994;202:340–5.

    Article  CAS  PubMed  Google Scholar 

  40. Schmidt G, Jann B, Jann K. Genetic and immunochemical studies on Escherichia coli O14:K7:H. Eur J Biochem. 1974;42:303–9.

    Article  CAS  PubMed  Google Scholar 

  41. Rai AK, Mitchell AM. Enterobacterial Common Antigen: Synthesis and Function of an Enigmatic Molecule. mBio 2020, 11.

  42. Bernatchez S, Szymanski CM, Ishiyama N, Li J, Jarrell HC, Lau PC, Berghuis AM, Young NM, Wakarchuk WW. A single bifunctional UDP-GlcNAc/Glc 4-epimerase supports the synthesis of three cell surface glycoconjugates in Campylobacter jejuni. J Biol Chem. 2005;280:4792–802.

    Article  CAS  PubMed  Google Scholar 

  43. Linton D, Dorrell N, Hitchen PG, Amber S, Karlyshev AV, Morris HR, Dell A, Valvano MA, Aebi M, Wren BW. Functional analysis of the Campylobacter jejuni N-linked protein glycosylation pathway. Mol Microbiol. 2005;55:1695–703.

    Article  CAS  PubMed  Google Scholar 

  44. Daniels CC, Rogers PD, Shelton CM. A review of Pneumococcal vaccines: current polysaccharide vaccine recommendations and future protein antigens. J Pediatr Pharmacol Ther. 2016;21:27–35.

    PubMed  PubMed Central  Google Scholar 

  45. Anish C, Beurret M, Poolman J. Combined effects of glycan chain length and linkage type on the immunogenicity of glycoconjugate vaccines. NPJ Vaccines. 2021;6:150.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Kay E, Cuccui J, Wren BW. Recent advances in the production of recombinant glycoconjugate vaccines. NPJ Vaccines. 2019;4:16.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Mitchell R, Kelly DF, Pollard AJ, Truck J. Polysaccharide-specific B cell responses to vaccination in humans. Hum Vaccin Immunother. 2014;10:1661–8.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Hong Y, Reeves PR. Diversity of O-antigen repeat unit structures can account for the substantial sequence variation of Wzx translocases. J Bacteriol. 2014;196:1713–22.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Hong Y, Hu D, Verderosa AD, Qin J, Totsika M, Reeves PR. Repeat-unit elongations to produce bacterial Complex Long Polysaccharide Chains, an O-Antigen Perspective. EcoSal Plus. 2023;11:eesp00202022.

    Article  PubMed  Google Scholar 

  50. Carter JA, Jimenez JC, Zaldivar M, Alvarez SA, Marolda CL, Valvano MA, Contreras I. The cellular level of O-antigen polymerase wzy determines chain length regulation by WzzB and WzzpHS-2 in Shigella flexneri 2a. Microbiol (Reading). 2009;155:3260–9.

    Article  CAS  Google Scholar 

  51. Whitfield C, Williams DM, Kelly SD. Lipopolysaccharide O-antigens-bacterial glycans made to measure. J Biol Chem. 2020;295:10593–609.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Cain BD, Norton PJ, Eubanks W, Nick HS, Allen CM. Amplification of the bacA gene confers bacitracin resistance to Escherichia coli. J Bacteriol. 1993;175:3784–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Stone KJ, Strominger JL. Mechanism of action of bacitracin: complexation with metal ion and C55 -isoprenyl pyrophosphate. Proc Natl Acad Sci USA. 1971;68:3223–7.

    Article  CAS  PubMed  PubMed Central  ADS  Google Scholar 

  54. Danese PN, Oliver GR, Barr K, Bowman GD, Rick PD, Silhavy TJ. Accumulation of the enterobacterial common antigen lipid II biosynthetic intermediate stimulates degP transcription in Escherichia coli. J Bacteriol. 1998;180:5875–84.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Yuasa R, Levinthal M, Nikaido H. Biosynthesis of cell wall lipopolysaccharide in mutants of Salmonella. V. A mutant of Salmonella typhimurium defective in the synthesis of cytidine diphosphoabequose. J Bacteriol. 1969;100:433–44.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  56. D’Elia MA, Millar KE, Bhavsar AP, Tomljenovic AM, Hutter B, Schaab C, Moreno-Hagelsieb G, Brown ED. Probing teichoic acid genetics with bioactive molecules reveals new interactions among diverse processes in bacterial cell wall biogenesis. Chem Biol. 2009;16:548–56.

    Article  PubMed  Google Scholar 

  57. Ranjit DK, Young KD. Colanic Acid Intermediates Prevent De Novo Shape Recovery of Escherichia coli spheroplasts, calling into question Biological roles previously attributed to Colanic Acid. J Bacteriol. 2016;198:1230–40.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Prats R, de Pedro MA. Normal growth and division of Escherichia coli with a reduced amount of murein. J Bacteriol. 1989;171:3740–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Caparros M, Quintela JC, de Pedro MA. Variability of peptidoglycan surface density in Escherichia coli. FEMS Microbiol Lett. 1994;121:71–6.

    Article  CAS  PubMed  Google Scholar 

  60. Ho TD, Waldor MK. Enterohemorrhagic Escherichia coli O157:H7 gal mutants are sensitive to bacteriophage P1 and defective in intestinal colonization. Infect Immun. 2007;75:1661–6.

    Article  CAS  PubMed  Google Scholar 

  61. Cherepanov PP, Wackernagel W. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene. 1995;158:9–14.

    Article  CAS  PubMed  Google Scholar 

  62. Keseler IM, Mackie A, Santos-Zavaleta A, Billington R, Bonavides-Martinez C, Caspi R, Fulcher C, Gama-Castro S, Kothari A, Krummenacker M, et al. The EcoCyc database: reflecting new knowledge about Escherichia coli K-12. Nucleic Acids Res. 2017;45:D543–50.

    Article  CAS  PubMed  Google Scholar 

  63. Reid AJ, Scarbrough BA, Williams TC, Gates CE, Eade CR, Troutman JM. General utilization of fluorescent polyisoprenoids with Sugar Selective Phosphoglycosyltransferases. Biochemistry. 2020;59:615–26.

    Article  CAS  PubMed  Google Scholar 

  64. King EJ. The colorimetric determination of phosphorus. Biochem J. 1932;26:292–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Chen PS, Toribara TY, Warner H. Microdetermination of Phosphorus. Anal Chem. 1956;28:1756–8.

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We thank Angela Mitchell (Texas A&M University) for providing anti-ECA antibody and for technical guidance in imaging ECA, Mario Feldman (Washington University in St. Louis) for providing plasmid pMAF12, and Kenn Gerdes (University of Copenhagen) for providing anti-MreB antibody. We also thank Kevin Young for critically reviewing the manuscript.

Funding

Work in the Wren lab was supported by the Biotechnology and Biological Sciences Research Council (BBSRC) Grants BB/R008124/1, BB/N001591/1, and BB/S004963/1. Work in the Troutman lab was supported by NIH R01GM123251 and F31GM130065. Work in the Jorgenson lab was supported by NIH R01GM061019 and P20GM103625. The UAMS DNA and flow cytometry core facilities are supported in part by the Center for Microbial Pathogenesis and Host Inflammatory Responses NIH grant P20GM103625. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Author information

Authors and Affiliations

Authors

Contributions

MJ and EK conceived the study. MJ and JB constructed Und-P pathway mutants and performed phenotypic experiments. EK measured SP4 expression in designer strains. MD and AR measured Und-P levels. BW and JT provided feedback on data interpretation. The manuscript was written through contributions of all authors. All authors read and approved the manuscript.

Corresponding author

Correspondence to Matthew A. Jorgenson.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kay, E.J., Dooda, M.K., Bryant, J.C. et al. Engineering Escherichia coli for increased Und-P availability leads to material improvements in glycan expression technology. Microb Cell Fact 23, 72 (2024). https://doi.org/10.1186/s12934-024-02339-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12934-024-02339-8

Keywords