Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2008 Dec 18.
Published in final edited form as: Biochemistry. 2007 Nov 23;46(50):14342–14348. doi: 10.1021/bi701956x

Polyisoprenol Specificity in the Campylobacter jejuni N-linked Glycosylation Pathway

Mark M Chen 1, Eranthie Weerapana 1, Ewa Ciepichal 2, Jacek Stupak 3, Christopher W Reid 3, Ewa Swiezewska 2, Barbara Imperiali 1,*
PMCID: PMC2585822  NIHMSID: NIHMS62979  PMID: 18034500

Abstract

Campylobacter jejuni contains a general N-linked glycosylation pathway in which a heptasaccharide is sequentially assembled onto a polyisoprenyl-diphosphate carrier and subsequently transferred to the asparagine side chain of an acceptor protein. The enzymes in the pathway function at a membrane interface and have in common amphiphilic membrane-bound polyisoprenyl-linked substrates. Herein we examine the potential role of the polyisoprene component of the substrates by investigating the relative substrate efficiencies of polyisoprene-modified analogs in individual steps in the pathway. Chemically defined substrates for PglC, PglJ and PglB are prepared via semisynthetic approaches. The substrates included polyisoprenols of varying length, double bond geometry, and degree of saturation to probe the role of the hydrophobic polyisoprene in substrate specificity. Kinetic analysis reveals that all three enzymes exhibit distinct preferences for the polyisoprenol carrier whereby cis-double bond geometry and α-unsaturation of the native substrate are important features, while the precise polyisoprene length may be less critical. These finding suggest that the polyisoprenol carrier plays a specific role in the function of these enzymes beyond a purely physical role as a membrane anchor. These studies underscore the potential of the C. jejuni N-linked glycosylation pathway as a system for investigating the biochemical and biophysical roles of polyisoprenol carriers common to prokaryotic and eukaryotic glycosylation.

Keywords: Campylobacter jejuni, PglB, PglC, PglJ, polyisoprene donor

The gram-negative bacterium Campylobacter jejuni contains a general N-linked protein glycosylation (pgl) pathway (1) with significant similarity to the corresponding eukaryotic pathway (2). In this pathway a phosphoglycosyltransferase (PglC) and a series of glycosyltransferases (PglA, PglJ, PglH and PglI) act sequentially to assemble a heptasaccharide on an undecaprenyl-diphosphate (Und-PP) carrier anchored at the cytoplasmic face of the inner membrane (Figure 1) (3, 4). The polyisoprenyl-linked heptasaccharide, GalNAc-α1,4-GalNAc-α1,4-(Glcβ1,3)-GalNAc-α1,4- GalNAc-α1,4-GalNAc-α1,3-Bac2,4diNAc (5), where Bac2,4diNAc is 2,4-diacetamido-2,4,6-trideoxyglucose, is then flipped by PglK to the periplasmic face of the membrane (6). In the periplasm, PglB, the bacterial oligosaccharyl transferase (OT), transfers the heptasaccharide to the asparagine side chain of acceptor proteins within a D/E-X1-N-X2-S/T consensus sequence, where X can be any amino acid except proline (7, 8). This pathway parallels the N-linked glycosylation process in eukaryotes, where a tetradecasaccharide is sequentially assembled on a dolichyl-diphosphate carrier prior to transfer to protein (9). In contrast to eukaryotic N-linked glycosylation, the enzymes from C. jejuni can be readily over-expressed in functional form in E. coli (10), which makes them more amenable to biochemical studies. Furthermore, each substrate in the pathway can be prepared using chemoenzymatic strategies (11) and purified to homogeneity, thereby enabling comparative kinetic studies of the function of the C. jejuni enzymes in vitro. The studies presented illustrate the use of C. jejuni as an ideal model system for investigating the complex process of N-linked protein glycosylation and suggest the opportunity for understanding the fundamental roles played by the polyisoprene component of the substrates in the membrane-associated pathway.

Figure 1.

Figure 1

Open in a new tab

Campylobacter jejuni N-linked glycosylation pathway.

The goal of this study is to address whether the polyisoprenol carrier that is common to both eukaryotic and prokaryotic protein glycosylation, plays a simple physical role as a hydrophobic membrane anchor or a more specific chemical role in substrate/enzyme interactions. The C. jejuni pathway integrates the functions of a series of enzymes that demonstrate a spectrum of membrane-association properties. The current efforts focus on three enzymes from this pathway (Figure 2), PglC, PglJ, and PglB, which act on undecaprenyl-linked substrates at various stages of the N-linked glycosylation process and differ significantly in the degree of membrane association (Figure S4 in the Supplementary Info). The phosphoglycosyltransferase PglC transfers a Bac2,4diNAc-phosphate to undecaprenyl-phosphate to afford the first membrane-associated intermediate in the pathway (4). PglC includes a single N-terminal transmembrane domain of 22 amino acids (TMHMM, ExPASY) and a large cytoplasmic domain. An intermediate step in the pathway is catalyzed by PglJ, a glycosyltransferase, which transfers a GalNAc to the Und-PP-Bac2,4diNAc-GalNAc disaccharide (3). PglJ is not predicted to include any transmembrane domains, but is predicted to include an N-terminal hydrophobic domain, which is postulated to interact with the bacterial membrane (Figure S4). The final step in the pathway is catalyzed by PglB, which transfers the completed heptasaccharide from the undecaprenyl-diphosphate carrier to the protein acceptor (7). PglB comprises 10–12 predicted transmembrane domains (TMHMM, ExPASY) and a small C-terminal periplasmic domain. These three enzymes provide a panel of proteins that span the spectrum of membrane association, ranging from no transmembrane domains to multiple integral membrane regions.

Figure 2.

Figure 2

Open in a new tab

PglC and PglA coupled reaction for the synthesis of polyisoprenyl-PP-Bac2,4diNAc-GalNAc, followed by PglB transfer of the glycan to the peptide acceptor. At position-6 of bacillosamine: H = Bac2,4diNAc, OH = 6-hydroxylBac2,4diNAc.

Additionally, the site of action of these enzymes varies, with PglC and PglB acting near the polyisoprenyl-phosphate group proximal to the membrane and PglJ catalyzing glycoside bond formation at a site away from the membrane at a distal saccharide unit. Using this system, the goal was to explore the extent of interaction of these enzymes with the polyisoprenyl moiety. If the native undecaprenyl-carrier fulfills a specific role in substrate recognition, we would anticipate that modifying the structure of the undecaprenyl moiety to polyisoprenols differing in length and saturation would influence enzyme activity. In contrast, if the undecaprenyl moiety only serves a physical role as a hydrophobic anchor, then the introduction of structurally altered polyisoprenols of comparable hydrophobicity should have a negligible effect on activity. The diverse physical properties of the enzymes in the system also allow us to probe how the membrane-association properties of each enzyme dictate the degree of interaction with the polyisoprenyl group in the substrates.

To address these questions, the native undecaprenyl-linked substrates for PglC, PglJ and PglB were synthesized using chemoenzymatic approaches and purified to homogeneity. Additionally, a panel of “non-natural” substrates for each enzyme was obtained by analogous approaches that incorporated polyisoprenols that differed in polyisoprene length, degree of saturation, and double-bond geometry (Figure 3). Kinetic analyses were performed using the various polyisoprenyl-linked substrates for each of the three enzymes (PglC, PglJ, and PglB). The studies revealed that the activity of each of the enzymes was significantly influenced by the structure of the polyisoprenol. Additionally, the results show remarkably common trends in the specificity for each of the enzymes studied regardless of the degree of membrane association of each protein.

Figure 3.

Figure 3

Open in a new tab

Polyisoprenols used in this study.

Understanding the polyisoprenol recognition of these enzymes will afford valuable general insight into the determinants that mediate the interaction of membrane-bound enzymes with their lipophilic substrates. Furthermore, C. jejuni is known to be involved in human gastroenterological disorders (12, 13), and evidence suggests that the N-linked glycans of this microorganism play a critical role in host adherence, invasion, and colonization which ultimately profoundly impact pathogenicity (14, 15). Therefore, this understanding of substrate specificity may facilitate the development of therapeutics towards gastroenterological disorders caused by C. jejuni, whilst also providing clues to manipulation of this pathway for future glycoengineering applications (10, 16).

MATERIAL AND METHODS

UDP-sugar Donors

UDP-Bac2,4diNAc was synthesized enzymatically from UDP-GlcNAc using the PglF, PglE, and PglD enzymes from the C. jejuni pathway as described previously (17). The preparation of UDP-6-hydroxylBac2,4diNAc, which is a synthetic intermediate in the chemical synthesis of UDP-Bac2,4diNAc was carried out as reported previously (4, 11). Radiolabeled UDP-[3H]GalNAc was purchased from American Radiolabeled Chemicals Inc.

Polyisoprenol Isolation & Phosphorylation

Solanesol and geranylgeraniol were purchased from Sigma-Aldrich. All remaining polyisoprenols were isolated from diverse plant sources (18) and purified from the natural extracts by chromatographic methods (19). Undecaprenol was from Magnolia kobus leaves, prenol-9 was from the Tilia cordata leaves, prenol-8 was from the wood of Betula veruccosa, and prenols-17–19 were from Ginkgo biloba leaves. Hydrogenation of undecaprenol resulted in a racemic mixture of dihydroprenol-11 (20). Dolichols-17–19 were synthesized via asymmetric hydrogenation of prenols-17–19 (21). Undecaprenol, dolichols, solanesol, and geranylgeraniol were chemically phosphorylated using phosphoramidite chemistry (22, 23). Prenol-9, prenol-8, were phosphorylated with phosphoric acid and trichloroacetonitrile (24). Dihydroprenol-11 was phosphorylated with phosphorus oxychloride (25). The purity of the polyisoprenyl-phosphates was evaluated by thin layer chromatography and determined to be >95% in all cases.

Enzyme Purification

The expression and purification of PglA (3), PglJ (3), PglC (4), and PglB (7) enzymes have been previously described in detail. For this study, both PglA and PglJ were further subjected to an additional chromatographic purification on a Sephacryl 300 gel filtration column (GE Healthcare). PglA and PglJ were evaluated to be 95% pure by SDS PAGE, while PglC and PglB were used as a semi-pure membrane fraction necessary for preserving enzyme stability.

Enzymatic synthesis of Radiolabeled Polyisoprenyl Diphosphate Disaccharide Donors

To an eppendorf tube containing 50 nmol of the polyisoprenyl-phosphate was added 6 µL DMSO and 14 µL of 14.3% (v/v) Triton X-100. After vortexing and sonication (water bath), 42.5 µL of H2O, 5.5 µL of 1 M Tris-acetate (pH 8.0), 2 µL of 1 M MgCl2, 20 µL of 5 mM UDP-Bac2,4diNAc, 20 µL of 2.5 µM UDP-GalNAc (specific activity = 152 nCi nmol−1), 70 µL of PglA, and 40 µL of PglC were added. The concentrations of enzyme stock solutions used were approximately 500 mg mL−1 and 250 mg mL−1 for PglA (3) and PglC (4) respectively. Reactions were left shaking at room temperature for 3 hours then quenched by the addition of 800 µL 2:1 chloroform:methanol and 100 µL of pure solvent upper phase (PSUP; 3% chloroform, 49% methanol, and 48% water with 100 mM KCl). After brief vortexing, the layers were allowed to separate and the aqueous layer was removed. The organic layer was washed three times with 200 µL of PSUP, dried under a stream of nitrogen, and then redissolved in 100 µL of 4:1 chloroform:methanol. The crude product was then purified by HPLC on a Microsorb 60 Si normal phase column (Varian), eluting with a linear gradient of 100% solvent C (4:1 chloroform:methanol) to 100% solvent D (10:10:3 chloform:methanol:2 M ammonium acetate pH 7.2) over 50 min at a flow rate of 1 mL/min. Fractions containing significant amount of radiolabeled product were combined and immediately aliquoted, dried, and stored at −80°C until needed.

PglJ Radioactivity Assay

To a tube containing a specified amount of dried radiolabeled polyisoprenyl-diphosphate Bac2,4diNAc-[3H]GalNAc were added 10 µL DMSO and 7 µL of 1.43% (v/v) Triton X-100. After vortexing and sonication (water bath), 5 µL of 1 M Tris-acetate (pH 8.0), 1 µL of 1 M MgCl2, 5µL of 50 mM DTT, 62 µL of H2O, and 5 µL of 500 mg mL−1 PglJ were added. For negative control reactions (not shown), the PglJ samples were replaced by 50 mM Tris-acetate buffer (pH 8.0). Four reactions containing different amounts of polyisoprenyl-diphosphate disaccharides were carried out in parallel. The reactions were initiated by the addition of 5 µL of 1 µM radiolabeled UDP-GalNAc (specific activity = 19.8 µCi nmol−1, KM(UDP-GalNAc) ~1 µM). All assays were carried out at 22 ± 1 °C, and 17 µL aliquots were taken at 2, 4, 6, and 8 min. Reactions were quenched by addition to an eppendorf tube containing 800 µL 2:1 chloroform:methanol and 200 µL of PSUP (26). After brief vortexing, the layers were allowed to separate and the aqueous layer was removed. The organic layer was washed two times with 100 µL of PSUP and dried under a stream of nitrogen. The residue was then redissolved in 200 µL of DMSO by vigorous vortexing followed by the addition of 5 mL of EcoLite scintillation fluid (MP Biomedicals) and subjected to scintillation counting (5 min per sample). All assays were carried out in duplicate or better.

PglB and PglC/PglA Coupled Radioactivity Assay

Procedures for these two assays are based on previous reports (4, 27), and are described in detail in the supplementary information.

Mass Spectral Characterization of Unnatural Polyisoprenyl Diphosphate Disaccharide Substrates

Mass spectra were acquired using a 4000 QTrap mass spectrometer (Applied Biosystems / Sciex, Concord, ON, Canada). Capillary electrophoresis was performed using a Prince CE system (Prince Technologies, The Netherlands). The separation was obtained on a 90 cm length bare fused-silica capillary (365 µm OD × 50 µm ID) with CE-MS coupling using a liquid sheath-flow interface and isopropanol:methanol (2:1) as the sheath liquid. Precursor-ion scanning (m/z 79, PO4) for the dolichol sample was done in the negative-ion mode, with a collision energy of 70 V. Samples were dissolved in 3:1 chloroform:methanol and an aqueous CE buffer consisting of 10 mM ammonium acetate was used for all experiments. Results are shown in supporting information (SI Table 2).

RESULTS

The N-linked glycosylation pathway in C. jejuni comprises a series of enzymes that act at the surface of the plasma membrane on substrates anchored to the membrane via an undecaprenyl-diphosphate moiety (1, 28). In order to explore the determinants mediating enzyme interaction with these membrane-associated substrates, the polyisoprenol specificity of these enzymes was investigated. Chemoenzymatic synthesis was applied to incorporate a series of polyisoprenols varying in length, degrees of saturation, and double-bond geometry, into selected substrates involved in the C. jejuni N-linked glycosylation pathway. These polyisoprenols were isolated from natural sources, and can be grouped into three structural classes (Figure 3). The first class includes the polyisoprenol that is naturally found in the C. jejuni substrates, undecaprenol, containing 11 unsaturated isoprene units, with cis (Z) double-bond geometry at the α-isoprene unit and an array of E and Z isoprene units in the remainder of the structure. In addition to undecaprenol, this class of polyisoprenols included prenol-9 and prenol-8, which are very similar in saturation and geometry, but contain fewer isoprene units. This class of polyisoprenols allows investigation of the effect of isoprene length on enzyme activity. In the second class are solanesol and geranylgeraniol, which are shorter in length than undecaprenol, including nine and four isoprene units respectively. Importantly, these two polyisoprenols are exclusively trans (E) in their double bond geometry. These polyisoprenols enable investigation of the effect of double bond geometry on enzyme activity. In the final class are dihydroprenol-11 and the dolichols, which are respectively either the same length as undecaprenol or longer, but are both saturated at the α-isoprene unit. This final class of polyisoprenol allows the exploration of the effect of α-isoprene saturation on enzyme activity. The dolichols are particularly interesting because these constitute part of the native substrates in the eukaryotic N-linked glycosylation pathways (9).

The polyisoprenols were incorporated into substrates for the enzymes PglC, PglJ and PglB (Figure S4) using a chemoenzymatic approach. First, the polyisoprenyl-alcohols were chemically phosphorylated (22, 23) to afford the polyisoprenyl-phosphates. These polyisoprenyl-phosphates all proved to be substrates for PglC, the enzyme which transfers a Bac2,4diNAc-phosphate to the polyisoprenol-phosphate providing the first membrane-associated intermediate in the pathway. Using PglC and the glycosyltransferase PglA from the C. jejuni pathway, radiolabeled polyisoprenyl-diphosphate Bac2,4diNAc-GalNAc acceptors were enzymatically synthesized. These polyisoprenyl-diphosphate-linked disaccharides are substrates for PglJ, which then transfers an additional GalNAc to afford polyisoprenyl-diphosphate Bac2,4diNAc-GalNAc2. Additionally, the polyisoprenyl-phosphate linked disaccharide can also act as a substrate for PglB. Although the native substrate for PglB is a polyisoprenyl-diphosphate linked heptasaccharide, previous studies have shown that PglB efficiently accepts a disaccharide substrate and facilitates the transfer to a peptide substrate in vitro (7). The use of the disaccharide donor together with a truncated peptide substrate provides a reliable assay for studying PglB activity.

The efficiencies of the enzymatic syntheses of the polyisoprenyl-diphosphate linked disaccharides using a combination of PglC and PglA varied among the polyisoprenyl-phosphates, and was optimized by the use of excess UDP-sugars. Typical yields were between 20–70% for a 50 nmol scale reaction, and the disaccharide products were separated from the unreacted polyisoprenyl-phosphates using normal phase HPLC and characterized by ESI-MS (Table S2 in the Supplementary Info). The [3H] radiolabel offers a convenient means to accurately quantify both substrates and products in the low micromolar concentration range. These studies demonstrate the power of chemoenzymatic synthesis to yield complex polyisoprenyl-diphosphate-linked glycan substrates for biochemical studies.

With the purified substrates in hand, enzymatic assays were performed to determine the specificity of PglC, PglJ and PglB for each of the different polyisoprenyl-phosphate-containing intermediates. To study PglB activity, a membrane fraction from E. coli in which PglB has been overexpressed (7), was used. The assay involved monitoring the transfer of radiolabeled glycan from the organic soluble polyisoprenyl-diphosphate donor to the aqueous soluble peptide acceptor. The peptide acceptor used in this study was the hexapeptide Ac-DFNVT-(pNF)-NH2 where pNF is para-nitrophenylalanine (27). The peptide was used at saturating concentrations and the amount of the polyisoprenyl-diphosphate-linked glycan substrates was varied to afford the results shown in Figure 4C. The studies demonstrate that the transfer efficiency of disaccharide to peptide from the unnatural polyisoprenols prenol-9 and prenol-8 is very similar to that of the native undecaprenol. These data suggest that PglB is not affected by the presence of a slightly truncated form of undecaprenol. The presence of dihydroprenol-11 and dolichols, both of which include a saturated terminal α-isoprene unit, results in a significant decrease in PglB activity suggesting that PglB is highly sensitive to the degree of saturation proximal to the site of enzymatic action. Finally, the solanesol-linked substrate demonstrates very low activity with PglB, suggesting that the all-trans geometry disrupts interactions between PglB and the polyisoprenol moiety. Both the KM and Vmax seem to be affected in the case of these non-natural substrates (Table 1).

Figure 4.

Figure 4

Open in a new tab

Comparison of enzyme reaction rates: A) Phosphoglycosyltransferases PglC and PglA coupled, B) Glycosyltransferase PglJ, and C) Oligosaccharyl transferase PglB. Radiolabeled disaccharide donor for PglB and acceptor for PglJ refers to polyisoprenyl-PP-Bac2,4diNAc-[3H]GalNAc. All assays were carried out in duplicate or better with representative data set shown.

Table 1.

Comparison of polyisoprenyl substrate kinetic parameters with Pgl enzymes. PglC and PglJ initial rates comparisons are measured at sub-saturating substrate concentrations of 60 µM polyisoprenyl-phosphate and 0.5 µM polyisoprenyl-PP-Bac2,4diNAc-GalNAc respectively. PglB KM values refer to that of the polyisoprenyl substrate at 100 µM saturating peptide acceptor concentration, KM(peptide) = 1.21 µM. n.d. = not determined.

Polyisoprene Common Name Geometry PglC Initial Rate nM/min PglJ Initial Rate nM/min PglB Vmax nM/min PglB KM µM PglB Vmax/KM min−1 × 103
Prenol-11 Undecaprenol trans 3, cis 7 210.6 ± 18.1 1.83 ± 0.20 63.2 ± 4.6 2.63 ± 0.36 24.0 ± 5.0
Prenol-9 trans 3, cis 5 214.4 ± 24.4 1.66 ± 0.11 56.0 ± 5.0 2.97 ± 0.50 18.9 ± 4.8
Prenol-8 trans 2, cis 5 183.8 ± 6.2 n.d 64.4 ± 6.9 2.77 ± 0.56 23.2 ± 7.2
Dihydroprenol-11 trans 3, cis 6 57.5 ± 5.6 0.67 ± 0.05 24.4 ± 2.0 5.87 ± 0.82 4.2 ± 0.9
Dihydroprenol-17-19 Dolichol trans 2, cis 13–15 50.6 ± 5.1 0.54 ± 0.03 16.6 ± 1.3 4.23 ± 0.53 3.9 ± 0.8
all-trans Prenol-9 Solanesol trans 8 3.8 ± 2.0 0.54 ± 0.07 8.5 ± 0.8 4.06 ± 0.68 2.1 ± 0.6
all-trans Prenol-4 Geranylgeraniol trans 3 0.6 ± 1.0 n.d n.d. n.d. n.d.
Open in a new tab

Similar to PglB, PglC was also used as a membrane fraction that yielded greater stability and activity relative to the purified form of the enzyme due to the stabilizing effect of components of the cell envelope fraction (4). UDP-6-hydroxylBac2,4diNAc was used as the glycosyl-phosphate donor since it is known to be accepted by PglC with similar efficiency as the native UDP-Bac2,4diNAc (4) and has the advantage of being more accessible by chemical synthesis. To assay for PglC activity, a previously described coupled assay was applied, which combined the activity of PglC with PglA, the next glycosyltransferase in the pathway. This coupled assay (4) was used due to the lack of radiolabeled UDP-Bac2,4diNAc to directly assay PglC activity prior to the report of UDP-Bac2,4diNAc biosynthesis using PglF, PglE, and PglD (17). A large excess of PglA was used to ensure that reaction rates reflect the rate limiting PglC step (data not shown). The assay for PglC activity monitored the transfer of radioactivity from the aqueous soluble UDP-[3H]GalNAc to the organic soluble polyisoprenyl-diphosphate substrate. Figure 4A illustrates the efficiency of product formation using each of the different polyisoprenyl-phosphates as substrates for PglC. Similar to the substrate selectivity demonstrated by PglB, PglC accepts prenol-8 and prenol-9 with equal efficiencies to the native undecaprenol-based substrate. Dolichols- and dihydroprenol-11-based substrates were accepted with much lower efficiency, suggesting that PglC is sensitive to the saturation of the α-isoprene unit. Finally, the all-trans solanesol- and geranylgeraniol-containing intermediates were extremely poor substrates for PglC suggesting that the isoprene-geometry plays a critical role in substrate recognition.

PglJ is the peripheral membrane protein that transfers the third sugar in the heptasaccharide biosynthesis. The PglJ enzyme samples used in this study were solubilized using Triton-X100 detergent and purified to homogeneity using Ni-NTA affinity chromatography as previously described (3). The purified radiolabeled disaccharide substrates used in the PglB assay above, were also used to assay PglJ activity. In this case, since the disaccharide acceptor is already radiolabeled, we measured the increase in radioactivity resulting from the transfer of an additional GalNAc residue from UDP-[3H]GalNAc. Similar to PglB and PglC, PglJ accepts the prenol-9 substrate with equal efficiencies to the native undecaprenol-based substrate (Figure 4B). The non-natural dihydroprenol-11, dolichols and solanesol-based substrates are also accepted by PglJ but with lower efficiencies compared to the corresponding undecaprenol and prenol-9 substrates.

Our studies reveal that the three enzymes studied exhibit distinct preferences for the structural features of the polyisoprenols within their target substrates. The native undecaprenol-based substrates were well accepted in all cases. Additionally, polyisoprenyl carriers of shorter length, such as prenol-9 and prenol-8, displayed similar turnover to the native substrate, indicating that these enzymes do not have stringent specificity for the number of isoprene units; the C40 substrate is virtually identical to the C55 native substrate. Although solanesol is of the same length as prenol-9, its incorporation into substrates for each of the enzymes proved highly deleterious, most likely due to the unfavorable all-trans double bond geometry. Furthermore, dihydroprenols such as dolichols, being saturated at the α-isoprene position, were also poorly accepted in comparison to the unsaturated counterparts. This clear specificity with respect to the polyisoprenyl carrier, especially towards the features at α-isoprene units, is consistent with the hypothesis that membrane-associated hydrophobic proteins interact extensively with their lipophilic substrates. Intriguingly, there was no clear correlation between the stringency of the substrate specificity and the number of predicted transmembrane domains possessed by each enzyme.

We considered the possibility that the differences observed in Vmax and KM do not actually reflect differences in catalytic efficiencies or their binding affinities to the enzymes, but rather a function of substrate solubility as well as, in the case of PglB and PglC, enzyme accessibility which is limited by the permeability of the cell membrane. However, the concentration of detergent used in the assay conditions was sufficient for fully dissolving the cell membrane, and therefore the interaction of the polyisoprenyl-linked substrates should not be influenced or limited by the membrane structure (29). Furthermore, the detergent is known to form mixed micelles with the substrate, such that each polyisoprenyl-linked substrate should be effectively solubilized (30).

DISCUSSION

The recognition of membrane-bound substrates by membrane-associated enzymes is a poorly understood phenomenon due to the difficulties associated with studying these enzymes in vitro with discrete substrate analogs. In C. jejuni, the topological diversity of enzymes from the N-linked glycosylation pathway, and our ability to chemoenzymatically synthesize and purify selected substrates that varied in the identity of the polyisoprenyl carrier, provided a unique opportunity for addressing this question.

The three enzymes for the study differed in their degree of membrane association, with PglB having 10–12 predicted transmembrane domains, PglC with one, and PglJ being only a peripheral membrane protein (Figure S4). Despite this variation in structure, we have demonstrated that the three enzymes display similar specificity toward their respective polyisoprenol-based substrates. Detailed analysis of the substrate preferences of these enzymes reveal that the double-bond geometry and α-unsaturation at the alcohol terminus of the polyisoprenols play a more important role in substrate recognition than the overall number of isoprene units (Figure 3). Even though the enzymes catalyze reactions that occur at the membrane surface, there appear to be important interactions within the hydrophobic component of the substrate that contribute to specificity.

These studies concur with a previous report on the N-linked glycosylation pathway of Saccharomyces cerevisiae, whereby polyisoprenyl specificity was found to be maintained by both Alg7 and the oligosaccharyl transferase (OT), the two eukaryotic enzymes that catalyze the analogous reactions to PglC and PglB (31). However, since Alg7 did not accept any α-unsaturated polyisoprenyl-phosphates or the all-trans solanesyl-phosphate, the corresponding OT substrates could not be accessed and therefore the study could not be pursued to the level of detail presented herein for the C. jejuni enzymes. To our knowledge, the specificity of OT towards α-unsaturated and all-trans polyisoprenols has not been previously tested in any other systems. A simplified generalization has been that α-unsaturated polyisoprenols are primarily found in bacteria and plants, while dihydropolyisoprenols are present in mammalian and yeast cells (32). Therefore, many examples of glycosyltransferases from eukaryotic and prokaryotic sources have been found to reflect this specificity towards their native type of polyisoprenyl carrier (33–35). In this case, the in vitro radioactivity-based assay allowed us to takes advantage of the versatility of a chemoenzymatic synthesis to evaluate unnatural substrates for the bacterial OT. Interestingly, all three enzymes appear capable of accepting the eukaryotic dolichol-based carrier in vitro albeit with lower efficiency.

In conclusion, the availability of purified substrates and purified enzymes of varying membrane affinity makes C. jejuni a viable model system for studying how membrane-associated enzymes interact with their lipophilic substrates. Our studies suggest that the polyisoprenyl carrier common to both eukaryotic and prokaryotic protein glycosylation plays a more specific chemical role in substrate/enzyme interactions rather than a simple physical role as a hydrophobic membrane anchor. The establishment of experimental approaches for investigating the role of the polyisoprenyl moiety in individual steps in the pgl pathway now sets the stages for studies aimed at evaluating the role of these essential biochemical components in the integrated action of the series of enzymes that ultimately results in the efficient biosynthesis of N-linked glycoproteins in prokaryotes.

Supplementary Material

1File002. SUPPORTING INFORMATION AVAILABLE.

Detailed assay protocols for PglC, PglA, and PglB; polyisoprenyl disaccharide mass spectral characterization; TMHMM, ExPASY transmembrane domains predictions; raw assay data. This material is available free of charge via the Internet at http://pubs.acs.org.

ACKNOWLEDGEMENT

We wish to thank Dr. Nelson Olivier and Dr. Jerry Troutman for critical reading of this manuscript, Dr. Jerry Troutman for optimization of PglJ assay conditions, and Dr. Kerney Jebrell Glover for helpful discussion.

This work was supported by the National Institute of Health Grants GM68692 and GM39334 to B.I.

ABBREVIATIONS

pgl

protein glycosylation

Und

undecaprenol

PP

diphosphate

GalNAc

N-acetyl galactosamine

Glc

glucose

Bac2,4diNAc

2,4-diacetamido bacillosamine

UDP

uridine diphosphate

OT

oligosaccharyl transferase

PSUP

pure solvent upper phase.

References

  • 1.Szymanski CM, Yao R, Ewing CP, Trust TJ, Guerry P. Evidence for a system of general protein glycosylation in Campylobacter jejuni. Mol Microbiol. 1999;32:1022–1030. doi: 10.1046/j.1365-2958.1999.01415.x. [DOI] [PubMed] [Google Scholar]
  • 2.Weerapana E, Imperiali B. Asparagine-linked protein glycosylation: From eukaryotic to prokaryotic systems. Glycobiology. 2006;16:91–101. doi: 10.1093/glycob/cwj099. [DOI] [PubMed] [Google Scholar]
  • 3.Glover KJ, Weerapana E, Imperiali B. In vitro assembly of the undecaprenylpyrophosphate-linked heptasaccharide for prokaryotic N-linked glycosylation. Proc Natl Acad Sci U S A. 2005;102:14255–14259. doi: 10.1073/pnas.0507311102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Glover KJ, Weerapana E, Chen MM, Imperiali B. Direct biochemical evidence for the utilization of UDP-bacillosamine by PglC, an essential glycosyl-1-phosphate transferase in the Campylobacter jejuni N-linked glycosylation pathway. Biochemistry. 2006;45:5343–5350. doi: 10.1021/bi0602056. [DOI] [PubMed] [Google Scholar]
  • 5.Young NM, Brisson JR, Kelly J, Watson DC, Tessier L, Lanthier PH, Jarrell HC, Cadotte N, St Michael F, Aberg E, Szymanski CM. Structure of the N-linked glycan present on multiple glycoproteins in the Gram-negative bacterium, Campylobacter jejuni. J Biol Chem. 2002;277:42530–42539. doi: 10.1074/jbc.M206114200. [DOI] [PubMed] [Google Scholar]
  • 6.Alaimo C, Catrein I, Morf L, Marolda CL, Callewaert N, Valvano MA, Feldman MF, Aebi M. Two distinct but interchangeable mechanisms for flipping of lipid-linked oligosaccharides. Embo J. 2006;25:967–976. doi: 10.1038/sj.emboj.7601024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Glover KJ, Weerapana E, Numao S, Imperiali B. Chemoenzymatic synthesis of glycopeptides with PglB, a bacterial oligosaccharyl transferase from Campylobacter jejuni. Chem Biol. 2005;12:1311–1315. doi: 10.1016/j.chembiol.2005.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kowarik M, Young NM, Numao S, Schulz BL, Hug I, Callewaert N, Mills DC, Watson DC, Hernandez M, Kelly JF, Wacker M, Aebi M. Definition of the bacterial N-glycosylation site consensus sequence. Embo J. 2006;25:1957–1966. doi: 10.1038/sj.emboj.7601087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Lennarz WJ. Lipid linked sugars in glycoprotein synthesis. Science. 1975;188:986–991. doi: 10.1126/science.167438. [DOI] [PubMed] [Google Scholar]
  • 10.Wacker M, Linton D, Hitchen PG, Nita-Lazar M, Haslam SM, North SJ, Panico M, Morris HR, Dell A, Wren BW, Aebi M. N-linked glycosylation in Campylobacter jejuni and its functional transfer into E. coli. Science. 2002;298:1790–1793. doi: 10.1126/science.298.5599.1790. [DOI] [PubMed] [Google Scholar]
  • 11.Weerapana E, Glover KJ, Chen MM, Imperiali B. Investigating bacterial N-linked glycosylation: synthesis and glycosyl acceptor activity of the undecaprenyl pyrophosphate-linked bacillosamine. J Am Chem Soc. 2005;127:13766–13767. doi: 10.1021/ja054265v. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Wassenaar TM, Blaser MJ. Pathophysiology of Campylobacter jejuni infections of humans. Microbes Infect 1. 1999:1023–1033. doi: 10.1016/s1286-4579(99)80520-6. [DOI] [PubMed] [Google Scholar]
  • 13.Nachamkin I. Chronic effects of Campylobacter infection. Microbes Infect. 2002;4:399–403. doi: 10.1016/s1286-4579(02)01553-8. [DOI] [PubMed] [Google Scholar]
  • 14.Szymanski CM, Burr DH, Guerry P. Campylobacter protein glycosylation affects host cell interactions. Infect Immun. 2002;70:2242–2244. doi: 10.1128/IAI.70.4.2242-2244.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kelly J, Jarrell H, Millar L, Tessier L, Fiori LM, Lau PC, Allan B, Szymanski CM. Biosynthesis of the N-linked glycan in Campylobacter jejuni and addition onto protein through block transfer. J Bacteriol. 2006;188:2427–2434. doi: 10.1128/JB.188.7.2427-2434.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Wacker M, Feldman MF, Callewaert N, Kowarik M, Clarke BR, Pohl NL, Hernandez M, Vines ED, Valvano MA, Whitfield C, Aebi M. Substrate specificity of bacterial oligosaccharyltransferase suggests a common transfer mechanism for the bacterial and eukaryotic systems. Proc Natl Acad Sci U S A. 2006;103:7088–7093. doi: 10.1073/pnas.0509207103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Olivier NB, Chen MM, Behr JR, Imperiali B. In Vitro Biosynthesis of UDP-N,N'-Diacetylbacillosamine by Enzymes of the Campylobacter jejuni General Protein Glycosylation System. Biochemistry. 2006;45:13659–13669. doi: 10.1021/bi061456h. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Wellburn AR, Hemming FW. The occurrence and seasonal distribution of higher isoprenoid alcohols in the plant kingdom. Phytochemistry. 1966;5:969–975. [Google Scholar]
  • 19.Chojnacki T, Jankowski W, Mankowski T, Sasak W. Preparative separation of naturally occurring mixtures of polyprenols on hydroxyalkoxypropyl-Sephadex. Anal Biochem. 1975;69:114–119. doi: 10.1016/0003-2697(75)90572-2. [DOI] [PubMed] [Google Scholar]
  • 20.Mankowski T, Jankowski W, Chojnacki T, Franke P. C55-Dolichol: occurrence in pig liver and preparation by hydrogenation of plant undecaprenol. Biochemistry. 1976;15:2125–2130. doi: 10.1021/bi00655a015. [DOI] [PubMed] [Google Scholar]
  • 21.Imperiali B, Zimmerman JW. Synthesis of Dolichols Via Asymmetric Hydrogenation of Plant Polyprenols. Tetrahedron Letters. 1988;29:5343–5344. [Google Scholar]
  • 22.Branch CL, Burton G, Moss SF. An expedient synthesis of allylic polyprenyl phosphates. Synthetic Communications. 1999;29:2639–2644. [Google Scholar]
  • 23.Ye XY, Lo MC, Brunner L, Walker D, Kahne D, Walker S. Better substrates for bacterial transglycosylases. J Am Chem Soc. 2001;123:3155–3156. doi: 10.1021/ja010028q. [DOI] [PubMed] [Google Scholar]
  • 24.Danilov LL, Druzhinina TN, Kalinchuk NA, Maltsev SD, Shibaev VN. Polyprenyl phosphates: synthesis and structure-activity relationship for a biosynthetic system of Salmonella anatum O-specific polysaccharide. Chem Phys Lipids. 1989;51:191–203. doi: 10.1016/0009-3084(89)90006-6. [DOI] [PubMed] [Google Scholar]
  • 25.Danilov LL, Chojnacki T. A Simple Procedure for Preparing Dolichyl Monophosphate by the Use of Poc13. Febs Letters. 1981;131:310–312. [Google Scholar]
  • 26.Osborn JM, Cynkin MA, Gilber JM, Muller L, Singh M. Synthesis of bacterial O-antigens. Methods Enzymol. 1972;28:583–601. [Google Scholar]
  • 27.Chen MM, Glover KJ, Imperiali B. From Peptide to Protein: Comparative Analysis of the Substrate Specificity of N-Linked Glycosylation in C. jejuni. Biochemistry. 2007;46:5579–5585. doi: 10.1021/bi602633n. [DOI] [PubMed] [Google Scholar]
  • 28.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;55:1695–1703. doi: 10.1111/j.1365-2958.2005.04519.x. [DOI] [PubMed] [Google Scholar]
  • 29.Bergman A, Mankowski T, Chojnacki T, De Luca LM, Peterson E, Dallner G. Glycosyl transfer from nucleotide sugars to C85- and C55-polyprenyl and retinyl phosphates by microsomal subfractions and Golgi membranes of rat liver. Biochem J. 1978;172:123–127. doi: 10.1042/bj1720123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Low P, Peterson E, Mizuno M, Takigawa T, Chojnacki T, Dallner G. Effectivity of dolichyl phosphates with different chain lengths as acceptors of nucleotide activated sugars. Biosci Rep. 1986;6:677–683. doi: 10.1007/BF01114763. [DOI] [PubMed] [Google Scholar]
  • 31.Palamarczyk G, Lehle L, Mankowski T, Chojnacki T, Tanner W. Specificity of solubilized yeast glycosyl transferases for polyprenyl derivatives. Eur J Biochem. 1980;105:517–523. doi: 10.1111/j.1432-1033.1980.tb04527.x. [DOI] [PubMed] [Google Scholar]
  • 32.Swiezewska E, Danikiewicz W. Polyisoprenoids: structure, biosynthesis and function. Prog Lipid Res. 2005;44:235–258. doi: 10.1016/j.plipres.2005.05.002. [DOI] [PubMed] [Google Scholar]
  • 33.Rush JS, Shelling JG, Zingg NS, Ray PH, Waechter CJ. Mannosylphosphoryldolichol-mediated reactions in oligosaccharide-P-P-dolichol biosynthesis. Recognition of the saturated alpha-isoprene unit of the mannosyl donor by pig brain mannosyltransferases. J Biol Chem. 1993;268:13110–13117. [PubMed] [Google Scholar]
  • 34.Mankowski T, Sasak W, Janczura E, Chojnacki T. Specificity of polyprenyl phosphates in the in vitro formation of lipid-linked sugars. Arch Biochem Biophys. 1977;181:393–401. doi: 10.1016/0003-9861(77)90244-2. [DOI] [PubMed] [Google Scholar]
  • 35.Mankowski T, Sasak W, Chojnacki T. Hydrogenated polyprenol phosphates - exogenous lipid acceptors of glucose from UDP glucose in rat liver microsomes. Biochem Biophys Res Commun. 1975;65:1292–1297. doi: 10.1016/s0006-291x(75)80370-6. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1File002. SUPPORTING INFORMATION AVAILABLE.

Detailed assay protocols for PglC, PglA, and PglB; polyisoprenyl disaccharide mass spectral characterization; TMHMM, ExPASY transmembrane domains predictions; raw assay data. This material is available free of charge via the Internet at http://pubs.acs.org.

NIHMS62979-supplement-1File002.pdf (319.4KB, pdf)

RESOURCES