The long-term goal of this research is to control activity of the NmW capsule polymerase for production of well-defined carbohydrates for glycoconjugate vaccines. This work describes the application of facile, high-throughput assay methods to advance this goal. Results described here are the culmination of 55 individual experiments in which two or three replicates were performed.
Differences in reactivity observed between UDP-Glo and CMP-Glo assays
In efforts to determine the optimal conditions to perform the enzyme reactions using these kits, a series of reactions (containing CMP-NeuNAc, UDP-Gal, DTT, hydrolyzed serogroup W polysaccharide acceptor, with or without enzyme) were performed in which the only component varied was the amount of enzyme. The results for the UDP-Glo assay (Additional file 2: Fig. S2A) show increase in activity as the amount of enzyme increases (max. with 1000 ng). The results for the CMP-Glo assay (Additional file 2: Fig. S2B), a measure of sialyltransferase activity, indicate a bell-shaped activity curve (max. with 50 ng). This was an unexpected result as it was assumed that the same enzyme concentration would be used for both assay kits. However, these results suggested that the activities of the two catalytic domains were not tightly correlated under these conditions. Nevertheless, the enzyme amount selected for further studies using the CMP-Glo assay was 50 ng. vs. 750 ng of enzyme to be used in the UDP-Glo assay because the luminescence output was comparable.
CMP-Glo kinetic assays using hydrolyzed serogroup W acceptor
With knowledge of how much serogroup W enzyme to use, both the optimal amount of nucleotide donors and the linearity of the enzyme reaction were investigated. The optimal amount of luminescence was obtained with 80 µM CMP-NeuNAc and UDP-Gal in both kits. In addition, the enzymatic reactions were found to be linear over 10 min (Additional file 3: Fig. S3). Initially, kinetic measurements used only the CMP-Glo assay. For all kinetic assays, one component (either UDP-Gal, CMP-NeuNAc or hydrolyzed acceptor) was varied while all other components were held constant. When both nucleotide donor sugars were constant (at 80 µM each) and the amount of hydrolyzed serogroup W acceptor was varied (0–2000 µg/mL), a Km value of 629.2 ± 101.4 µg/mL and a Vmax of 0.8965 ± 0.05823 µM/min (Fig. 1A) were obtained. In kinetic studies with varied CMP-NeuNAc (0–80 µM) and constant acceptor, a Km and Vmax values were obtained 13.84 ± 9.675 µM and 0.6205 ± 0.1331 µM/min (Fig. 1B). This data set showed more variability as evidenced by the error bars and reduced R2 value. The related standard curves show the data is reliable (R2 > 0.95) (Fig. 1C, D).
Hydrolyzed serogroup W polysaccharide acceptor contains mostly sialylated material
Because of the continued variability in the data, further confirmation that the change in luminescence observed was enzyme-mediated was needed. A series of reactions in the absence of selected components was performed using both bioluminescence kits. Galactosyltransferase activity (as observed using UDP-Glo) was seen only in the presence of all components as expected (Fig. 2A). The results of monitoring sialyltransferase activity (using CMP-Glo) were unexpected. There was an enzyme-mediated increase in activity in the absence of UDP-Gal (Fig. 2B). To gain more understanding into this finding, the products of enzymatic elongation of DMB-labeled hydrolyzed acceptor by the serogroup W capsule polymerase was visualized by anion exchange HPLC-FD. The goal was to observe whether there was any change in the chromatogram in the presence of the capsule polymerase, acceptor and with no UDP-Gal present or with no CMP-NeuNAc. As reported previously by Romanow et al., there are signature peak retention time shifts observed in elongated fluorescent products [4]. Decreased retention time indicates addition of galactose (due to the decrease in polarity by addition of the neutral sugar) and increased retention time indicates addition of sialic acid. Our data suggests that the hydrolyzed acceptor being used was primarily galactosylated (Fig. 2C, D). In the absence of CMP-NeuNAc and the presence of UDP-Gal, there is only a shift of one peak, and this is towards decreased retention time. In contrast, when CMP-NeuNAc is included and UDP-Gal omitted, there is a shift of nearly all remaining peaks towards increased retention time suggesting sialylation. At this point, it was unclear whether there was preferential hydrolysis of the polysaccharide or whether there was preferential labeling during incubation with DMB [22]. The DMB dye will only label free reducing end sialic acids so this phenomenon may influence the products that are visualized. We transitioned to a well-defined oligosaccharide which is a known substrate of the enzyme: a trimer of α, 2–8 linked sialic acid [4, 5].
CMP-Glo assay optimization with sialic acid trimer
While the UDP-Glo kit includes commercially available ultrapure UDP-Gal (essential in avoiding high background rates) there is no commercially available ultrapure CMP-NeuNAc. Our source of CMP-NeuNAc was the highest purity commercially available, [guaranteed 97% by Nacalai-Tesque and verified by HPLC analysis (not shown)] yet this seemingly small 3% impurity was having a large effect on results because of the sensitivity of the assay. To circumvent this, CMP-NeuNAc solutions were pre-treated with alkaline phosphatase (AP) (Additional file 4: Fig. S4A, Additional file 5). This enzyme removes phosphoryl groups from nucleotide mono- and diphosphates [24]. There were decreased levels of background CMP after this pre-treatment. DP3 trimer was also subjected to AP treatment with no change observed (Additional file 4: Fig. S4B). Despite this, there was still a considerable amount of unexplainable background luminescence (results not shown). The decision was made to focus solely on the UDP-Glo assay for subsequent studies with DP3 acceptor and continue AP pre-treatment of CMP-NeuNAc because better luminescence was observed (Additional file 6: Fig. S5A, B).
UDP-Glo assay optimization with sialic acid trimer
Similar optimization assays were performed with sialic acid trimer using the UDP-Glo system. The trends mirrored those observed with hydrolyzed acceptor. Namely, there was an increase in activity with increasing levels of enzyme present in the reaction (Fig. 3A). The highest signal was seen with 4 mM DP3 as an acceptor and there was very little background observed in the control reactions (Fig. 3B). Results for the optimum amount of nucleotide donor sugar to use and the time course of the reaction were like our previous observations with hydrolyzed serogroup W sugar. The optimal luminescence was obtained with 80 µM CMP-NeuNAc and UDP-Gal (Fig. 3C) and the enzymatic reaction was found to be linear over 10 min (Additional file 7: Fig. S6).