Identification of ectonucleotide pyrophosphatase/phosphodiesterase 3 (ENPP3) as a regulator of N-acetylglucosaminyltransferase GnT-IX (GnT-Vb)

Abstract

Our previous studies on a β1,6-N-acetylglucosaminyltransferase, GnT-IX (GnT-Vb), a homolog of GnT-V, indicated that the enzyme has a broad GlcNAc transfer activity toward N-linked and O-mannosyl glycan core structures and that its brain-specific gene expression is regulated by epigenetic histone modifications. In this study, we demonstrate the existence of an endogenous inhibitory factor for GnT-IX that functions as a key regulator for GnT-IX enzymatic activity in Neuro2a (N2a) cells. We purified this factor from N2a cells and found that it is identical to ectonucleotide pyrophosphatase/phosphodiesterase 3 (ENPP3), as evidenced by mass spectrometry and by the knockdown and overexpression of ENPP3 in cultured cells. Kinetic analyses revealed that the mechanism responsible for the inhibition of GnT-IX caused by ENPP3 is the ENPP3-mediated hydrolysis of the nucleotide sugar donor substrate, UDP-GlcNAc, with the resulting generation of UMP, a potent and competitive inhibitor of GnT-IX. Indeed, ENPP3 knockdown cells had significantly increased levels of intracellular nucleotide sugars and displayed changes in the total cellular glycosylation profile. In addition to chaperones or other known regulators of glycosyltransferases, the ENPP3-mediated hydrolysis of nucleotide sugars would have widespread and significant impacts on glycosyltransferase activities and would be responsible for altering the total cellular glycosylation profile and modulating cellular functions.

Keywords: Glycobiology; Glycosylation; Glycosyltransferases; Nucleoside Nucleotide Metabolism; Phosphodiesterases.

FIGURE 1.

Identification of GnT-IX-inhibitory activity in N2a cells. Effects of N2a cell lysate on GnT-IX (A) and GnT-V (B) activities, respectively. The lysate of the HEK 293T cells transfected with each construct was used as the enzyme source. Enzymatic activity was assayed in the presence of various quantities of N2a cell lysate with or without boiling for 3 min prior to assays. C, effects of de-N-glycosylation of the N2a cell lysate on its endogenous GnT-IX-inhibitory activity. N2a cell lysate (∼50 μg) was treated with PNGase F (0.5 unit) for 37 °C overnight prior to the assay and then assessed for its endogenous GnT-IX-inhibitory activity. All data are presented as the mean of duplicate measurements. Act., activity.

FIGURE 2.

Purification and identification of ENPP3 as an inhibitory factor for GnT-IX in N2a cells. A, schematic representation of the protocol for the purification of the GnT-IX-inhibitory factor from N2a cells. B, SDS-PAGE of the de-N-glycosylated purified proteins, which were then subjected to mass spectrometry analysis for protein identification. A 1:20 aliquot of the final purified protein fraction was pretreated with PNGase F (0.5 unit) for 37 °C overnight and then subjected to SDS-PAGE under reducing conditions followed by silver staining. Protein identification was carried out by LC-electrospray ionization-ion trap-MS/MS analysis on the in-gel digested proteins. MW, molecular weight markers. *, bands representing contaminating keratins.

FIGURE 3.

Confirmation of ENPP3-mediated GnT-IX inhibition. A–C, endogenous ENPP3 in N2a cells was knocked down (KD) by introducing siRNAs. siControl, siRNA negative control duplex; siENPP3_1 and 2, specific Enpp3 siRNAs, MSS277508 and MSS209893, respectively (Invitrogen). A, ENPP3 mRNA level in the KD cells. Real-time PCR was performed on the reverse-transcribed total RNA from each KD cell. The ENPP3 mRNA level was normalized to the corresponding GAPDH mRNA level, and the percentage level relative to the control is shown. B, phosphodiesterase activity in the KD cells. Enzyme activity was assayed using pNP-5′-TMP as the substrate on the cell lysate from each KD cell. C, GnT-IX-inhibitory activity in the KD cells. The GnT-IX-inhibitory activity was assayed on the cell lysate from each KD cell. The GnT-IX-transfected HEK 293T-cell lysate was used as the enzyme source. D–F, the C-terminally Myc-His-tagged intact and T205A-mutated ENPP3s were transiently transfected to HEK 293T cells. D, ENPP3 protein level in the transfected cells. Western blot analysis was performed on the cell lysate (5 μg) from each transfectant (n = 3). E, phosphodiesterase activity in the transfected cells. The activity was assayed using pNP-5′-TMP as the substrate on the cell lysate from each transfectant. F, GnT-IX-inhibitory activity in the transfected cells. The inhibitory activity of the cell lysate from each transfectant was assayed. The GnT-IX-transfected HEK 293T-cell lysate was used as the enzyme source. N.D., not detected. All data except in D are shown as the mean ± S.D. (error bars) (n = 3).

FIGURE 4.

ENPP3-mediated UDP-GlcNAc hydrolysis. The intact and T205A-mutated ENPP3 recombinant proteins were generated and partially purified (see “Experimental Procedures”). A, ion pair reversed phase HPLC profile on the hydrolysate of UDP-GlcNAc by ENPP3. The reaction was carried out by mixing 100 ng of the partially purified intact or T205A-mutated recombinant ENPP3 with 20 mm UDP-GlcNAc, and the resulting product was analyzed by ion pair reversed phase HPLC, as described under “Experimental Procedures.” The arrows indicate the elution positions of authentic compounds, UMP, UDP-GlcNAc, and UDP. B, deduced reaction of ENPP3-catalyzed hydrolysis of UDP-GlcNAc. GlcNAc-1-P, GlcNAc 1-phosphate.

FIGURE 5.

Mechanism of ENPP3-mediated GnT-IX inhibition. The partially purified recombinant GnT-IX and -V (see “Experimental Procedures”) were used as the enzyme sources. Shown is inhibition of GnT-IX (A) and GnT-V (B) activities caused by UMP and GlcNAc 1-phosphate (GlcNAc-1-P). Enzymatic activity was assayed at a fixed concentration of 2 mm UDP-GlcNAc and 25 μm acceptor substrate, GnM-S-NBD and GnGnbi-PA for GnT-IX and -V, respectively. Shown is competitive inhibition of GnT-IX (C) and GnT-V (D), respectively, caused by UMP. Data are shown in double-reciprocal plots. Enzymatic activity was assayed at a fixed concentration of 25 μm acceptor substrate as described above and various concentrations of UDP-GlcNAc. K_i values for UMP of GnT-IX and -V are indicated in the panels. Apparent K_m values for UDP-GlcNAc of GnT-IX and -V estimated from the plots were 1.5 and 7.6 mm, respectively. All data are shown as the mean of duplicate measurements.

FIGURE 6.

The level of intracellular nucleotide sugars in ENPP3-KD cells. Endogenous ENPP3 in N2a cells was knocked down by introducing siRNAs, negative control duplex, and specific Enpp3 siRNA (MSS277508). Intracellular nucleotide sugars were extracted, purified, and separated and quantitatively determined by ion pair reversed phase HPLC as described under “Experimental Procedures.” The inset exhibits the magnified figure to clearly demonstrate the changes in the minor nucleotide sugars. All data are shown as the mean ± S.D. (error bars) (n = 3). Significant difference versus control cells as determined by Student's t test is shown. *, p < 0.05.

FIGURE 7.

Comparison of HPLC elution profiles of the PA-glycans from ENPP3-KD and control cells. ENPP3-KD N2a cells were engineered as described in the legend to Fig. 6. A, normal phase HPLC profile of the PA-glycans from ENPP3-KD and control cells. Fractions F1–F6 were collected as indicated by the partitioned bars, and each collected fraction was further analyzed by reversed phase HPLC (B–G). Although many peaks were found to show changes in their expression level in the ENPP3-KD cells, we picked up 21 peaks, G1–G21, which displayed characteristic changes in the expression level and were judged to be quantitatively sufficient for subsequent MS analysis. The numbered arrowheads in A indicate the elution position of PA-glucose oligomers with the corresponding degree of polymerization. The inset in C shows an enlargement to clearly demonstrate the changes in the elution profile. Peaks eluted at less than 40 min and 15 min in A and B–G, respectively, were due to contaminating materials or by-products of the pyridylamination reaction.

FIGURE 8.

A hypothetical model of the ENPP3-mediated modulation of glycan biosynthesis. A schematic representation of the ENPP3-mediated nucleotide sugar degradation system in the regulation of glycan biosynthesis is shown. ENPP3 is predicted to regulate the cellular glycosylation process potentially through the following two mechanisms: 1) ENPP3-catalyzed hydrolysis of the intracellular nucleotide sugars that would reduce the availability of the donor substrates for glycosyltransferases and act on the enzymes in an inhibitory manner and 2) the inhibition of glycosyltransferases caused by a hydrolysate of nucleotide sugars by ENPP3, such as NMP, which can be a potent inhibitor for certain kinds of glycosyltransferases with high sensitivity (low K_i value) toward it. P, phosphate.