Glycobiology Advance Access originally published online on October 3, 2007
Glycobiology 2008 18(1):125-134; doi:10.1093/glycob/cwm109
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Unexpected Basis for Impaired Glc3Man9GlcNAc2-P-P-Dolichol Biosynthesis by Elevated Expression of GlcNAc-1-P Transferase
2 Department of Pharmacology, University of Texas Southwestern Medical Center, 6001 Forest Park Road, Dallas, TX 75390-9041, USA
1 To whom correspondence should be addressed: Tel: +214-645-6172; Fax: +214-645-6173; e-mail: mark.lehrman{at}utsouthwestern.edu
Received on August 17, 2007; revised on September 29, 2007; accepted on September 30, 2007
 |
Abstract |
|---|
GlcNAc-1-P transferase (GPT) transfers GlcNAc-1-P from UDP-GlcNAc to dolichol-P (Dol-P), forming GlcNAc-P-P-Dol to initiate synthesis of the lipid-linked oligosaccharide Glc3Man9GlcNAc2-P-P-dolichol (G3M9Gn2-P-P-Dol). Elevated expression of GPT in CHO-K1 cells is known to cause accumulation of the intermediate M5Gn2-P-P-Dol, presumably by excessively consuming Dol-P and thereby hindering Dol-P-dependent synthesis of Man-P-Dol (MPD) and Glc-P-Dol (GPD), which provide the residues for extending M5Gn2-P-P-Dol to G3M9Gn2-P-P-Dol. If so, elevated GPT expression should increase oligosaccharide-P-P-Dol quantities and reduce monosaccharide-P-Dol quantities, while requiring GPT enzymatic activity. Here we report that elevated GPT expression failed to appreciably alter the quantities of the two classes of dolichol-linked saccharide, and that neither a GPT inhibitor nor introduction of an inactivating mutation into GPT prevented M5Gn2-P-P-Dol accumulation, arguing against excessive Dol-P consumption. Unexpectedly, we noticed similarities between the phenotypes of GPT overexpressers and of CHO-K1 cells lacking Lec35p (encoded by MPDU1, the congenital disorder of glycosylation (CDG)-If locus), which is required for utilization of MPD and GPD. By compensatory overexpression of Lec35p, G3M9Gn2-P-P-Dol synthesis in GPT overexpressers could be restored. However, GPT overexpression did not affect the levels of Lec35 mRNA or protein. These results suggest that GPT may impair Lec35p function, and imply that upper as well as lower limits on GPT expression exist in normal cells. Since the mammalian GPT gene can undergo spontaneous amplification, the data also indicate a potential basis for forms of pseudo-CDG-If.
Key words: congenital disorder of glycosylation / dolichol / GPT / Lec35 / lipid-linked oligosaccharide
|
Introduction |
|---|
N-linked glycosylation is essential for eukaryotic physiology, and requires fastidious synthesis of the lipid-linked oligosaccharide (LLO) glucose3mannose9GlcNAc2-P-P-dolichol (G3M9Gn2-P-P-Dol). Synthesis of G3M9Gn2-P-P-Dol starts with dolichol-P (Dol-P), and continues by the sequential addition of two residues of GlcNAc from UDP-GlcNAc (the first of which includes transfer of phosphate and formation of the pyrophosphate linkage by the enzyme GlcNAc-1-P transferase (GPT/Alg7p), a reaction inhibited by tunicamycin (TN)) and five residues of mannose from GDP-mannose. This results in synthesis of the M5Gn2-P-P-Dol LLO intermediate, which is extended by the addition of four residues of mannose from mannose-P-Dol ((MPD), synthesized from Dol-P and GDP-mannose) and three residues of glucose from glucose-P-Dol ((GPD), synthesized from Dol-P and UDP-glucose), to yield G3M9Gn2-P-P-Dol. N-linked glycosylation in eukaryotes involves cotranslational transfer of glycan from G3M9Gn2-P-P-Dol to asparaginyl residues in Asn-Xaa-Ser/Thr sequons of nascent polypeptides as they enter the endoplasmic reticulum (ER) lumen (Hubbard and Ivatt 1981
; Kornfeld and Kornfeld 1985), a reaction carried out by oligosaccharyltransferase.
Regulation of GPT/Alg7p expression may be particularly important because both diminished and elevated levels are detrimental. Inhibition of GPT by TN blocks LLO synthesis and N-glycosylation (Elbein 1987), and ablation of the GPT genes in Saccharomyces cerevisiae (Kukuruzinska and Robbins 1987), Schizosaccharomyces pombe (Zou et al. 1995), and mouse cells (Marek et al. 1999) results in lethality. On the other hand, elevated expression of GPT by spontaneous gene amplification (Waldman et al. 1987) or transfection (Zhu et al. 1992; Datta and Lehrman 1993) causes M5Gn2-P-P-Dol to accumulate with concordant decreases of G3M9Gn2-P-P-Dol. These results could be explained if excessive consumption of Dol-P by GPT competitively hindered synthesis of MPD and GPD. This predicts that elevating GPT expression should: (i) result in much more M5Gn2-P-P-Dol than the amount of G3M9Gn2-P-P-Dol which occurs in normal cells (ii) greatly reduce quantities of MPD and GPD, and (iii) have no effect on dolichol-linked saccharides if GPT's catalytic activity is blocked. In CHO-K1 cells overexpressing GPT due to gene amplification, identification of dolichol conjugates by [3H]mevalonate labeling showed that MPD and GPD were reduced by about half, and LLOs levels were increased by a factor of 
1.5, compared to parental cells (Rosenwald et al. 1990, Table 4A). While these differences support the idea that GPT, MPD synthase, and GPD synthase all compete for a common pool of Dol-P, the mild extents of the changes leave some doubt as to whether limitation for MPD and GPD was the cause of M5Gn2-P-P-Dol accumulation.
Here, we used a quantitative nonradioactive technique, fluorophore-assisted carbohydrate electrophoresis (FACE), to assess LLOs, MPD, and GPD. We report that M5Gn2-P-P-Dol accumulation in CHO-K1 cells with overexpressed GPT is inconsistent with a model involving excessive Dol-P consumption. Unexpectedly, our results suggest that elevated GPT blocks the function of Lec35/MPDU1, a metazoan locus that encodes a protein (Lec35p) required for utilization of MPD and GPD (Anand et al. 2001). This inhibition infers the need for upper as well as lower limits on GPT expression, and suggests that inappropriate overexpression of GPT might underlie some unclassified forms of congenital disorder of glycosylation (CDG) Type I (Jaeken and Matthijs 2001; Freeze and Aebi 2005), a family of LLO dysfunction diseases, by causing accumulation of M5Gn2-P-P-Dol.
|
Results |
|---|
M5Gn2-P-P-Dol accumulation in GPT overexpressers does not correlate with an appreciable overall LLO increase, or losses of MPD or GPD, and does not require GPT catalytic activity
LLO, MPD, and GPD levels were analyzed by FACE (Gao and Lehrman 2006) to avoid potential complications associated with radioactive precursor methods (Lehrman 2007). Dolichol-linked saccharides were recovered by organic extraction and then separated into monophosphoryl (MPD and GPD) and diphosphoryl (LLO) fractions by ion-exchange (DEAE-cellulose) chromatography. After cleavage of the phosphate bonds, the reducing termini of the released saccharides were coupled to FACE fluorophores, separated by high-voltage polyacrylamide gel electrophoresis, and detected with a charge-coupled device imager by exposure to ultraviolet light.
GPT expression is proportional to TN resistance, which is primarily due to buffering of TN by the enzyme rather than a compensatory increase in enzymatic activity (Zhu et al. 1992; Dan et al. 1996). Figure 1, panel A, shows FACE-LLO analyses of parental CHO-K1 cells, a GPT transfectant (Tn-10) isolated previously (Zhu et al. 1992) with resistance to 16 µg/mL TN, and a MPD synthase-deficient Lec15 mutant (Camp et al. 1993) known to accumulate M5Gn2-P-P-Dol. As expected, G3M9Gn2-P-P-Dol was the predominant LLO in parental cells, while like Lec15, the Tn-10 cells exhibited accumulation of M5Gn2-P-P-Dol. Additional GPT transfectants were isolated in this study, and G3M9Gn2-P-P-Dol synthesis was impaired in proportion to their resistance to TN (panel B). Since conversion of Gn-P-P-Dol to M5Gn2-P-P-Dol requires nucleotide sugars, but not monosaaccharide-P-Dols, excessive consumption of Dol-P by GPT should result in formation of unusually large amounts of M5Gn2-P-P-Dol. Yet, quantities of this LLO in the highest GPT expressers (for example, Tn-10 in panel A and the 16 µg/mL TN survivor in panel B) were increased only approximately twofold over the quantity of G3M9Gn2-P-P-Dol in parental cells. Since a comparable increase was also seen in the Lec15 mutant (panel A), the increase of M5Gn2-P-P-Dol over G3M9Gn2-P-P-Dol in GPT overexpressers was most likely an indirect consequence of the LLO assembly defect.
|
The levels of Gn-P-P-Dol (the product of GPT) were examined directly with a FACE monosaccharide profiling gel (panel C). Although its presence in CHO-K1 cells was sensitive to TN as anticipated, Gn-P-P-Dol was unaffected by overexpression of GPT. Chitobiosyl-AMAC derived from Gn2-P-P-Dol can also be assessed on monosaccharide profiling gels, but it was undetectable in this experiment (data not shown). To address the possibility that excess Gn-P-P-Dol might have been converted to large amounts of an LLO intermediate smaller than M5Gn2-P-P-Dol, and therefore not visible on FACE oligosaccharide profiling gels run under typical conditions (such as in panels A and B), CHO-K1 and Tn-10 LLOs were reexamined on a FACE gel displaying all materials above the dye front (consisting of unreacted ANDS). As shown in panel D, G3M9Gn2-P-P-Dol was the predominant TN-sensitive LLO detected in CHO-K1 cells as expected. While some smaller intermediates, likely M3–4Gn2-P-P-Dol, were detected in Tn-10 cells in addition to M5Gn2-P-P-Dol, in aggregate their amounts did not appear to greatly exceed the LLO amount in CHO-K1. Therefore, there was no evidence of excessive synthesis of Gn-P-P-Dol resulting from GPT overexpression. In panel D, several rapidly migrating TN-resistant species are visible in the CHO-K1 lanes, yet absent in the Tn-10 lanes. Their identities remain to be determined, but it is plausible that their synthesis is sensitive to overexpression of GPT for the same reason that G3M9Gn2-P-P-Dol synthesis is sensitive.
Monosaccharide-P-Dol analyses by FACE were validated by comparing the MPD synthase-deficient Lec15 mutant line with Lec35 mutant cells, which also accumulate Man5GlcNAc2-P-P-Dol but do not have a defect in MPD synthesis (Zeng and Lehrman 1990). In the Lec35 mutant, MPD and GPD were readily detected (Figure 1, panel E). By comparison, in Lec15 only GPD was present at appreciable levels. Using this method, a GPT overexpresser (Tn-10) which accumulated M5Gn2-P-P-Dol had MPD and GPD levels comparable to those in parental CHO-K1 cells and the Lec35 mutant (panel F).
Consequently, GPT overexpression did not appear to result in appreciable increases of total LLO, or decreases of MPD or GPD, arguing against excessive consumption of Dol-P by GPT. We reasoned further that if M5Gn2-P-P-Dol accumulated due to depletion of the Dol-P pool by GPT, titration with TN should restore synthesis of G3M9Gn2-P-P-Dol. To inhibit GPT in cells, 2 µg/mL TN is sufficient (see below). However, this concentration and even higher concentrations of TN sufficient to diminish M5Gn2-P-P-Dol did not restore G3M9Gn2-P-P-Dol synthesis (Figure 2, panel A). Moreover (panel B), elevated expression of GPTArg303Lys, a mutant lacking catalytic activity but able to bind TN (Dan et al. 1996), caused accumulation of M5Gn2-P-P-Dol (although at the level of expression attained, the block was not complete with some M9Gn2-P-P-Dol and G3M9Gn2-P-P-Dol still produced). Taken together, the data of Figures 1 and 2 indicate that GPT overexpression does not cause M5Gn2-P-P-Dol accumulation by depleting MPD or GPD.
|
|
M5Gn2-P-P-Dol accumulation in GPT overexpressers resembles that of Lec35 mutants |
|---|
 TopAbstractIntroductionResults M5Gn2-P-P-Dol accumulation in... Compensatory overexpression of...Elevated expression of GPT...DiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
In Lec15 mutants, approximately 20% of the M5Gn2-P-P-Dol is extended in a GPD-requiring process to G3M5Gn2-P-P-Dol (Anand et al. 2001
). In contrast, very little of this LLO intermediate was detectable in a GPT overexpresser (Zhu et al. 1992), suggesting that both the MPD- and GPD-requiring reactions were impaired. This phenotype might be predicted for a mechanism excessively consuming Dol-P, but it was also highly reminiscent of the phenotype of Lec35 mutants, which accumulate M5Gn2-P-P-Dol because Lec35p is required in mammalian cells for utilization of both MPD and GPD.
To explore further the similarities between the LLO phenotypes of GPT overexpressers and Lec35 mutants, we took advantage of the partial in vitro correction of the Lec35 mutant phenotype which occurs upon preparation of microsomal membranes, but not upon plasma membrane permeabilization with streptolysin-O (SLO) (Figure 3, panels A, B, D, and E, and (Anand et al. 2001)). In contrast, Lec15 mutants fail to mannosylate M5Gn2-P-P-Dol in either system (Anand et al. 2001). As shown in panels C and F, the M5Gn2-P-P-Dol accumulation defect of GPT overexpressers was partially corrected in microsomal membranes, but retained after permeabilization with SLO, highly similar to the Lec35 mutant result. The failure to extend M5Gn2-P-P-Dol in SLO-treated Lec35 mutants and GPT overexpressers was not due to a generalized inefficiency of the SLO system, because conversion of LLO intermediates to G3M9Gn2-P-P-Dol with parental CHO-K1 cells was actually more effective with the SLO system (panel A) than with microsomes (panel B). By providing additional support for the similarity between GPT overexpressers and Lec35 mutants, these results raised the unexpected possibility that elevated GPT expression interfered with Lec35p.
|
|
Compensatory overexpression of Lec35p restores synthesis of G3M9Gn2-P-P-Dol in GPT overexpressers |
|---|
|
TopAbstractIntroductionResultsM5Gn2-P-P-Dol accumulation in...Compensatory overexpression of...Elevated expression of GPT...DiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Since Lec35p is required in mammalian cells for synthesis of G3M9Gn2-P-P-Dol, GPT would not be expected to impair Lec35p at the ratios of these two proteins resulting from normal endogenous expression. The ability of GPT to cause accumulation of M5Gn2-P-P-Dol was therefore hypothesized to require an abnormally high ratio of GPT to Lec35p. Consequently, we reasoned that compensatory expression of Lec35p in GPT overexpressers should restore synthesis of G3M9Gn2-P-P-Dol. To test this idea, we took advantage of the fact that the Lec35.1 mutant line used in the above experiments is characterized by a disrupted, internally deleted copy of the Lec35 gene (Anand et al. 2001
). Lec35.1 was transfected with normal copies of Lec35 under the control of a promoter suppressible by tetracycline (the "Tet-OFF" system) or the related agent, doxycycline (DOX). Thus, in these transfectants all Lec35 activity, and consequently all ability to extend M5Gn2-P-P-Dol to G3M9Gn2-P-P-Dol, is dependent upon the transfected copy of Lec35 and cannot be due to the reactivation of the endogenous deleted gene. In the absence of DOX, these cells express about 200 times the Lec35 mRNA of normal cells, but close to the normal amount in the presence of DOX (Anand et al. 2001). As shown in Figure 4 (panel A, lane 6), in the absence of exogenously expressed GPT, the limited quantity of transfected Lec35 mRNA synthesized in the presence of DOX was sufficient to restore extension of M5Gn2-P-P-Dol to G3M9Gn2-P-P-Dol.
|
Two DOX-suppressible Lec35.1 transfectants, designated 6C and 10A, were also supertransfected with cDNA encoding a normal copy of GPT. Stable transfectants were initially selected with 2 µg/mL TN (the minimum concentration needed to kill untransfected cells), and then screened with twofold increases up to 16 µg/mL TN to identify transfectants with varying expression of GPT. LLOs in selected colonies were then analyzed by FACE after culture in the absence or presence of DOX. A representative set of FACE analyses including double-transfectants resistant to 8 µg/mL TN is presented in Figure 4, panel A. DOX did not affect G3M9Gn2-P-P-Dol synthesis in parental CHO-K1 cells (lanes 1 and 2), or M5Gn2-P-P-Dol accumulation in GPT overexpressers without transfected Lec35 cDNA (lanes 3 and 4). As discussed above, prior to supertransfection with GPT, G3M9Gn2-P-P-Dol was detected in Lec35 transfectants both in the absence and presence of DOX (lanes 5 and 6). However, in a double transfectant overexpressing both GPT and Lec35, G3M9Gn2-P-P-Dol was synthesized in the absence of DOX (lane 7), while M5Gn2-P-P-Dol accumulated if Lec35 expression was suppressed by the presence of DOX (lane 8).
Figure 4 (panel B) summarizes results for parental CHO-K1 cells and the DOX-suppressible Lec35 transfectants 6C and 10A supertransfected to express GPT by different extents (discerned by relative resistance to TN). Without exogenously expressed GPT ("–" on the X-axis), G3M9Gn2-P-P-Dol was the predominant LLO for these three cell types. In 6C and 10A cells with the highest GPT expression (surviving 16 µg/mL TN), increasing Lec35 expression by omitting DOX did not prevent accumulation of M5Gn2-P-P-Dol. However, in 6C and 10A cells with intermediate expression of GPT (cells surviving 8 µg/mL TN), control of Lec35 expression (and hence the ratio of GPT to Lec35p) by DOX determined whether M5Gn2-P-P-Dol or G3M9Gn2-P-P-Dol was predominant. In other words, in cells transfected to express both GPT and Lec35p, the relative ratio of these two proteins determined whether or not MPD- and GPD-dependent reactions of LLO synthesis were completed.
Although GPT appeared to interfere with Lec35p function, there was no evidence that the reciprocal was true. First, cells expressing high levels of Lec35 mRNA (Lec35Tet-OFF cells in the absence of DOX) did not have less G3M9Gn2-P-P-Dol than CHO-K1 cells (Figure 4, panel A, compare lanes 1 and 5). Second, there were no significant differences observed for GPT activity (formation of [3H]GlcNAc-P-P-Dol from exogenously added UDP-[3H]GlcNAc in both SLO-permeabilized cells and in microsomal membrane preparations) by comparing parental CHO-K1, Lec35 mutants, and cells with high levels of Lec35 mRNA (data not shown).
|
Elevated expression of GPT does not significantly affect synthesis or stability of Lec35p |
|---|
To determine if elevated GPT impaired Lec35p by altering its synthesis or stability, antibody methods were required. Since antibodies for native rodent Lec35p are unavailable, MYC and hemagglutinin (HA) tags were fused to the amino terminus of the Lec35p coding sequence. These constructs were transfected into Lec35.1 mutants, so that any restoration of G3M9Gn2-P-P-Dol synthesis was strictly dependent upon the tagged Lec35p. The transfectants were then subjected to selection for restored synthesis of G3M9Gn2-P-P-Dol as described under Materials and methods. Both methods successfully generated functional tagged versions of Lec35p (data not shown), indicating that the free amino terminus does not serve a critical function. One MYC-Lec35p transfectant, designated A92, was chosen for further study.
A92 was supertransfected with C-terminal FLAG-tagged GPT (the C-terminal FLAG does not appreciably affect GPT function (Dan et al. 1996) in order to obtain transfectants expressing both MYC-Lec35p and GPT-FLAG, but with a relatively high amount of the latter so that the extension of M5Gn2-P-P-Dol was inhibited. Colonies were selected for resistance to TN as described under Materials and methods. Two colonies, designated A92.12 and A92.17, which expressed GPT-FLAG mRNA and protein at high levels were identified for further analysis (Figure 5, panel A (RNA blots) and panel B (immunoblot)). In contrast to A92, which synthesized G3M9Gn2-P-P-Dol (indicating functional correction by MYC-Lec35p), A92.12 and A92.17 accumulated M5Gn2-P-P-Dol in a manner similar to the Lec35.1 mutant (panel C), demonstrating that the ratio of GPT-FLAG to MYC-Lec35p was relatively high (see Figure 4). MPD and GPD levels in A92, A92.12, and A92.17 were as high as in CHO-K1 (panel D), consistent with the results of Figure 1, panel F. Importantly, in A92.12 and A92.17 both MYC-Lec35 mRNA (Figure 5, panel A) and protein (panel B) were comparable to, or greater than, levels in A92 cells. The immunoblot procedure detected an anti-MYC reactive polypeptide of 22 kDa; MYC-Lec35p has an expected size of 27 kDa, but is anticipated to migrate somewhat aberrantly due to its high degree of hydrophobicity. To verify more directly that Lec35p synthesis and stability were not strongly impaired by GPT overexpression, cells were pulsed with [35S]Translabel for 30 min to label newly-synthesized MYC-Lec35p, and then subjected to immunoprecipitation with anti-MYC antibody. This procedure detected [35S]MYC-Lec35p in A92 cells but not the Lec35.1 mutant line (panel E). As shown in panel F (0 h of chase) the amount of newly synthesized MYC-Lec35p in A92 cells was similar to that in A92.12 and A92.17, and MYC-Lec35p appeared highly stable in each cell type (up to 24 h of chase). Any perceived differences of MYC-Lec35p are attributable to experimental variation in [35S]protein load, which can be judged from the background bands. Taken together, these data suggest that elevated expression of GPT can strongly hinder Lec35p's function in G3M9Gn2-P-P-Dol production, without significantly affecting the synthesis or stability of Lec35 mRNA or protein.
|
|
Discussion |
|---|
Overexpression of GPT in CHO-K1 cells by either gene amplification or transfection of cDNA impairs G3M9Gn2-P-P-Dol synthesis, and causes M5Gn2-P-P-Dol to accumulate. The simplest explanation for this phenomenon invokes excessive consumption of Dol-P by GPT, depleting pools of MPD and GPD needed to extend M5Gn2-P-P-Dol to G3M9Gn2-P-P-Dol. A corollary of this prediction is greatly increased synthesis of GlcNAc-P-P-Dol, which should be efficiently extended to M5Gn2-P-P-Dol because nucleotide sugars but not dolichol-P sugars are required. We found, however, that overexpression of GPT sufficient to cause M5Gn2-P-P-Dol accumulation did not appreciably affect levels of MPD or GPD, did not result in significantly increased amounts of total LLO, and did not require the catalytic activity of GPT. The absence of an observed increase in Gn-P-P-Dol levels (Figure 1, panel C) also argues against potential inhibition by excess Gn-P-P-Dol of flipping activities, attenuation of which could have accounted for M5Gn2-P-P-Dol accumulation.
Instead, our data support the unexpected conclusion that overexpressed GPT impairs G3M9Gn2-P-P-Dol synthesis by interference with Lec35p, a protein required for utilization of both MPD and GPD (Anand et al. 2001). Phenotypes characteristic of Lec35 mutants were also observed for GPT overexpressers, and overexpression of Lec35p counteracted the effect of overexpression of GPT. However, GPT overexpression did not affect the synthesis or stability of Lec35p. It should be noted that our data do not rule out the possibility that GPT inhibits the function of a protein, other than Lec35p, involved in mannosylation of M5Gn2-P-P-Dol with MPD. In this scenario, overexpression of Lec35p would compensate for impaired function of the third protein by enhancing its ability to utilize MPD. However, both models are similar in that the M5Gn2-P-P-Dol accumulation is proposed to result from GPT inhibiting a key protein involved in mannosylation.
Since both Lec35p and GPT are essential for G3M9Gn2-P-P-Dol synthesis in mammals, and genetic deficiency of Lec35p causes CDG-If, it is unlikely that the inhibition of Lec35p by GPT represents a normal physiological process. Attempts to identify direct interactions between GPT and Lec35p were unsuccessful (data not shown). We speculate that they might compete for docking to a necessary scaffold or other factor, yet to be identified. Both proteins would be accommodated at normal expression levels, but upon elevation of GPT, Lec35p would be excluded with GPT acting as a dominant negative inhibitor. However, this model predicts that high levels of Lec35p should interfere with GPT function, while as explained in the text we observed no evidence for such reciprocal inhibition. The mechanism and significance of the apparent effect of GPT upon Lec35p therefore remain unclear. The results were also surprising because the expected mechanism (excessive consumption of Dol-P by elevated expression of GPT) was not supported, suggestive of a qualitative difference between endogenously and exogenously expressed GPT which in some way governs access to Dol-P.
As reviewed in the Introduction, analyses of dolichol conjugates by [3H]mevalonate labeling revealed only moderate changes in the concentrations of MPD, GPD, or LLO in GPT overexpressers (Rosenwald et al. 1990), in general agreement with FACE results reported here. A similarly moderate change for LLO levels was reported by [3H]mannose labeling (Waldman et al. 1987, Table 2B). However, two aspects of that report were discordant with our own. First, in GPT overexpressers, a 50-fold decrease in cellular [3H]MPD synthesis was reported. Second, the presence of TN restored extension of [3H]M5Gn2-P-P-Dol to [3H]G3M9Gn2-P-P-Dol (as well as restoring [3H]MPD synthesis) to normal. There are no clear explanations for these discrepancies. However, in the [3H]mannose labeling studies but not the [3H]mevalonate or FACE studies, medium with low (0.1 mM) glucose was used. We have found that in some cases low glucose conditions can significantly influence the analysis of dolichol-linked saccharides (Lehrman 2007). It is also plausible that inhibiting GPT with TN might have influenced synthesis of MPD and LLO by activating an ER stress response, which can have dramatic effects on G3M9Gn2-P-P-Dol synthesis (Shang et al. 2007).
Since a minimal level of GPT is necessary for adequate G3M9Gn2-P-P-Dol synthesis, and excess GPT causes M5Gn2-P-P-Dol to accumulate, we infer that cells may have mechanisms which govern both the lower and upper limits of GPT expression. Resistance to TN correlates with GPT expression (Zhu et al. 1992), so an estimate of the acceptable upper limit can be obtained by comparison of TN-resistant cells with parental CHO-K1 cells, which are viable in up to approximately 0.2 µg/mL TN (Zhu et al. 1992). GPT transfectants with 10-fold TN resistance have substantial accumulation of M5Gn2-P-P-Dol (2 µg/mL TN survivors, Figure 1, panel B). Transfectants with only approximately fivefold TN resistance (Datta and Lehrman 1993) accumulated M9Gn2-P-P-Dol (indicative of a GPD utilization deficiency). Thus, the acceptable upper limit may be less than approximately fivefold elevated expression.
As mentioned earlier, CHO-K1 cells resistant to toxic concentrations of TN can be isolated without prior intentional mutagenesis due to spontaneous amplification of a chromosomal region including the GPT locus, resulting in increased GPT gene expression (Scocca et al. 1988; Zhu and Lehrman 1989). Conversely, the CHO-K1 Lec35 gene contains numerous Alu repetitive elements, sequences which favor spontaneous gene disruption (Lehrman et al. 1987). The defect in the Lec35.1 line used here was traced to a deletion in an Alu-rich region of the Lec35 gene (Anand et al. 2001), which occurred spontaneously in the CHO-K1 population. At least for CHO-K1 cells in culture, cells with increased GPT or with Lec35 disruptions appear to have no growth disadvantages. Thus, is it conceivable that cells in which one of the two GPT loci had been amplified, alone or in conjunction with disruption of one of the Lec35 loci, would have LLO dysfunction while growing at a normal rate. In humans, this might be predicted to cause a nonheritable pseudo-CDG-If mosaic phenotype. These would escape detection by standard sequence-based screening of blood cells of unclassified CDG-I patients because the phenotype would not necessarily be expressed in blood cells, and sequences of GPT DNA and RNA would appear normal. A similar pseudo-CDG-If phenotype would be predicted for mutations that increased GPT promoter activity, although in such cases, genomic DNA sequencing could detect the abnormality and the disease should be heritable.
|
Materials and methods |
|---|
Reagents and cell lines
DOX was from Invitrogen, Carlsbad, CA. TN was from Sigma-Aldrich, St. Louis, MO. AMAC and ANDS were from Molecular Probes, Invitrogen. Parental CHO-K1 cells were from the American Type Culture Collection, Rockville, MD. Some CHO-K1 derived lines were prepared earlier: Lec35.1 and Lec15.2 (Camp et al. 1993
), 6C and 10A (Anand et al. 2001), and Tn-10 (Zhu et al. 1992). Unless indicated otherwise, cells were grown at 37°C in a humidified 5% CO2 atmosphere in Ham's F-12 medium (Invitrogen) supplemented with 2% fetal bovine serum and 8% calf serum (Atlanta Biologicals, Lawrenceville, GA).
Permeabilization of cells with SLO
SLO treatment was done exactly as described earlier (Anand et al. 2001). Briefly, cells were incubated with SLO solution for 4 min at 4°C. After removing the solution and rinsing with ice-cold PBS, buffer containing 50 mM K-HEPES (pH 7.2), 78 mM KCl, and 3 mM MgCl2 (prewarmed to 37°C) was added, incubating for 5 min. Incubation with nucleotide sugars for LLO synthesis was done as described (Anand et al. 2001).
Preparation and use of microsomal membranes
Microsomal membranes were prepared as described (Lehrman et al. 1988). In brief, cells were swollen for 15 min on ice in hypo-osmotic solution containing 2 mM Tris-Cl (pH 7.4), the cells were disrupted with a motorized homogenizer, debris, and heavy membrane fractions were removed by centrifugation at 3,000 x g for 10 min, and microsomes were collected by further centrifugation at 100,000 x g for 60 min. Incubation with nucleotide sugars for LLO synthesis was done as described (Zeng and Lehrman 1990).
FACE analyses of LLOs, MPD, and GPD
The methods for extraction and enrichment of dolichol-linked saccharides, cleavage of phosphate bonds, conjugation to FACE fluorophores, electrophoretic separation, and imaging have been described in detail (Gao and Lehrman 2002; Gao and Lehrman 2003; Gao and Lehrman 2006). Briefly, monosaccharides representing MPD or GPD (released from dolichol-P-monosaccharide fractions), or representing Gn-P-P-Dol (released from dolichol-P-P-monosaccharide fractions), were conjugated to AMAC and separated on borate-impregnated FACE monosaccharide profiling gels. Oligosaccharides released from dolichol-P-P-oligosaccharide fractions were conjugated to ANDS and separated on FACE oligosaccharide profiling gels. In general, cells were grown to 90% of confluence in 150 mM dishes, and 50% of the sample per dish was used per lane of the FACE gels. In any given figure panel, all cells were grown and processed similarly to allow direct comparison.
Transfection of GPT, GPT-FLAG, and GPTArg303Lys
GPT constructs were prepared previously in the pJB20 vector, which drives expression of the insert cDNA with a cytomegalovirus promoter (Slonina et al. 1993; Dan et al. 1996). Transfection was done by the calcium phosphate method (Zhu et al. 1992). This was followed by selection of colonies stably expressing GPT with 2 µg/mL TN, the minimum concentration able to efficiently eliminate untransfected cells. In some experiments, resistant colonies were screened for resistance to 4, 8, or 16 µg/mL TN to obtain different levels of GPT expression.
Construction and transfection of epitope-tagged Lec35
pTRE-Lec35 (Anand et al. 2001), bearing the hamster Lec35 cDNA sequence, was the template for PCR. For MYC-Lec35p,> an insert fragment was amplified with 5'ATACGGATCCATGGCCGGTGAGGCGGACGGAC 3' (forward, BamHI site underlined) and 5' CTGCAAGCTTCTATTGCTCCTTTTTATGTTTG 3' (reverse, HindIII site underlined). Both the PCR product and pCMV5-MYC (gift of Dr. Helen Yin, UT-Southwestern (Rozelle et al. 2000)) were digested with BamHI and HindIII and then ligated together. For HA-Lec35p, an insert fragment was amplified with 5'GGGGCCGGCCAATGGCCGGTGAGGCGGACGGAC-3' (forward, FseI site underlined) and 5'GGGGCGCGCCCTATTGCTCCTTTTTATGTTTG 3' (reverse, AscI site underlined). Both this PCR product and pCS2-HA (gift of Dr. Hongtao Yu, UT-Southwestern (Kang et al. 2006)) were digested with FseI and AscI and then ligated together.
After propagation in Escherichia coli, the resulting plasmids were transfected into Lec35.1 cells by the calcium phosphate method. Transfectants having restored synthesis of G3M9Gn2-P-P-Dol were selected for 4 days with medium containing a combination of 20 µg/mL PHA-E/1 µg/mL swainsonine (Ware and Lehrman 1996). Spurious Lec35/Lec1 double mutants, which can survive PHA-E/swainsonine (Zeng and Lehrman 1991), were eliminated by counter selection for 2 days with medium containing 10 µg/mL concanavalin A. Colonies surviving both procedures were reselected with PHA-E/swainsonine to obtain single colonies stably expressing tagged Lec35p, and then screened with mRNA blots and immunoblots to identify those with the highest expression. Restoration of G3M9Gn2-P-P-Dol synthesis in transfectants expressing each form of Lec35p was verified by [3H]mannose-labeling of cells and HPLC fractionation of the glycans. A MYC-Lec35p transfectant, designated A92, was selected for further study because in preliminary experiments, MYC-Lec35p was more easily detected than HA-Lec35p by pulse labeling and immunoprecipitation.
A92 cells were supertransfected with GPT-FLAG (Dan et al. 1996) and selected for resistance to 4 µg/mL TN. Stable expression of GPT in these tagged double transfectants was verified by immunoblotting with anti-FLAG antibody. Randomly chosen colonies were also screened with anti-MYC antibody, resulting in identification of two tagged double transfectants (designated A92.12 and A92.17), which retained MYC-Lec35p expression similar to that in A92 (Figure 5, panel B).
Northern blotting
Cells were grown to confluence in 100 mm dishes. Total RNA was harvested with the Qiagen "RNeasy" minikit system, which typically yielded 50–60 µg RNA per dish. Northern blotting was performed by loading 15 µg RNA per well of a 9-cm-long 1% agarose gel, with electrophoresis in 20 mM Na–MOPS bufffer, pH 7.0, at 60 V for 3.5 h. RNA was transferred by capillary action to an Amersham XL-Hybond membrane, then hybridized overnight with DNA probes prepared by PCR and labeled with [32P]dCTP with the Amersham Rediprime II Random Prime Labeling System. GPT probes (716 bp) were prepared from a hamster cDNA clone (Zhu et al. 1992) by PCR with the primers (forward) 5' ACTGCTTTGTGGAGGAGCAGTGTA 3' and (reverse) 5' ATAGCTCATCTCCAGTTTGCCCGT 3'. Lec35 probes (744 bp) were prepared from a hamster cDNA clone (Anand et al. 2001) by PCR with the primers (forward) 5' ATGGCCGGTGAGGCGGACGGACCG 3' and (reverse) 5' CTATTGCTCCTTTTTATGTTTGTG 3'. After washing, the membrane was exposed to Kodak O-MAT Blue film and also imaged on a Fujifilm phosphorimager for quantitative analysis.
Immunoblotting
Cells were harvested at 90% confluence in RIPA buffer (10 mM Tris–Cl (pH 7.4), 100 mM NaCl, 1 mM Na3EDTA, 0.5% Na-deoxycholate, 0.1% SDS, 1% Triton X-100), 1 mL per 100 mm dish. Lysates were kept on ice for 20 min, then clarified by centrifugation for 20 min at 20,000 x g at 4°C. Supernatants were mixed with concentrated SDS-PAGE sample buffer for a final concentration of 2% SDS, denatured for 60 min at 45°C in the presence of 100 mM β-mercaptoethanol, and subjected to SDS-PAGE. Proteins were transferred to a PVDF membrane. After incubation for 16 h at 4°C in blocking solution (5% nonfat milk, 10 mM Tris–Cl (pH 7.4), 0.15 M NaCl, and 0.2% Triton X-100), the membrane was probed with 0.3 µg/mL anti-MYC (Invitrogen, cat. # 46–0603) or 0.2 µg/mL anti-FLAG (Sigma-Aldrich, cat. # T-3165) monoclonal antibody for 1 h. After washing four times with blocking solution (omitting milk), the membrane was incubated with a 1:5000 dilution of horse radish peroxidase-conjugated antimouse IgG (Amersham Biosciences, Piscataway, NJ, cat. # NXA931), treated with ECL reagent (Amersham Biosciences, cat. # RPN2106), and detected by autoradiography.
Immunoprecipitation of MYC-Lec35p from [35S]methionine-labeled cells
Cells in 100 mm dishes, grown to 70–80% confluence, were incubated 20 min in methionine-free F12 medium. Cells were then incubated 30 min ("pulse") with methionine-free medium supplemented with 125 µCi/mL [35S]-Translabel (methionine/cysteine mixture, from MP Biomedical, Irvine, CA). In some experiments, cells were then "chased" in normal medium without [35S]-Translabel. After rinsing twice with ice-cold PBS, the cells were lysed with 1 mL RIPA buffer supplemented with protease inhibitors (Roche, Penzberg, Germany, cat. # 11836153001, 1 tablet per 10 mL) just prior to use. Lysates were gently mixed for 30 min at 4°C, then precleared by the addition of 1 µL normal mouse serum and 20 µL (packed volume) protein G-Sepharose 4B (Zymed) prewashed in RIPA buffer, then rotated for 1 h at 4°C, and centrifugated at 3,000 rpm in a refrigerated microcentrifuge. The supernatant was recovered, and added to a tube containing 20 µL (packed volume) of protein G-Sepharose 4B preloaded with anti-MYC monoclonal IgG1 (Invitrogen) by prior mixing with 1.2 µg of the antibody in 100 µL RIPA buffer. The suspension was incubated with rotation 18 h at 4°C. After recovery by centrifugation, the beads were washed twice with 500 µL RIPA buffer, and eluted sequentially with 60 µL, and then 20 µL of SDS-PAGE sample buffer containing 2% SDS and 0.1 M β-mercaptoethanol. The two elutions were pooled, heated at 45°C for 10 min, and subjected to SDS-PAGE (10% acrylamide). Gels were fixed in a solution of 50% methanol/10% glacial acetic acid, dried, and exposed to Kodak O-MAT Blue film.
Image preparation
Images for RNA blots, immunoblots, FACE gels, and autoradiography were in some cases cropped and rejoined to remove irrelevant lanes, and/or image brightness and contrast were adjusted, using tools in Microsoft Office Powerpoint 2003. However, all image components in any given figure panel were treated identically to allow direct comparison.
|
Funding |
|---|
National Institute of Health (GM38545); Welch Foundation (I-1168).
|
Conflict of interest statement |
|---|
None declared.
|
Acknowledgements |
|---|
We thank Suzanne Browne (RNA blots), Karen Cuellar (pulse-chase experiments), and Biswanath Pramanik (cell culture) for excellent technical assistance. We also thank Dr. Sharon Krag for reading and commenting on the manuscript.
|
Abbreviations |
|---|
AMAC, 2-aminoacridone; ANDS, 7-amino-1,3-naphthalenedisulfonic acid; CDG, congenital disorder of glycosylation; Dol-P, dolichol phosphate; DOX, doxycycline; FACE, fluorophore-assisted carbohydrate electrophoresis; G3M9Gn2-P-P-Dol, glucose3mannose9N-acetylglucosamine2-P-P-dolichol; GPD, glucose-P-Dol; GPT, GlcNAc-1-P transferase; HA, hemagglutinin; LLO, lipid-linked oligosaccharide; MPD, mannose-P-Dol; SLO, streptolysin-O; TN, tunicamycin
|
References |
|---|
Anand M, Rush JS, Ray S, Doucey MA, Weik J, Ware FE, Hofsteenge J, Waechter CJ, Lehrman MA. Requirement of the Lec35 gene for all known classes of monosaccharide-P-dolichol-dependent glycosyltransferase reactions in mammals. Mol Biol Cell (2001) 12:487–501.
Camp LA, Chauhan P, Farrar JD, Lehrman MA. Defective mannosylation of glycosylphosphatidylinositol in Lec35 Chinese hamster ovary cells. J Biol Chem (1993) 268:6721–6728.
Dan N, Middleton RM, Lehrman MA. Hamster UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase has multiple transmembrane spans and a critical cytosolic loop. J Biol Chem (1996) 271:30717–30725.
Datta AK, Lehrman MA. Both potential dolichol recognition sequences of hamster GlcNAc-1-phosphate transferase are necessary for normal enzyme function. J Biol Chem (1993) 268:12663–12668.
Elbein AD. Inhibitors of the biosynthesis and processing of N-linked oligosaccharide chains. Annu Rev Biochem (1987) 56:497–534.[CrossRef][Web of Science][Medline]
Freeze HH, Aebi M. Altered glycan structures: The molecular basis of congenital disorders of glycosylation. Curr Opin Struct Biol (2005) 15:490–498.[CrossRef][Web of Science][Medline]
Gao N, Lehrman MA. Analyses of dolichol pyrophosphate-linked oligosaccharides in cell cultures and tissues by fluorophore-assisted carbohydrate electrophoresis. Glycobiology (2002) 12:353–360.
Gao N, Lehrman MA. Letter to the glycoforum: Alternative sources of reagents and supplies for fluorophore-assisted carbohydrate electrophoresis (FACE). Glycobiology (2003) 13:1G–3G.
Gao N, Lehrman MA. Non-radioactive analysis of lipid-linked oligosaccharide compositions by fluorophore-assisted carbohydrate electrophoresis (FACE). Methods Enzymol (2006) 415:3–20.[Medline]
Hubbard SC, Ivatt RJ. Synthesis and processing of asparagine-linked oligosaccharides. Annu Rev Biochem (1981) 50:555–583.[CrossRef][Web of Science][Medline]
Jaeken J, Matthijs G. Congenital disorders of glycosylation. Annu Rev Genomics Human Genet (2001) 2:129–151.[CrossRef]
Kang J, Gocke C, Yu H. Phosphorylation-facilitated sumoylation of MEF2C negatively regulates its transcriptional activity. BMC Biochem (2006) 7:5.[CrossRef][Medline]
Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem (1985) 54:631–664.[CrossRef][Web of Science][Medline]
Kukuruzinska MA, Robbins PW. Protein glycosylation in yeast: Transcript heterogeneity of the ALG7 gene. Proc Natl Acad Sci USA (1987) 84:2145–2149.
Lehrman MA. Teaching dolichol-linked oligosaccharides more tricks with alternatives to metabolic radiolabeling. Glycobiology (2007) 17:75R–85R.
Lehrman MA, Russell DW, Goldstein JL, Brown MS. Alu-Alu recombination deletes splice acceptor sites and produces secreted low density lipoprotein receptor in a subject with Familial Hypercholesterolemia. J Biol Chem (1987) 262:3354–3361.
Lehrman MA, Zhu X, Khounlo S. Amplification and molecular cloning of the hamster tunicamycin- sensitive N-acetylglucosamine-1-phosphate transferase gene: The hamster and yeast enzymes share a common peptide sequence. J Biol Chem (1988) 263:19796–19803.
Marek W, Vijay K, Marth D. A recessive deletion in the GlcNAc-1-phosphotransferase gene results in peri-implantation embryonic lethality. Glycobiology (1999) 9:1263–1271.
Rosenwald AG, Stoll J, Krag SS. Regulation of glycosylation: Three enzymes compete for a common pool of dolichyl phosphate in vivo. J Biol Chem (1990) 265:14544–14553.
Rozelle AL, Machesky LM, Yamamoto M, Driessens MHE, Insall RH, Roth MG, Luby-Phelps K, Marriott G, Hall A, Yin HL. Phosphatidylinositol 4,5-bisphosphate induces actin-based movement of raft-enriched vesicles through WASP-Arp2/3. Curr Biol (2000) 10:311–320.[CrossRef][Web of Science][Medline]
Scocca JR, Hartog KO, Krag SS. Evidence of gene amplification in tunicamycin-resistant chinese hamster ovary cells. Biochem Biophys Res Commun. (1988) 156:1063–1069.[CrossRef][Web of Science][Medline]
Shang J, Gao N, Kaufman RJ, Ron D, Harding HP, Lehrman MA. Translational balancing by the eIF2
kinase PERK couples ER glycoprotein synthesis to lipid-linked oligosaccharide flux. J Cell Biol (2007) 176:605–616.
Slonina AE, Scott TC, Lehrman MA. Gene inactivation in Lec35.1 (mannosylation-defective) Chinese hamster ovary cells. J Biol Chem (1993) 268:6729–6733.
Waldman BC, Oliver C, Krag SS. A clonal derivative of tunicamycin-resistant Chinese hamster ovary cells with increased N-acetylglucosamine-phosphate transferase activity has altered asparagine-linked glycosylation. J Cell Physiol (1987) 131:302–317.[CrossRef][Web of Science][Medline]
Ware FE, Lehrman MA. Expression cloning of a novel suppressor of the Lec15 and Lec35 glycosylation mutations of Chinese hamster ovary cells. J Biol Chem (1996) 271:13935–13938.
Zeng Y, Lehrman MA. A block at Man5GlcNAc2-pyrophosphoryldolichol in intact but not disrupted castanospermine and swainsonine-resistant Chinese hamster ovary cells. J Biol Chem (1990) 265:2296–2305.
Zeng Y, Lehrman MA. Isolation of Chinese hamster ovary cell lines producing Man3GlcNAc2 asparagine-linked glycans. Anal Biochem (1991) 193:266–271.[CrossRef][Web of Science][Medline]
Zhu X, Lehrman MA. Gene amplification and molecular cloning of N-acetylglucosamine- 1-phosphate transferase (GPT) from Chinese hamster ovary (CHO) cells. J Cell Biol (1989) 107:851.[CrossRef][Web of Science]
Zhu X, Zeng Y, Lehrman MA. Evidence that the hamster tunicamycin resistance gene encodes UDP-GlcNAc:dolichol phosphate N-acetylglucosamine-1-phosphate transferase. J Biol Chem (1992) 267:8895–8902.
Zou J, Scocca JR, Krag SS. Asparagine-linked glycosylation in Schizosaccharomyces pombe: Functional conservation of the first step in oligosaccharide-lipid assembly. Arch Biochem Biophys (1995) 317:487–496.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Noffz, S. Keppler-Ross, and N. Dean Hetero-oligomeric interactions between early glycosyltransferases of the dolichol cycle Glycobiology, May 1, 2009; 19(5): 472 - 478. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||