Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (10)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Dal Nogare, A. R.
Right arrow Articles by Lehrman, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dal Nogare, A. R.
Right arrow Articles by Lehrman, M. A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Glycobiology Pages 625-632  

Conserved sequences in enzymes of the UDP-GlcNAc/MurNAc family are essential in hamster UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase
Introduction
Results
Discussion
Materials and methods
Acknowledgments
Abbreviations
References

Conserved sequences in enzymes of the UDP-GlcNAc/MurNAc family are essential in hamster UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase

Conserved sequences in enzymes of the UDP-GlcNAc/MurNAc family are essential in hamster UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase

Anthony R.Dal Nogare1,3, Ning Dan2,3,5, Mark A.Lehrman2,4

Departments of Internal Medicine1 and Pharmacology2, University of Texas Southwestern Medical Center At Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75235-9041, USA

Received on November 13, 1997; revised on December 31, 1997; accepted on January 2, 1998

The UDP-GlcNAc/MurNAc family of eukaryotic and prokaryotic enzymes use UDP-GlcNAc or UDP-MurNAc-pentapeptide as donors, dolichol-P or polyprenol-P as acceptors, and generate sugar-P-P-polyisoprenols. A series of six conserved sequences, designated A through F and ranging from 5 to 13 amino acid residues, has been identified in this family. To determine whether these conserved sequences are required for enzyme function, various mutations were examined in hamster UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase (GPT). Scramble mutations of sequences B-F, generated by scrambling the residues within each sequence, demonstrated that each is important in GPT. While E and F scrambles appeared to prevent stable expression of GPT, scrambling of B-D resulted in GPT mutants that could be stably expressed and bound tunicamycin, but lacked enzymatic activity. Further, the C and D scramble mutants had an unexpected sorting defect. Replacement of sequences B-F with prokaryotic counterparts from either the B.subtilis mraY or E.coli rfe genes also affected GPT by preventing expression of the mutant protein (B, F) or inhibiting its enzymatic activity (C-E). For the C-E replacements, no acquisition of acceptor activity for polyprenol-P, the fully unsaturated natural bacterial acceptor, was detected. These studies show that the conserved sequences of the UDP-GlcNAc/MurNAc family are important, and that the eukaryotic and prokaryotic counterparts are not freely interchangeable. Since several mutants were efficiently expressed and bound tunicamycin, yet lacked enzymatic activity, the data are consistent with these sequences having a direct role in product formation.

Key words: UDP-GlcNAc/dolichol/tunicamycin/ glycosyltransferase

Introduction

The identification of highly conserved sequences in an enzyme family can greatly facilitate the study of the structures and functions of the individual members. For example, such insights into type II glycosyltransferases of the Golgi apparatus have led to creative studies of enzyme function as well as to the cloning of new transferases. UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase (GPT) is an endoplasmic reticulum glycosyltransferase which catalyzes the committed step of dolichol-linked oligosaccharide biosynthesis in eukaryotes, by transferring GlcNAc-1-P from UDP-GlcNAc to the C95 polyisoprenoid acceptor lipid dolichol-P to yield GlcNAc-P-P-dolichol (Lehrman, 1991). GPT is distinct from the type II transferases in that it has multiple transmembrane segments and a catalytic site which is oriented toward the cytosol (Dan et al., 1996, and references cited therein).

This study was designed to test a hypothesis formulated by a Commentary in this journal, that a set of six conserved sequences (designated A-F) found in eukaryotic GPTs and in prokaryotic enzymes encoded by the mraY and rfe genes are important for enzyme function (Lehrman, 1994). The bacterial enzymes use UDP-GlcNAc (rfe) or the structurally related UDP-MurNAc-pentapeptide (mraY) as the donor, and transfer GlcNAc-1-P or MurNAc-1-P pentapeptide, respectively, to the acceptor lipid C55 polyprenol-P (a polyisoprenoid lipid that differs from dolichol-P by unsaturation of the [alpha]-isoprene unit). Thus, the prokaryotic enzymes catalyze reactions analogous to those of the eukaryotic GPTs. However, it is clear that eukaryotic GPTs are much more effective with dolichol-P as acceptor compared with polyprenol-P, while the opposite is true for the rfe protein (Rush et al., 1997, and references cited therein). In hamster GPT the conserved sequences occur at or near predicted membrane/water interfaces. This suggests roles in substrate binding and/or catalysis since the donor (UDP-GlcNAc) is highly polar while the acceptor (dolichol-P) is hydrophobic. Based on the sequence similarities, this group of enzymes has been termed the UDP-GlcNAc/MurNAc family (Lehrman, 1994).

With the advent of successful genome sequencing efforts, the number of prokaryotic gene products that have been identified with these sequences has increased substantially since the initial report (Table I). The list now includes several gene products, other than mraY or rfe, of unknown or ambiguous function. However, with the exception of some recently published data from this laboratory on a GPT mutant affecting sequence B (Dan and Lehrman 1997), there has been no experimental confirmation that these sequences are necessary for the function(s) of any of the members of the UDP-GlcNAc/MurNAc family. Sequences B-F are of particular interest because they are clustered within a stretch of [sim]150 residues (Lehrman, 1994). In this report the effects of mutations of these sequences in hamster GPT are examined. To aid the reader, 'scramble" mutations in GPT will be denoted in the text with a superscript 's," and 'replacement" mutations will be denoted with a superscript 'r."

Results

Sequences B, C, and D are important for GPT activity, but not folding or expression

Sequences B-F of hamster GPT were scrambled so as to eliminate similarity with the consensus shown in Table I, with only minimal alteration of the amino acid composition (Table II). Each mutant enzyme was assessed in vitro by GPT activity assays and immunoblots after transient transfection in COS-6 cells and stable transfection in CHO-K1 cells, as well as by the ability to confer resistance to the specific inhibitor tunicamycin (Tn) in vivo in stable transfectants. Cellular resistance to Tn is caused primarily by the ability of GPT to 'buffer" Tn (Dan et al., 1996) rather than a compensatory increase of enzyme activity. Tn binding does not require a catalytically functional enzyme. Therefore, Tn resistance can be used to assess folding of mutant forms of GPT, independent of catalytic activity.

Table I. Sequence similarities in the UDP-GlcNAc/MurNAc family
A series of six sequence similarities (A through F) are shown, based on the predicted coding regions listed below
H-GPT Chinese hamster ovary GPT [J05590]
Y-GPT Saccharomyces cerevisiae GPT [Y00126]
Bs-mraY Bacillus subtilis mraY protein, UDP-N-Acetylmuramoyl-pentapeptide:undecaprenol-P phospho-N-acetylmuramoyl-pentapeptide transferase [Z15056]
Ec-mraY Escherichia coli mraY protein [X51584]
Ec-rfe E.coli rfe protein, UDP-GlcNAc:Undecaprenol-P GlcNAc-1-P transferase [M76129]
Ec-rfe′ Ec-rfe translated from upstream initiator at nt 324
Sua-GPT Sulfolobus acidocaldarius GPT homolog [P39465]
Bb-mraY Borrelia burgdorferi mraY protein [U43739]
Sta-llm Staphylococcus aureus llm protein (affects lysis rate and methicillin resistance) [A55856]
Ss-hyp Synechocystis sp. hypothetical protein [D64005]
Ye-trsF Yersinia enterocolitica trsF protein [S51265]
Ml-rfe Mycobacterium leprae rfe protein [P45830]
Hi-mraY Haemophilus influenzae mraY protein [P45062]
Hi-rfe Haemophilus influenzae rfe protein [P45341]
Mt-unk Mycobacterium tuberculosis unknown protein [Z73419]
Pa-rfbA Pseudomonas aeruginosa rfbA protein [U17293]
Rm-mraY Rhizobium meliloti mraY protein [L25875]
A: H-GPT 19 LEU Leu GLY PHE Val ala Thr Val thr¦
  Y-GPT 27 ala Val GLY PHE gly ILE Ala gly Tyr
  Bs-mraY 11 LEU Met GLY PHE Leu ILE Ser Val Leu
  Bb-mraY 22 Ile Phe Ala PHE Leu Leu Ser Leu Ile
  Ss-hyp 58 LEU Val Thr ala Leu ILE Gly Met ala
  Consensus   LEU H GLY PHE H ILE S H H
B: H-GPT 123 ARG his LYS¦ LEU Leu
  Y-GPT 153 ARG his LYS Phe Phe
  Bs-mraY 114 Lys gln LYS LEU Ile
  Hi-mraY 131 ARG trp LYS Tyr Phe
  Ss-hyp 153 Lys gln LYS LEU Phe
  Consensus   ARG X LYS LEU H
C: H-GPT 136 LEU LEU MET VAL TYR PHE¦
  Y-GPT 166 LEU LEU MET VAL TYR Tyr
  Bs-mray 127 Phe Tyr ala VAL TYR his
  Ec-mraY 177 Ile LEU Leu ala TYR PHE
  Ec-rfe′ 110 Val Met MET VAL Phe gly
  Consensus   LEU LEU MET VAL TYR PHE
D: H-GPT 163 Leu ASP¦ LEU GLY Ile Leu TYR Tyr Val Tyr met
  Y-GPT 194 Val ASP LEU GLY Leu Trp TYR Tyr Val Tyr met
  Bs-mraY 149 Phe ASP LEU GLY Trp ala TYR Phe Ile Leu val
  Ec-rfe[prime] 131 Met val LEU GLY Pro Phe gly Tyr Phe Leu thr
  Bb-mraY 161 Ile ASP LEU GLY Leu Phe TYR ile Pro Phe gly
  Ss-hyp 188 Leu pro LEU GLY Phe Leu Phe gln Leu Val ala
  Ye-trsF 130 Ile Glu LEU GLY Ile Phe gly ser Leu Phe gly
  Hi-mraY 170 Pro gln LEU GLY Leu Phe TYR Ile Val ser leu
  Hi-rfe 125 Leu thr LEU GLY ser Ile gly Leu Ile Ile thr
  Pa-RFB 130 Val ASP LEU GLY Trp Leu gly his Val Leu ala
  Consensus   H ASP LEU GLY H H TYR H H H X
E: H-GPT 181 ¦thr ASN ALA Ile ASN Ile
  Y-GPT 212 pro ASN Ser Ile ASN Ile
  Bs-mraY 167 ser ASN ALA Val ASN Leu
  Ec-mraY 220 gly ASN - Met ASN Phe
  Ec-rfe 39/            
  Ec-rfe[prime] 149 ile ASN ALA Phe ASN Ile
  Bb-mraY 179 ser ASN Ser Phe ASN Leu
  Sta-llm 144 thr ASN ALA Ile ASN Leu
  Ss-hyp 206 ser ASN ALA thr ASN Leu
  Ye-trsF 148 leu ASN leu Tyr ASN Phe
  Ml-rfe 186 val ASN ALA Ile ASN Phe
  Hi-mraY 188 gly ASN ALA Val ASN Leu
  Hi-rfe 143 ile ASN ALA Phe ASN Met
  Mt-unk 186 val ASN ALA Met ASN Phe
  Pa-RBF 148 leu ASN leu Tyr ASN Phe
  Sua-GPT 157 ala ASN ALA Phe ASN Met
  Consensus   X ASN ALA H ASN Ile
F: H-GPT 245 PRO Ser gln VAL PHE Val GLY ASP Thr¦phe Cys Tyr Phe  
  Y-GPT 279 PRO ALA thr VAL PHE Val Ala ASP Thr tyr Cys Tyr Phe
  Bs-mraY 224 PRO ALA lys VAL PHE MET GLY ASP Thr GLY SER Leu ala
  Bs-mraY 257 ile Gly gly VAL PHE Val ile Glu Thr leu SER Val Ile
  Ec-mraY 260 PRO ALA gln VAL PHE MET GLY ASP val GLY SER Leu ala
  Ec-mraY 293 met Gly gly VAL PHE Val val Glu Thr leu SER Val Ile
  Ec-rfe 100/                          
  Ec-rfe[prime] 210 arg tyr lys VAL PHE MET GLY ASP Ala GLY SER thr Leu
  Sua-GPT 216 PRO ALA lys thr PHE pro GLY asn ile GLY Thr Tyr Phe
  Bb-mraY 251 PRO ALA lys Ile Met MET GLY ASP Thr GLY SER Leu ala
  Bb-mraY 284 leu ALA gly VAL PHE Ile ile Glu Thr met SER Val Ile
  Sta-llm 210 PRO ALA lys Ile PHE Leu GLY ASP Ser GLY Ala Leu Met
  Ss-hyp 262 PRO ALA arg VAL PHE MET GLY ASP Thr GLY SER Leu ala
  Ss-hyp 295 ile Ser gly Ile PHE Leu Ala Glu Ser leu SER Val Ile
  Ye-trsF 208 PRO ALA lys Ile PHE MET GLY ASP Ala GLY SER gly Phe
  Ml-rfe 251 arg ALA lys Ile PHE MET GLY ASP Ser GLY SER Met Leu
  Hi-mraY 261 PRO ALA gln VAL PHE MET GLY ASP val GLY SER Leu ala
  Hi-mraY 193 met Gly gly VAL PHE Val val Glu Ala leu SER Val Ile
  Hi-rfe 226 lys tyr lys VAL PHE MET GLY ASP Ala GLY SER thr Leu
  Mt-unk 251 arg ALA lys Ile PHE MET GLY ASP Ser GLY SER Met Leu
  Pa-RFB 205 PRO ALA arg Ile PHE MET GLY ASP Ala GLY SER gly Phe
  Rm-mraY 32 PRO ALA ala Ile PHE MET GLY ASP Thr GLY SER Leu ala
  Rm-mraY 65 ile Gly gly Leu PHE Val ile Glu Thr leu SER Val Ile
  Consensus   PRO ALA X VAL PHE MET GLY ASP S GLY SER H H
The number of the first residue of each peptide is listed. Use of three uppercase letters indicates residues identical in at least 50% of the sequences. One uppercase letter indicates similar residues. No uppercase letters indicates dissimilar residues. For each consensus (in boldface): H = strongly hydrophobic, S = small side chain (weakly hydrophobic or weakly polar), and X = any amino acid. The underline (used only for H-GPT) denotes a proposed hydrophobic transmembrane segment (see Zhu and Lehrman 1990), and '¦" indicates a border between a hydrophobic and a hydrophilic segment. GenBank accession numbers are in square brackets next to the sequence names. Note that Ec-rfe [M76129] contains a potential in-frame upstream initiator codon GUG, which normally encodes valine but can also encode an initiator methionine, at nt 324 of its 5[prime] untranslated region. Translation from this codon yields a longer open reading (Ec-rfe[prime]) that has apparent matches for sequences C and D in addition to matches for E and F already noted for Ec-rfe.

Table II. Activities of hamster GPT scramble mutations
Sequence
scrambled		 Transfection
Mutation		 Transient
Fold-GPT activity
(% SE)		 Stable
Fold-GPT activity
(% SE)		 Stable
Tn resistance
(µg/ml)		 Stable
Activity:Tn
resistance ratio
None (Vector) 1.0 (8%) 1.0 (7%) 0.5 -
None (Normal GPT) 3.0 (5%) 3.6 (10%) 8 0.35
B from:RHKLL
to: HRLKL		 1.1 (5%) 0.7 (2%) 10 (0)
C from:LLMVYF
to: ASMYVT<		 1.1 (1%) 0.8 (6%) 6 (0)
D from:LDLGILYYVYM
to: DLGLLAVVYMY		 1.0 (13%) 0.7 (9%) 10 (0)
E from:TNAINI
to: NTIAIN		 0.7 (4%) 0.6 (4%) 0.5 -
F from:PSQVFVGD
to: SPVQVFDG		 0.9 (11%) 0.9 (13%) 0.5 -
Hamster GPT was scrambled in each of five sequences as indicated, and transfected transiently in COS-6 or stably into CHO-K1 cells. GPT activities were performed in duplicate. The results are expressed as fold-change with respect to the vector control with the percent standard error given in parentheses. The actual activities for vector transfectants were 639 c.p.m. ± 8% (COS) and 1520 c.p.m. ± 7% (CHO). The actual activites for normal GPT were 1843 c.p.m. ± 5% (COS) and 5508 c.p.m. ± 10% (CHO). For stable transfectants growth was tested in 2-fold increases of tunicamycin (Tn), and the Tn resistance was determined as the highest concentration that did not inhibit growth of the cells. For cells with appreciable Tn resistance above vector controls (that is, greater than 0.5 µg/ml), indicating expression of properly folded enzyme, Tn resistance was used to normalize the GPT activity (final column). This calculation was corrected for the contribution of endogenous enzyme by subtracting 0.5 from the Tn resistance value and by subtracting 1.0 from the fold-increase of enzyme activity. Thus, normal transfected enzyme had a value of 0.35. A zero in parentheses indicates a GPT mutant that conferred resistance to Tn, but had no significant enzyme activity. This method of analysis was described earlier in detail (Dan et al., 1996).

Table III. Summary of properties of hamster GPT mutations
Sequence
mutagenized		 Scramble (S) or
replacement (R)		 Enzyme active
in vitro?		 Tn Resistance
in vitro?		 Normal levels
of protein detected?		 Sorting defect
proposed?
B S No Yes Yes No
B R No No No No
C S No Yes No Yes
C R No Yes Yes No
D S No Yes No Yes
D R No Yes Yes No
E S No No No No
E R No Yes Yes No
F S No No No No
F R No No No No

All five scramble mutations eliminated detectable GPT activity in microsomes from transfected COS and CHO cells (Table II). These mutants could be divided into two groups. For three mutants (Bs, Cs, Ds), expression was verified in CHO cells by Tn resistance (Table II) and in COS (data not shown) and CHO cells (Figure 1) by detection on immunoblots. Thus, the Bs, Cs, and Ds mutants were properly folded but catalytically inactive. The two other mutants, Es and Fs, were expressed weakly in COS (data not shown) and CHO (Figure 1) transfectants as judged by immunoblotting. However, in contrast to Bs-Ds, they did not confer resistance to Tn (Table II). It was not clear whether this was due to malfolding, toxicity of the mutant proteins, or a direct effect on the Tn binding site.


Figure 1. Immunoblot analysis of GPT scramble mutation transfectants. Microsomes (50 µg membrane protein/lane) isolated from the stable transfectants listed in Table II were characterized by immunoblot analysis, with an antibody directed against residues 42-56, as described previously (Zhu et al., 1992). Included for comparison are untransfected CHO cells (control) and cells transfected with normal GPT. Note that an empty well between samples E and F was electronically excised from the image.

Sequences B-F were also analyzed with these single mutations at highly conserved residues: R123N and K125N (sequence B), L137A and Y140L (sequence C), D164N and Y169V (sequence D), N182S and N185S (sequence E), and F249L and D252N (sequence F). After stable transfection into CHO cells and normalization of enzyme assay data to Tn resistance, no mutation resulted in a loss of more than half of the enzymatic activity (data not shown). The inability of single mutations at these highly conserved residues to inhibit activity strongly is considered in the Discussion.

Abnormal sorting of the Cs and Ds mutants

Immunoblots of microsomes repeatedly suggested that the signals for the Cs and Ds scramble mutants were weaker than for normal GPT or the Bs scramble mutant (Figure 1), even though the corresponding transfectants had similar resistance to Tn (6-10 µg/ml; Table II). Since GPT spans the ER membrane multiple times and would not be expected to exist freely in the aqueous phase, such results could be explained if the Cs and Ds mutants proteins had mis-sorted to another membranous organelle in which they could still buffer Tn and protect the cell. Furthermore, CHO microsomes used for immunoblot analysis are recovered from a 1000 × g supernatant fraction by centrifugation at 100,000 × g. Thus, the possibility was raised that the Cs and Ds mutants mis-sorted to an organelle that sedimented at 1000 × g, most likely the nucleus.

The localization of the Cs and Ds mutants was investigated by confocal immunofluorescence microscopy with an antibody specific for residues 42-56 (cytosolic loop 1/2) as done earlier (Dan et al., 1996). As expected, the endogenous level of GPT was not detectable in untransfected CHO cells (Figure 2), whereas both transfected normal GPT (Figure 2) and the Bs mutant (data not shown) gave a typical ER fluorescence pattern. This pattern is equivalent to that obtained by staining for the ER chaperone BiP (Dan et al., 1996). In contrast, immunofluorescence of both the Cs (data not shown) and Ds mutants (Figure 2) revealed staining of the nuclear matrix with little staining of the ER. By adjusting the confocal microscope in multiple focal planes, uniform staining throughout the matrix was observed with no apparent staining of the nuclear membrane. By comparison, overexpressed normal GPT (Zhu et al., 1992) can sometimes be found in the nuclear membrane but not the matrix.


Figure 2. Indirect immunofluorescence confocal microscopy. Untransfected CHO-K1 cells (CHO), and CHO-K1 cells transfected with either normal GPT (GPT) or GPT with a scramble of sequence D (GPT-D) were analyzed. See Table II for the properties of these cells. Cells were fixed, permeabilized with Triton X-100, and characterized by indirect immunofluorescence confocal microscopy exactly as described earlier (Dan et al., 1996) using an anti-peptide polyclonal antibody specific for residues 42-56 of GPT. Staining of each cell with nonimmune antibody gave a background pattern equivalent to that shown in the figure for untransfected CHO cells with residue 42-56 antibody (data not shown). As described in the text, examination of the GPT-D cells in multiple focal plains consistently revealed nuclear matrix staining. All three panels were generated under identical microscopic and photographic conditions for direct comparison.

Attempts to identify the Cs and Ds mutants on immunoblots of 1000 × g pellets were not successful (data not shown; see Discussion). To demonstrate the depletion of the Cs and Ds mutants from microsomes independently, an assay was developed to detect such GPT mutants as Tn binding proteins in 100,000 × g microsomal samples. This assay took advantage of the ability of enzymatically inactive GPT mutants to buffer Tn in vitro and lessen the inhibition of normal GPT by Tn. Thus, 10 µg microsomal membrane protein from a CHO transfectant overexpressing active GPT (see Table II) was assayed in the absence or presence of 20 µg of microsomal membrane protein from CHO cells stably transfected with GPT mutations Bs-Es, with either 0 or 5 ng/ml Tn (Figure 3). Experiments with 10 ng/ml Tn gave results similar to those with 5 ng/ml Tn (data not shown). The endogenous GPT in 20 µg mutant microsomes did not appreciably alter the total GPT activity in the experiment, most of which was due to the overexpressed GPT in the 10 µg of microsomes from the normal transfectant. In all cases Tn reduced the total GPT activity. However, the presence of Tn buffering activity could be detected by the ability of certain mutant membrane preparations to increase relative GPT activity in the presence of Tn (right panel) compared to the absence of Tn (left panel). As expected, microsomes from Bs mutant cells conferred substantial resistance to Tn, and microsomes from the Es mutant had no effect. Further, in accord with the results of Figure 2, microsomes from cells harboring the Cs and Ds mutants were as ineffective as Es mutant microsomes although the Cs and Ds cells were highly resistant to Tn.


Figure 3. Detection of properly folded but enzymatically inactive GPT in microsomes by buffering of Tn in vitro. 20 µg of microsomal membranes from each of the indicated GPT scramble mutations (same cells as indicated in Table II) were incubated in standard assay buffer (Lehrman et al., 1988) in the absence (left panel) or presence (right panel) of 5 ng/ml Tn for 10 min at room temperature. Similar results were obtained with 10 ng/ml Tn (data not shown). Then, 10 µg samples of microsomal membrane protein from GPT-overexpressing cells and UDP-[3H]GlcNAc were added, and the assay was continued as described (Lehrman et al., 1988) for 30 min at 37°C. In each case, the activity of 10 µg microsomal membranes from the overexpressing 'normal GPT" transfectant in the presence of 20 µg mutant microsomal membranes was divided by the activity of the 10 µg of overexpressing 'normal GPT" microsomal membranes alone. The averaged results (±SE) of determinations done in two separate experiments are shown. Under these conditions the activity of 10 µg microsomes in the absence of mutant microsomes was inhibited 90% by 5 ng/ml Tn.

Taken together, the results of Figures 2 and 3 suggest that the weak detection of the Cs and Ds mutants on immunoblots of microsomal fractions was due to abnormal sorting. However, this abnormality did not hinder the ability of these mutants to confer resistance to Tn in intact cells. The possible relevance of this observation is considered under the Discussion.

Replacement of the B-F sequences with corresponding sequences from bacterial homologs

If these sequences serve identical functions in eukaryotic and prokaryotic enzymes they should be freely interchangeable. Conversely, if they serve distinct functions such as the recognition of dolichol-P (eukaryotes) vs. polyprenol-P (prokaryotes), they should not be interchangeable. The E.coli rfe protein seemed to be a better candidate than the mraY protein for sequence replacements with hamster GPT since the rfe reaction product, GlcNAc-P-P-polyprenol, had a closer structural similarity to the product of GPT, GlcNAc-P-P-dolichol. However, E.coli rfe lacked sequences similar to B-D. Therefore, the Er and Fr replacements were made from E.coli rfe protein, and the Br-Dr replacements were made from the B.subtilis mraY protein, exactly as indicated by Table I.

For reasons which remain unclear, the Br and Fr replacements could not be expressed (data not shown). In contrast, good expression of the Cr, Dr, and Er replacements was achieved with both transient (data not shown) and stable (Figure 4) transfectants as judged with immunoblots, and in stable transfectants by resistance to Tn, indicative of correct folding. However, the Cr, Dr, and Er replacements had little or no enzyme activity when assayed in intact membranes with endogenous dolichol phosphate as the acceptor (Figure 4). Thus, the hamster sequences were not freely interchangeable with bacterial counterparts.


Figure 4. Expression of bacterial replacement mutants. Stable CHO-K1 transfectants harboring vector, normal GPT, or the bacterial replacements C, D, or E, were characterized by: GPT activity with endogenous acceptor, 20 µg membrane protein per assay, in duplicate ± standard error (A); immunoblotting with a polyclonal antibody directed against residues 42-56, 5 µg membrane protein per lane (B); or cellular resistance to tunicamycin (C).

To account for the loss of enzyme activity for the Cr, Dr, and Er replacements, the possibility was considered that the switch to bacterial sequences may have caused the enzyme to change its acceptor preference from dolichol-P to polyprenol-P. Therefore, assays were performed with NP-40 in the presence of exogenously added dolichol-P or polyprenol-P. While 0.1% (w/v) NP-40 is typically used to assay normal GPT with exogenous dolichol-P (Dan et al., 1996), 0.04% (w/v) NP-40 was used to test the Cr-Er replacements since this concentration was more effective for polyprenol-P (Figure 5A). In accord with the results of Figure 4, the Cr-Er replacements were inactive with various concentrations of dolichol-P as acceptor (data not shown). The Cr and Er replacements also exhibited detergent-dependent dominant-negative inhibition of the type described earlier (data not shown) (Dan and Lehrman 1997). Further, comparisons of the ratios of activities with polyprenol-P to dolichol-P at various acceptor concentrations revealed no acquisition of a preference for polyprenol-P, such as shown in Figure 5B. Therefore, polyisoprenol-P binding could not be solely attributed to sequences C, D, or E.


Figure 5. Comparison of activities of bacterial replacement mutants with dolichol-P and polyprenol-P. (A) To determine optimal detergent concentrations, GPT assays with various concentrations of NP-40 (percent w/v) were performed as described previously (Dan et al., 1996) with 20 µg microsomal membrane protein from the 'normal" transfectant of Figure 4 and 5 µg exogenous acceptor, either dolichol-P (C85-105; squares), dolichol-P (C55; diamonds), or polyprenol-P (C55; circles). Assays without exogenous acceptor are also shown (triangles). (B) Assays were performed with the vector and replacement transfectants shown in Figure 4 as described for (A), with 0.04% (w/v) NP-40. The ratios of activities with polyprenol-P (C55) to dolichol-P (C55) are shown. The vector transfectant, rather than the normal transfectant, was used for this comparison; as shown in Figure 4 its total activity was much more similar to the activities of the replacement transfectants. In separate experiments (data not shown), the normal transfectant gave similar results.

Discussion

These studies were performed to determine whether the B - F sequences of hamster GPT, which are conserved with other members of the UDP-GlcNAc/MurNAc family, are important for function. By designing both scramble and bacterial replacement mutations within hamster GPT, we found that sequences B-E were necessary for function. Both strategies produced C and D mutant enzymes that were expressed well and could bind Tn, but lacked enzymatic activity. In the cases of the B and E sequences only one strategy produced a stably expressed protein that was catalytically inactive. The reason for this is not clear. Since neither strategy produced an F mutant that could be stably expressed, it remains to be seen whether the F sequence has a functional role. However, since the F sequence is the longest and most broadly conserved of the sequences (Table I), an important role is likely. In these studies we chose not to characterize the A sequence as it was the most weakly conserved sequence, and its position in GPT was distant from the other sequences (Lehrman, 1994).

The most likely roles for these sequences are in either substrate binding or formation of the pyrophosphate linkage. Bacterial replacement mutations for any sequences involved in acceptor substrate binding would be expected to alter GPT activity, since the eukaryotic and prokaryotic enzymes normally use different acceptors. A similar argument can be made for donor substrate binding in the case of the Br, Cr, and Dr mutants, as these were based on the sequence of B.subtilis mraY which uses UDP-MurNAc-pentapeptide. However, since both GPT and E.coli rfe use UDP-GlcNAc, the inactivity of the Er mutant suggests that the E sequence is not involved in nucleotide-sugar binding.

Since neither the Cr, Dr, nor Er mutants acquired activity for polyprenol-P (Figure 5), these sequences do not appear to have sole responsibility for determination of acceptor specificity. However, the data do not rule out the possibility that together and/or in combination with B and F, these sequences might govern the preference for dolichol-P or polyprenol-P. Since these sequences are predicted to be at membrane/water interfaces and may all be on the same side of the membrane (Dan et al., 1996), it is possible that the membrane spans that bear them may be in close proximity, allowing the sequences to form a functional pocket for acceptor binding. Thus, multiple blocks of amino acids could be involved, and each block may be necessary but not sufficient. This scenario may also explain why the various single amino acid replacements had modest effects on activity; in each case the remaining amino acids in the pocket would be able to mediate acceptor binding. Experiments are in progress to test hamster GPT mutants with multiple bacterial replacements to determine whether this is indeed the case.

The Cs and Ds mutants mis-sorted, apparently to the nucleus. Since the ER is contiguous with the nuclear membrane, mislocalization to the nucleus is not unexpected. Indeed, it is normal to observe GPT in the nuclear envelope when overexpressed (Zhu et al., 1992). GPT is a highly hydrophobic protein that crosses the membrane multiple times, and is unlikely to exist freely in the aqueous phase (Dan et al., 1996). Thus, detection of the Cs and Ds mutants in the nuclear matrix by fluorescence microscopy (with an antibody specific for residues 42-56 of cytoplasmic loop 1/2) was unexpected, since the nuclear matrix is not known to contain membranous structures.

By analogy to the well-known mechanism for cleavage of the SREBP precursor in the ER membrane to yield a soluble cytoplasmic fragment that enters the nucleus (Brown and Goldstein 1997), we considered the possibility that a soluble fragment of GPT containing residues 42-56 was cleaved from the Cs and Ds mutants. Since this epitope is on a cytoplasmic loop of [sim]25 amino acids (Dan et al., 1996), such a fragment would be small enough to enter the nucleus without a nuclear localization signal. Presumably the rest of the GPT polypeptide would remain membrane-associated, possibly in the nuclear membrane, and still bind tunicamycin. However, we were unable to detect any relevant proteolytic fragments of the Cs and Ds mutants in either the 1000 × g or 100,000 × g fractions, with antibodies directed against amino acids 42-56, 397-408 (lumenal C-terminus), or a FLAG tag attached to the C-terminus (data not shown). Nonetheless, the results with the Cs and Ds mutants may indicate proteolytic release of loop 1/2. Such a mechanism might have no physiological significance, and could result from unregulated cleavage of the mutant protein. However, given what is known about SREBP, we considered a model in which GPT served as a regulatory sensor for dolichol-P levels in addition to its role in GlcNAc-P-P-dolichol synthesis. In this model, if the Cs and Ds mutations had prevented dolichol-P binding, such mutations would simulate conditions of cellular dolichol-P depletion. By analogy to SREBP precursor, which is cleaved to release SREBP when cellular cholesterol levels are low, it seemed possible that a soluble peptide derived from loop 1/2 of GPT might represent a nuclear regulatory factor designed to boost levels of dolichol-P biosynthetic enzymes. To explore this, dolichol-P biosynthesis in the Cs and Ds mutants was examined by labeling with [3H]mevalonate. The results did not, however, indicate any elevation of dolichol-P biosynthetic rates or steady-state levels (D. Crick and C. J. Waechter, personal communication).

In summary, the conserved sequences of the UDP-GlcNAc/MurNAc family are essential in hamster GPT, and they are not freely interchangeable with prokaryotic counterparts. Thus, the weakly conserved residues are important. Since a number of the mutants were efficiently expressed and folded, the data are consistent with these sequences having roles in product formation.

Materials and methods

Materials

UDP-[3H]GlcNAc (26 Ci/mmol) was from DuPont. Muta-Gene 2 kits were from Bio-Rad. Geneticin (G418-sulfate) and powdered cell culture media were from Life Technologies, formerly Gibco/BRL. Serum was from Atlanta Biologicals. Citifluor glycerol was purchased from Ted Pella, Inc. Various enzymes required for manipulation of recombinant DNA were from New England Biolabs, Fisher Scientific, or Boehringer-Mannheim. FITC-labeled antibodies used for immunofluorescence microscopy were obtained from Sigma (goat anti-rabbit IgG, cat. #F-9887). Dolichol-P (C85-105) was from Sigma, and dolichol-P (C55) and polyprenol-P (C55) were gifts of Dr. Charles Waechter, University of Kentucky.

Cell culture

CHO-K1 cells were normally maintained in Ham's F-12 medium buffered at pH 7.2 with 15 mM Na-HEPES with 2 % fetal bovine serum and 8 % calf serum as described previously (Lehrman et al., 1988). COS-6 cells were cultured in DME medium containing 10% fetal bovine serum.

Transient and stable transfection

Stable transfection of normal GPT cDNA and various mutants in plasmid pJB20 (which contains a G418-resistance marker) (Slonina et al., 1993) was performed by the calcium phosphate procedure as described earlier (Zhu et al., 1992). Stable transfectants were selected with either 1 mg/ml Geneticin or 1 µg/ml Tn (Sigma #T7765) (Dan et al., 1996); after initial selections with 1 µg/ml Tn, individual colonies were screened with higher concentrations of Tn to obtain the highest expressers as described in detail (Dan et al., 1996). Stable transfectants were subcloned by limiting dilution. Transient transfection of COS-6 cells by a minor modification of the DEAE dextran method (Zhu and Lehrman, 1990) was also performed with cDNA subcloned into pJB20.

Oligonucleotide-directed mutagenesis

Mutagenesis of the 1.6 kb EcoRI-PstI hamster GPT cDNA fragment in pTZ18U was conducted with the Muta-Gene 2 kit and various mutagenic oligonucleotides ranging from 28 to 36 nt in length, with an essential modification as described previously (Datta and Lehrman 1993). As these methods are now standard, the sequences of the 20 oligonucleotides used in this study are not presented here. The correct orientation and sequence of each mutant was confirmed by standard DNA sequencing methods. When practical, unique restriction sites were also introduced to facilitate mutant screening.

Other methods

The following standard methods used in this study have been described previously: microsomal membrane preparation (Lehrman et al., 1988); membrane protein concentration determination (Lehrman et al., 1988); GPT assays with UDP-[3H]GlcNAc, either with intact membranes using endogenous acceptor, or with detergent-treated membranes and exogenous acceptor (Dan et al., 1996); indirect immunofluorescence confocal microscopy with cells permeabilized with Triton X-100 (Dan et al., 1996); and immunoblot analysis (Zhu et al., 1992).

Acknowledgments

This work was supported by Grant GM38545 from the National Institutes of Health and I-1168 from the Robert Welch Foundation. We thank Mr. Biswanath Pramanik for assistance with cell culture, Drs. Charles Waechter, Dean Crick, and Jeff Rush for their advice and comments on the manuscript, and Drs. Paul Rick and Charles Waechter for pointing out the potential upstream initiator codon in Ec-rfe. We also thank Drs. Charles Waechter and Dean Crick for analysis of dolichol metabolism in some of the transfectants.

Abbreviations

CHO, Chinese hamster ovary; ER, endoplasmic reticulum; GPT, GlcNAc-1-P transferase; Tn, tunicamycin.

References

Brown,M.S. and Goldstein,J.L. (1997) The SREBP Pathway: regulation of cholesterol metabolism by proteolysis of a membrane-bound transcription factor. Cell, 89, 331-340. MEDLINE Abstract

Datta,A.K. and Lehrman,M.A. (1993) Both potential dolichol recognition sequences of hamster GlcNAc-1-phosphate transferase are necessary for normal enzyme function. J. Biol. Chem., 268, 12663-12668. MEDLINE Abstract

Dan,N. and Lehrman,M.A. (1997) Oligomerization of hamster UDP-GlcNAc:Dolichol-P GlcNAc-1-P transferase, an enzyme with multiple transmembrane spans. J. Biol. Chem., 272, 14214-14219. MEDLINE Abstract

Dan,N., Middleton,R.M. and Lehrman,M.A. (1996) Hamster UDP-GlcNAc:dolichol-P GlcNAc-1-P transferase has multiple transmembrane spans and a critical cytosolic loop. J. Biol. Chem., 271, 30717-30725. MEDLINE Abstract

Lehrman,M.A. (1991) Biosynthesis of N-acetylglucosamine-P-P-dolichol, the committed step of asparagine-linked oligosaccharide assembly. Glycobiology, 1, 553-562. MEDLINE Abstract

Lehrman,M.A. (1994) A family Of UDP-GlcNAc/MurNAc:polyisoprenol-P GlcNAc/MurNAc-1-P transferases. Glycobiology, 4, 768-771. MEDLINE Abstract

Lehrman,M.A., Zhu,X. and Khounlo,S. (1988) 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., 263, 19796-19803. MEDLINE Abstract

Rush,J.S., Rick,P.D. and Waechter,C.J. (1997) Polyisoprenyl phosphate specificity of UDP-GlcNAc:undecaprenyl phosphate N-acetylglucosaminyl 1-P transferase from E.coli. Glycobiology, 7, 315-322. MEDLINE Abstract

Slonina,A.E., Scott,T.C. and Lehrman,M.A. (1993) Gene inactivtion in Lec35.1 (mannosylation-defective) Chinese hamster ovary cells. J. Biol. Chem., 268, 6729-6733. MEDLINE Abstract

Zhu,X. and Lehrman,M.A. (1990) Cloning, sequence, and expression of a cDNA encoding hamster UDP-GlcNAc:dolichol phosphate N-acetylglucosamine-1-phosphate transferase. J. Biol. Chem., 265, 14250-14255. MEDLINE Abstract

Zhu,X., Zeng,Y. and Lehrman,M.A. (1992) Evidence that the hamster tunicamycin resistance gene encodes UDP-GlcNAc:dolichol phosphate N-acetylglucosamine-1-phosphate transferase. J. Biol. Chem. 267, 8895-8902. MEDLINE Abstract

3Contributed equally to this article
4To whom correspondence should be addressed
5Current address: Department of Pharmacology, Institute of Medical Science, Henan Medical University, Zhengzhou, Henan 450052, P.R. China

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 18 May 1998
Copyright©Oxford University Press, 1998.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 J. Bacteriol.Home page
B. Al-Dabbagh, D. Mengin-Lecreulx, and A. Bouhss
Purification and Characterization of the Bacterial UDP-GlcNAc:Undecaprenyl-Phosphate GlcNAc-1-Phosphate Transferase WecA
J. Bacteriol., November 1, 2008; 190(21): 7141 - 7146.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
J. Lehrer, K. A. Vigeant, L. D. Tatar, and M. A. Valvano
Functional Characterization and Membrane Topology of Escherichia coli WecA, a Sugar-Phosphate Transferase Initiating the Biosynthesis of Enterobacterial Common Antigen and O-Antigen Lipopolysaccharide
J. Bacteriol., April 1, 2007; 189(7): 2618 - 2628.
[Abstract] [Full Text] [PDF]

Home page
 GlycobiologyHome page
N. P. Price and F. A. Momany
Modeling bacterial UDP-HexNAc: polyprenol-P HexNAc-1-P transferases
Glycobiology, September 1, 2005; 15(9): 29R - 42R.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
I. Chantret, J. Dancourt, A. Barbat, and S. E. H. Moore
Two Proteins Homologous to the N- and C-terminal Domains of the Bacterial Glycosyltransferase Murg Are Required for the Second Step of Dolichyl-linked Oligosaccharide Synthesis in Saccharomyces cerevisiae
J. Biol. Chem., March 11, 2005; 280(10): 9236 - 9242.
[Abstract] [Full Text] [PDF]

Home page
 MicrobiologyHome page
C. L. Marolda, J. Vicarioli, and M. A. Valvano
Wzx proteins involved in biosynthesis of O antigen function in association with the first sugar of the O-specific lipopolysaccharide subunit
Microbiology, December 1, 2004; 150(12): 4095 - 4105.
[Abstract] [Full Text] [PDF]

Home page
 MicrobiologyHome page
A. O. Amer and M. A. Valvano
Conserved aspartic acids are essential for the enzymic activity of the WecA protein initiating the biosynthesis of O-specific lipopolysaccharide and enterobacterial common antigen in Escherichia coli
Microbiology, February 1, 2002; 148(2): 571 - 582.
[Abstract] [Full Text] [PDF]

Home page
 MicrobiologyHome page
A. O. Amer and M. A. Valvano
Conserved amino acid residues found in a predicted cytosolic domain of the lipopolysaccharide biosynthetic protein WecA are implicated in the recognition of UDP-N-acetylglucosamine
Microbiology, November 1, 2001; 147(11): 3015 - 3025.
[Abstract] [Full Text] [PDF]

Home page
 GlycobiologyHome page
D. C. Crick, S. Mahapatra, and P. J. Brennan
Biosynthesis of the arabinogalactan-peptidoglycan complex of Mycobacterium tuberculosis
Glycobiology, September 1, 2001; 11(9): 107R - 118R.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
A. O. Amer and M. A. Valvano
The N-Terminal Region of the Escherichia coli WecA (Rfe) Protein, Containing Three Predicted Transmembrane Helices, Is Required for Function but Not for Membrane Insertion
J. Bacteriol., January 15, 2000; 182(2): 498 - 503.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (10)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Dal Nogare, A. R.
Right arrow Articles by Lehrman, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dal Nogare, A. R.
Right arrow Articles by Lehrman, M. A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?