SEARCH

SEARCH BY CITATION

Keywords:

  • Golgi mannosidase II;
  • glycosyl hydrolase;
  • Arabidopsis thaliana;
  • N-glycosylation;
  • Golgi targeting

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

N-glycosylation is one of the major post-translational modifications of proteins in eukaryotes; however, the processing reactions of oligomannosidic N-glycan precursors leading to hybrid-type and finally complex-type N-glycans are not fully understood in plants. To investigate the role of Golgi α-mannosidase II (GMII) in the formation of complex N-glycans in plants, we identified a putative GMII from Arabidopsis thaliana (AtGMII; EC 3.2.1.114) and characterized the enzyme at a molecular level. The putative AtGMII cDNA was cloned, and its deduced amino acid sequence revealed a typical type II membrane protein of 1173 amino acids. A soluble recombinant form of the enzyme produced in insect cells was capable of processing different physiologically relevant hybrid N-glycans. Furthermore, a detailed N-glycan analysis of two AtGMII knockout mutants revealed the predominant presence of unprocessed hybrid N-glycans. These results provide evidence that AtGMII plays a central role in the formation of complex N-glycans in plants. Furthermore, conclusive evidence was obtained that alternative routes in the conversion of hybrid N-glycans to complex N-glycans exist in plants. Transient expression of N-terminal AtGMII fragments fused to a GFP reporter molecule demonstrated that the transmembrane domain and 10 amino acids from the cytoplasmic tail are sufficient to retain a reporter molecule in the Golgi apparatus and that lumenal sequences are not involved in the retention mechanism. A GFP fusion construct containing only the transmembrane domain was predominantly retained in the ER, a result that indicates the presence of a motif promoting ER export within the last 10 amino acids of the cytoplasmic tail of AtGMII.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

N-glycosylation is considered to be one of the most important post-translational protein modifications. In both animal and plant cells, a highly ordered biosynthetic pathway is responsible for the addition of N-linked oligosaccharides to selected asparagine residues of newly synthesized proteins and their subsequent maturation (Kornfeld and Kornfeld, 1985). The basic steps of the N-glycosylation pathway are evolutionarily highly conserved, with both oligomannosidic and complex-type N-glycans being present in animal and plant glycoproteins. However, the structures of mature complex-type N-glycans differ between plants and mammals because of major differences in the final steps of the biosynthetic pathway. Plant complex-type N-glycans are generally smaller, lack β1,4-galactose and sialic acid but contain plant-specificβ1,2-xylose and/or core α1,3-fucose residues (Lerouge et al., 1998; Wilson, 2002). The processing steps leading to the conversion of oligomannosidic N-glycan precursors to hybrid and complex-type structures are fairly well documented in animal cells, but they are not nearly as well understood in plants. To understand the mechanism of these fundamental biological processes in plants, it is important to understand the specificity and regulation of the Golgi-bound glycan-processing enzymes.

According to the current model of N-glycan processing, specific glycosyltransferases and glycosidases work in a sequential manner determined by their substrate specificity and their distribution along the secretory pathway. In plants, five closely coordinated Golgi-located enzymes (GnTI, GMII, GnTII, XylT and FucT; see Figure 9 below) are responsible for the formation of the major complex N-glycan structures (Wilson et al., 2001). It has been proposed that these enzymes act in the following order: GnTI [RIGHTWARDS ARROW] GMII [RIGHTWARDS ARROW]GnTII [RIGHTWARDS ARROW] XylT/FucT (Lerouge et al., 1998; Tezuka et al., 1992). It is well documented that GnTI generates hybrid-type structures (e.g. GlcNAcMan5GlcNAc2) and so initiates the formation of complex N-glycans (Johnson and Chrispeels, 1987; Kornfeld and Kornfeld, 1985). However, the finding that XylT and FucT efficiently act on hybrid structures (Bencur et al., 2005; Leiter et al., 1999) indicates that these two enzymes can act at an earlier stage than initially proposed. This suggests that an alternative pathway GnTI [RIGHTWARDS ARROW] XylT/FucT [RIGHTWARDS ARROW] GMII [RIGHTWARDS ARROW] GnTII also exists in vivo. However, mainly because of the incomplete characterization of plant GMII, a key intermediate enzyme, it was not possible to reconstruct the pathways fully in vitro. Thus an important aspect of this fundamental biological process remained unidentified.

image

Figure 9. Possible routes in the formation of complex N-glycans. Left branch: traditional pathway proposed by Lerouge et al. (1998) and Tezuka et al. (1992). Right branch: additional alternative pathway proposed by our laboratory (Bencur et al., 2005) and further confirmed in this study. It should be pointed out that cross-talk between the two pathways is possible. GnTI, N-acetylglucosaminyltransferase I; GMII, Golgi α-mannosidase II; GnTII, N-acetylglucosaminyltransferase II; XylT, β1,2-xylosyltransferase; FucT, α1,3-fucosyltransferase.

Download figure to PowerPoint

In animals, GMII cleaves two mannose residues from GlcNAcMan5GlcNAc2 N-linked oligosaccharides generating GlcNAcMan3GlcNAc2 structures (for a recent review, see Moremen, 2002). It is well documented that GMII may act immediately after GnTI, thus generating the specific substrate for GnTII which catalyses the conversion of hybrid- to complex-type N-glycans (for a recent review, see Katsuko, 2002). Hybrid-type oligosaccharide structures are predominant when GMII activity is inhibited by the plant alkaloid swainsonine, indicating the central role of GMII in N-glycan processing in mammals (Moremen, 2002). Evidence for the existence of an α-mannosidase II gene in plants was provided by the purification of such a protein from mung bean (Vigna radiata) seedlings, which shows enzymatic properties comparable to those of the mammalian counterpart (Kaushal et al., 1990). However, the limited amount of purified protein obtained in these studies did not allow a detailed analysis of the enzyme.

The aim of this work was to obtain a detailed characterization of plant GMII, the last molecularly uncharacterized enzyme of five closely coordinated enzymes responsible for the formation of the major complex-type N-glycans. Arabidopsis thaliana GMII was cloned, recombinantly expressed in insect cells and characterized at a molecular level. Incubation with various substrates representing different stages in the N-glycan processing pathway revealed that AtGMII efficiently processes different hybrid-type structures. A detailed glycan analysis of AtGMII knockout mutants using matrix-assisted laser-desorption ionization-time-of-flight (MALDI-TOF) analyses revealed the predominant presence of hybrid-type glycan structures carrying β1,2-xylose and core α1,3-fucose residues. These results demonstrate the central role of AtGMII in the formation of complex N-glycans and establish that at least two routes in the processing of N-glycans are possible in plants.

Using transient expression of N-terminal AtGMII fragments fused to a reporter molecule, we show that a truncated cytoplasmic (C) and transmembrane (T) region of the enzyme is sufficient to retain a reporter molecule in the Golgi and that lumenal sequences are not involved in the retention mechanism. Furthermore, evidence for the presence of a motif promoting ER export within the 10 amino acids of the cytoplasmic tail was obtained.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Cloning and identification of Golgi α-mannosidase II (AtGMII) cDNA from Arabidopsis thaliana

GMII cDNAs from several mammalian species as well as from insects have been cloned and characterized previously (for a recent review, see Moremen, 2002). A database search based on the amino acid sequence of the human and Drosophila melanogaster GMII resulted in the identification of four homologous proteins in the A. thaliana genome. All four genes (At3g26720, At5g13980, At5g14950 and At5g66150) were previously annotated as α-mannosidases and categorized as glycosyl hydrolase family 38 members (Carbohydrate-Active enZYmes website: http://afmb.cnrs-mrs.fr/CAZY/). Pair-wise BLAST analyses (http://www.ncbi.nlm.nih.gov/blast) revealed that three of them were more related to lysosomal mannosidases and one (At5g14950) displayed a higher homology to human and D. melanogaster GMII. To clone the entire At5g14950 cDNA, 5′-RACE was performed. The cloned cDNA contained an open reading frame of 3522 bp, which is identical to the annotated At5g14950 nucleotide sequence (accession number NM_121499), and 151 bp of 5′ untranslated region. Comparison of the cloned cDNA with the annotated genomic sequence indicated a gene structure in which there are four exons (Figure 1a). The At5g14950 open reading frame encodes a polypeptide of 1173 amino acids (134 kDa). A hydrophobicity plot as well as a hydrophobic cluster analysis indicated a type II transmembrane protein topology which is typical for Golgi glycosylation enzymes. The unusually long putative N-terminal cytoplasmic tail (50 amino acids) is followed by a 22 amino acid transmembrane domain and a large catalytic domain which contains nine putative N-glycosylation sites (Figure 1b). The identity to human and D. melanogaster GMII is 38% (55% and 52% similarity, respectively) with a number of highly conserved amino acid motifs (Figure 2). Data derived from the crystal structure of D. melanogaster GMII suggest that some of these conserved residues are involved in the coordination of the zinc ion required for catalysis and in the formation of the active site (van den Elsen et al., 2001).

image

Figure 1. Arabidopsis thaliana Golgi α-mannosidase II (AtGMII). (a) Schematic presentation of AtGMII gene organization and insertion site of T-DNA in hgl1-1 (2741 bp) and hgl1-2 (1297 bp) mutants. Black boxes are exons. Primer locations are shown by small arrows and primer numbers are depicted in italics. (b) Protein sequence of AtGMII; the predicted transmembrane is shown in bold, putative N-glycosylation sites are underlined and the dibasic motif [RK](X)[RK] is shown in italics and underlined.

Download figure to PowerPoint

image

Figure 2. Alignment of the predicted GMII protein sequences derived from Arabidopsis thaliana (At; DQ029214), Drosophila melanogaster (Ds; JC4037) and human (Hs; Q16706) using clustalw (http://www.ch.embnet.org/software/ClustalW.html). Identical and similar residues are shaded in black and light grey, respectively. Conserved residues, which are involved in zinc ion binding and active site formation in the D. melanogaster enzyme, are marked at the top with an asterisk.

Download figure to PowerPoint

BLAST analyses have identified a putative GMII orthologue in Oryza sativa (accession number BAD45807), and closely related genes are represented in EST collections from other plant species (e.g. Triticum aestivum, Glycine max and Lycopersicon esculentum).

Heterologous expression of AtGMII in insect cells

To determine the enzymatic properties of AtGMII, a cDNA encoding the putative lumenal catalytic domain lacking the N-terminal 92 amino acids was subcloned into the pVTBacHis vector for its expression in insect cells as a soluble secreted protein containing an N-terminal His tag and an enterokinase cleavage site. A recombinant baculovirus was generated and used to infect Spodoptera frugiperda Sf21 cells. Subsequently, cell supernatant as well as lysate was analysed for α-mannosidase activity in vitro, using a saturating concentration of the synthetic substrate p-nitrophenyl α-D-mannopyranoside (PNP-Man). While high enzyme activity (847 units per mg of total solubilized cellular protein) was found in the culture supernatant, only marginal activity could be detected in the cell lysate (8.0 units per mg of total solubilized cellular protein). In contrast, culture supernatant from uninfected Sf21 cells contained comparatively little α-mannosidase activity (≤1.5 units per mg of total solubilized cellular protein) and no significant hydrolysis of the substrate was detected in lysates of uninfected cells (0.8 units per mg of total solubilized cellular protein). Immunoblotting with antibodies to the enterokinase cleavage site of the fusion protein revealed a major immunoreactive band of approximately 137 kDa in the culture supernatant (Figure 3a). This is in good agreement with the theoretical molecular mass of this form of AtGMII (129 kDa), considering that the protein contains nine potential N-glycosylation sites. The small amount of AtGMII present in the cell lysate could not be detected with this antibody. These results demonstrate that recombinant AtGMII displays significant catalytic activity and can be efficiently produced at high levels in insect cells as a secreted, soluble protein.

image

Figure 3. Expression of AtGMII and characterization of enzyme activity. (a) Heterologous expression of soluble forms of AtGMII in insect cells. Equivalent amounts of culture supernatants (Sn) and protein extracts from Sf21 insect cells (C) either infected (GMII) with recombinant baculoviruses encoding N-terminally truncated AtGMII or uninfected (control) were subjected to Western blotting using antibodies against the enterokinase cleavage site of the recombinant protein. A band corresponding to the expected size of 137 kDa was detected in the supernatant of infected cells. (b) Coomassie Blue-stained SDS/PAGE gel of purified AtGMII. Protein (2 μg) was incubated overnight in the presence (+) or absence (−) of PNGase F prior to electrophoretic analysis. (c) GMII enzyme activity assay. Man5Gn-glycopeptide was incubated without (upper panel) or with (lower panel) purified recombinant AtGMII and analysed using matrix-assisted laser-desorption ionization-time-of-flight (MALDI-TOF) mass spectrometry. The labelled peaks represent (M + Na)+ ions. Other peaks are potassium adducts of the same glycans. In the lower panel, peaks corresponding to the substrate (Man5Gn) and the products of GMII activity (Man4Gn and MGn) are visible. Mass differences correspond to the expected value of 162 for each mannose residue removed.

Download figure to PowerPoint

The secreted recombinant protein was purified from culture supernatants of AtGMII-producing Sf21 cells by means of nickel-chelate affinity chromatography. Endoglycosidase treatment revealed that the purified recombinant AtGMII is sensitive to PNGase F, confirming the presence of N-linked glycans (Figure 3b). Using saturating substrate concentrations, purified recombinant AtGMII cleaved PNP-Man with a specific activity of 6379 units per mg enzyme protein. This compares well with recombinant D. melanogasterα-mannosidase II (van den Elsen et al., 2001), which displayed a specific activity of 10238 units per mg enzyme protein under the same assay conditions.

The different types of intracellular α-mannosidases vary in their metal ion requirements. The activity of AtGMII towards PNP-Man was not sensitive to EDTA and not significantly stimulated by any divalent cation tested (Ca2+, Co2+, Mn2+, Ni2+ or Zn2+), as previously found for GMII isolated from mung bean seedlings (Kaushal et al., 1990) and recombinant D. melanogaster GMII (Rabouille et al., 1999). However, analysis of purified AtGMII by inductively coupled plasma mass spectrometry revealed that the enzyme contains stoichiometric amounts of zinc, as previously shown for recombinant D. melanogaster GMII (Bencur et al., 2005; van den Elsen et al., 2001). Furthermore, AtGMII was inhibited to >97% by 1 mm Cu2+, as also reported for the mung bean and D. melanogaster variants of the enzyme (Kaushal et al., 1990; Rabouille et al., 1999). AtGMII was highly sensitive to the classical class II α-mannosidase inhibitor swainsonine, with 50% inhibition of the enzyme activity (IC50) at an inhibitor concentration of 18 nm. Essentially the same IC50 (16 nm) was determined for the inhibition of the recombinant D. melanogaster enzyme by swainsonine, which is in good agreement with previously published data (Rabouille et al., 1999). The IC50 of swainsonine for mung bean GMII was reported to be in the range of 10–25 nm (Kaushal et al., 1990). These results indicate that AtGMII exhibits similar catalytic properties to those of its orthologues from other species.

The physiological substrate for Golgi α-mannosidase II is GlcNAcMan5GlcNAc2 (Man5Gn), which it converts by sequential removal of two α1,6- and α1,3-linked mannose residues from the α1,6-mannose branch of the substrate into GlcNAcMan3GlcNAc2 (MGn). To test whether the purified protein is able to remove these α1,3- and α1,6-mannosyl residues, AtGMII was incubated with a Man5Gn-glycopeptide. Analysis of the reaction products by MALDI-TOF mass spectrometry clearly showed the presence of α-mannosidase II activity (Figure 3c). AtGMII was also able to cleave β1,2-xylosylated Man5Gn-glycopeptide (Man5GnX-GP), albeit to a lesser extent (Table 1). A similar preference for Man5Gn-glycopeptide over its β1,2-xylosylated derivative was found for D. melanogaster GMII (data not shown). Assays performed with pyridylaminated substrates (Man5Gn-PA and Man5GnX-PA) gave the same results (Table 1). Importantly, neither AtGMII nor the D. melanogaster enzyme cleaved Man5-glycopeptide, indicating the strict requirement for the presence of a terminal β1,2-linked GlcNAc residue on the α1,3-mannose arm of the substrate. These results establish that AtGMII has the capacity to display genuine α-mannosidase II activity in planta.

Table 1.  Substrate specificity of AtGMII for N-glycan structures
StructureNameGMII activity (nmol min−1 mg−1)Relative GMII activity (%)
  1. The values obtained with the glycopeptide (−GP) substrates are derived from matrix-assisted laser-desorption ionization-time-of-flight (MALDI-TOF) mass spectrometry analysis and represent the mean ± standard deviation for four independent experiments. The values obtained with the pyridylaminated (−PA) oligosaccharides are derived from HPLC analysis and represent means of two independent experiments. n.d., not detectable (<0.1 nmol min−1 mg−1 or <0.1%, respectively). As the available amounts of the respective glycopeptides were limited, those assays were performed at non-saturating substrate concentrations.

inline image

Generation and characterization of AtGMII-deficient plants

In a screen of collections of T-DNA insertion lines for the presence of putative AtGMII knockouts, two lines, Salk_052443 and SALK_141821, were identified in the SIGnAL (SALK Institute Genome Analysis Laboratory, http://signal.salk.edu/tabout.html) collection that contained a T-DNA insertion in exon 4 and intron 3 of At5g14950, respectively (Figures 1a and 4b). Subsequently, homozygous plants were generated and the disruption of the GMII gene was confirmed by PCR from genomic DNA (Figure 4a). RT-PCR analysis of RNA extracted from wild-type and knockout plants demonstrated the absence of a functional GMII transcript in both mutants (Figure 4c).

image

Figure 4. Identification of homozygous T-DNA insertion lines. (a) PCR from genomic DNA isolated from Columbia (Col-0) wild-type (1) and T-DNA insertions lines hgl1-1 (2) and hgl1-2 (3). Primers 6 and 11 amplify a 1035-bp fragment from the wild-type AtGMII gene; primers 1 and 2 amplify a 666-bp fragment. The location of the primers is depicted in Figure 1(a). M, lambda DNA EcoRI + HindIII marker. (b) PCR analysis of genomic DNA from Col-0 (1) and hgl1-1 (2) with AtGMII-specific primer 1 and T-DNA-specific primer LBa1, and PCR analysis of genomic DNA from Col-0 (3) and hgl1-2 (4) with AtGMII-specific primer 11 and T-DNA-specific primer LBa1. M, lambda DNA EcoRI + HindIII marker. (c) RT-PCR analysis of RNA extracted from wild-type (1) and homozygous mutants hgl1-1 (2) and hgl1-2 (3). Primers 1 and 2 amplify a 666-bp fragment. As a control, the fucosyltransferase B (FTB) mRNA (319-bp fragment) was monitored. (d) Western blot analysis of proteins extracted from hgl1-1 (1) and the complementation mutant hgl1-1 + AtGMII cDNA (2). Immunodetection was performed using an antibody directed against the myc tag, which was incorporated at the C-terminus of GMII.

Download figure to PowerPoint

To detect changes in the N-glycosylation pattern caused by inactivation of the GMII gene, endogenous glycoproteins isolated from the mutant lines were subjected to total N-glycan analysis by MALDI-TOF mass spectrometry. The mass spectra of wild-type A. thaliana plants contained three major peaks which were assigned to the complex-type N-glycans Man3XylFucGlcNAc2 (m/z 1212), GlcNAcMan3XylFucGlcNAc2 (m/z 1415) and GlcNAc2Man3XylFucGlcNAc2 (m/z 1618) (Figure 5a). The spectra of the AtGMII mutant lines, termed hybrid glycosylation1-1 and 1-2 (hgl1-1 and hgl1-2), lacked these three major complex-type N-glycan peaks. Instead, Man5XylFucGlcNAc2 and GlcNAcMan5XylFucGlcNAc2 structures were most abundant (Figure 5b and Table 2). No further processing of hybrid-type N-glycans to complex-type N-glycans was observed. This result demonstrated the inactivation of AtGMII by the T-DNA insertion and confirmed that A. thaliana contains only one functional GMII protein (encoded by At5g14950) catalysing the removal of both α1,3- and α1,6-mannosyl residues from the α1,6-mannose branch of N-glycans. It is noteworthy that the N-glycan profile could be completely restored to wild-type when the hgl1-1 mutant was complemented with wild-type AtGMII cDNA (Figures 4d and 5c). Like other mutant plants defective in the formation of hybrid and complex N-glycans (von Schaewen et al., 1993; Strasser et al., 2004), hgl1 mutants are viable and fertile and do not show any obvious phenotype under standard growth conditions.

image

Figure 5. Matrix-assisted laser-desorption ionization-time-of-flight (MALDI-TOF) mass spectra of N-linked oligosaccharides released from endogenous glycoproteins. The labelled peaks represent (M + Na)+ ions. Other peaks are potassium adducts of the same glycans. Only the three major peaks are marked. (a) Wild-type Arabidopsis thaliana; Man3XylFucGlcNAc2 (MMXF); GlcNAcMan3XylFucGlcNAc2 (GnMXF/MGnXF); GlcNAc2Man3XylFucGlcNAc2 (GnGnXF). See Wilson et al. (2001) or http://www.proglycan.com for N-glycan abbreviations. (b) hgl1-1: GMII knockout; Man5GlcNAc2 (Man5); Man5XylFucGlcNAc2 (Man5XF); GlcNAcMan5XylFucGlcNAc2 (Man5GnXF). (c) hgl1-1 + GMII: hgl1-1 mutant line complemented with wild-type AtGMII cDNA.

Download figure to PowerPoint

Table 2.  Relative amounts of N-glycans detected in Arabidopsis thaliana wild type, hgl1-1 mutants and hgl1-1 mutant after complementation (hgl1-1 + GMII). Hybrid-type N-glycan structures are shown in bold.
m/z (M + Na)+Hybrid-type and complex-type structuresWild type (%) hgl1-1 (%) hgl1-1 + GMII (%)
933.8Man3GlcNAc20.9
1065.9Man3XylGlcNAc21.62.3
1137.0GlcNAcMan3GlcNAc21.4
1212.1Man3XylFucGlcNAc226.425.4
1269.1GlcNAcMan3XylGlcNAc21.50.9
1390.2Man5XylGlcNAc22.2
1415.5GlcNAcMan3XylFucGlcNAc215.020.9
1461.3GlcNAcMan5GlcNAc20.5
1472.1GlcNAc2Man3XylGlcNAc21.41.1
1536.4Man5XylFucGlcNAc231.5
1593.4GlcNAcMan5XylGlcNAc21.4
1618.5GlcNAc2Man3XylFucGlcNAc226.624.1
1739.6GlcNAcMan5XylFucGlcNAc222.1
 Total73.958.674.7
 Oligomannosidic structures
1258.1Man5GlcNAc210.815.510.9
  1. Percentages were calculated based on peak areas from matrix-assisted laser-desorption ionization-time-of-flight (MALDI-TOF) mass spectra. For wild type and hgl1-1, mean values of at least three independent experiments are shown.

1420.2Man6GlcNAc26.17.44.1
1582.4Man7GlcNAc22.77.54.0
1744.5Man8GlcNAc23.58.93.3
1906.7Man9GlcNAc23.02.13.0
 Total26.141.425.3

Identification of Golgi retention sequences of AtGMII

In order to study the localization and retention of AtGMII in the Golgi apparatus, a series of deletion constructs were expressed in plants. The following experiments are based on the vast store of knowledge that exists concerning the retention of mammalian glycosyltransferases and glycosidases within the Golgi apparatus, and in particular on the observation that the transmembrane domain (T) and its surrounding regions, namely the lumenal stem region (S) and the cytoplasmic domain (C), contribute to Golgi retention (e.g. Nilsson and Warren, 1994; Nilsson et al., 1996). Although little is known about Golgi retention in plants, the mechanisms that underlie this as yet not fully understood process seem to be similar to those found in animals. This was shown by the identification of Golgi retention sequences within the CTS region of plant glycosyltransferases (Dirnberger et al., 2002; Essl et al., 1999). Furthermore, it was previously demonstrated that human glycosylation enzymes are retained in the plant Golgi apparatus, as shown by the expression of active mammalian glycosyltransferases in plant cells (Bakker et al., 2001; Gomez and Chrispeels, 1994; Palacpac et al., 1999; Wee et al., 1998).

In this study, constructs were designed to define (minimal) sequences that contribute to the Golgi retention of AtGMII. First we investigated whether the putative CTS region of AtGMII, which lacks the large catalytic domain, is sufficient to retain a reporter molecule in the plant Golgi apparatus. A CTS-GFP fusion was constructed and inserted into a tobacco mosaic virus (TMV)-based expression vector (Figure 6). TMV vectors have been shown previously to be well suited to the study of protein targeting to cell organelles in the tobacco-related model plant Nicotiana benthamiana (Dirnberger et al., 2002; Essl et al., 1999). Transient expression of the chimeric CTS-GFP was monitored by confocal microscopy. JIM84 antibodies, which recognize the Lewisa carbohydrate epitope predominantly present in the plant Golgi apparatus (Horsley et al., 1993), were used for co-localization studies. Indeed, as shown by in situ indirect immunofluorescence in CTS-GFP-expressing epidermal cells, a significant overlap between the GFP fluorescence and JIM84 staining was obtained (Figure 7). Hence, the Golgi retention signals for AtGMII appear to be included within the 92 N-terminal amino acids. This result is consistent with those of studies on D. melanogaster GMII, where CTS-hemagglutinin-tagged fusions were found in the Golgi apparatus of insect cells (Rabouille et al., 1999). In order to elucidate the individual contributions of the C, T and S regions to Golgi targeting and retention of AtGMII, a series of deletion constructs were subsequently transiently expressed in N. benthamiana (Figure 7). The unusually long C domain of 50 amino acids led us to generate a truncated C construct (C10), containing only 10 amino acids proximal to the T domain. These 10 amino acids contain a dibasic [RK] motif which might promote ER exit, as has been demonstrated for other dibasic motifs in mammalian glycosylation enzymes (Giraudo and Maccioni, 2003). Interestingly, a punctuate staining pattern which co-localized with the Lewisa marker was observed for CT-GFP as well as for C10T-GFP, which suggests that the putative S region and the first 40 amino acids from the C region are not necessary for Golgi targeting and retention. In contrast, the staining obtained for the successive deletion constructs TS-GFP, T-GFP and C-GFP was clearly different from that for CTS-GFP and CT-GFP. In the case of C-GFP expression, we observed cytoplasmic staining identical to that obtained for wild-type GFP. In contrast, the expression of T-GFP and TS-GFP displayed reticular staining throughout the cytoplasm. This ER-like network pattern revealed only marginal co-localization with Lewisa; however, it co-localized with BiP (Figure 8), a well established ER marker (Denecke et al., 1991). These results suggest a requirement of the T region for entry into the secretory pathway; however, the T and TS regions alone are not able to promote exit from the ER. Furthermore, the results indicate the presence of an ER exit motif within the 10 amino acids of the C region, as the C10T-GFP construct is transported to the Golgi.

image

Figure 6. Illustration of recombinant viral vectors used for localization studies in Nicotiana benthamiana plants. The GFP fusion constructs were placed under the control of the coat protein (CP) promoter of the TMV-based expression vector p4GD-pl (Casper and Holt, 1996). Numbers indicate the putative lengths of cytoplasmic (C), transmembrane (T) and stem (S) regions of Arabidopsis thaliana GMII in amino acids. MP, movement protein; T7, T7 promoter for in vitro transcription.

Download figure to PowerPoint

image

Figure 7. Co-localization of AtGMII-GFP fusions with plant Lewisa epitopes in Nicotiana benthamiana epidermal cells by confocal laser scanning microscopy. Left column, GFP staining obtained with CTS-, CT- and C10T-GFP constructs (C, cytoplasmic; T, transmembrane; S, stem); middle column: staining obtained by in situ indirect immunofluorescence using the JIM84 antibody, which detects Lewisa epitopes; right column: overlay of both staining patterns. With all three constructs, significant co-localization with the Golgi marker was obtained. Scale bar, 8 μm.

Download figure to PowerPoint

image

Figure 8. Subcellular localization of transmembrane domain (T)-GFP in epidermal cells of Nicotiana benthamiana by confocal laser scanning microscopy. Left: staining obtained with the T-GFP construct; middle: staining obtained by in situ indirect immunofluorescence using the anti-BiP antibody; right: overlay of both staining patterns. A significant co-localization with the ER marker was obtained. Scale bar, 8 μm.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

In animals, Golgi α-mannosidase II is a key enzyme in the processing of N-linked oligosaccharides, with a strict linear processing order of GnTI [RIGHTWARDS ARROW] GMII [RIGHTWARDS ARROW]  GnTII prior to further processing. However, despite an absolute requirement for the action of GnTI prior to GMII, several lines of evidence indicate that alternate enzyme systems provide a bypass of GMII activity. GMII null mice are capable of synthesizing complex-type oligosaccharides in many adult tissues, although at a reduced level. The activity of an alternative enzyme, termed α-mannosidase III, which converts Man5GlcNAc2 to Man3GlcNAc2, was proposed to compensate GMII deficiency in several tissues of the null mice (Chui et al., 1997). A gene encoding an α-mannosidase III was also identified in Spodoptera frugiperda insect cells (Kawar et al., 2001); however, the physiological role of these enzymes is still unknown. Another enzyme, designated α-mannosidase IIx, which might bypass GMII activity, was identified in mice (Misago et al., 1995). The enzyme converts Man6GlcNAc2 to Man4GlcNAc2 and thus provides another alternative pathway for the production of complex N-glycans in mammals (Oh-eda et al., 2001).

In contrast, the GMII-deficient A. thaliana plants generated and analysed in this work synthesize only hybrid-type structures with no indications of further processing to complex N-glycans. The biochemical analysis of AtGMII clearly showed that the enzyme is distinct from either α-mannosidase III or α-mannosidase IIx. Furthermore, a BLAST search in the A. thaliana genome did not reveal any sequences which could encode alternative processing α-mannosidases. Together, these data strongly indicate the absence of alternative pathways which can bypass GMII in A. thaliana and emphasize the central role of AtGMII in the formation of complex N-glycans in plants.

hgl1 mutant plants are viable and fertile and do not show any obvious phenotype under standard growth conditions. This observation is in good agreement with observations made in other A. thaliana mutants showing defects in hybrid and complex N-glycosylation. XylT, FucT or XylT/FucT double knockout lines do not show an obvious phenotype (Strasser et al., 2004). Similarly, cgl mutants which lack GnTI activity and thus synthesize only oligomannosidic N-glycans do not exhibit a significant phenotype, although cgl plants are slightly more susceptible to certain stress factors (von Schaewen et al., 1993). These observations are in stark contrast to those in mammals, where defects in complex N-glycosylation lead to severe phenotypes (for recent reviews, see Lowe and Marth, 2003; Moremen, 2002; Stanley and Ioffe, 1995). Interestingly, A. thaliana mutants that have alterations in enzymes involved in N-glycan assembly or in N-glycan processing in the ER exhibit phenotypes such as growth defects and developmental abnormalities (e.g. Lerouxel et al., 2005) or even embryo lethality (Boisson et al., 2001). This demonstrates the importance of initial N-glycosylation for normal plant development; however, the function of hybrid and complex N-glycans in plants remains unknown. Thus, mutant A. thaliana plants with a distinct defect in the N-glycosylation pathway may provide a valuable tool to investigate the physiological roles of these oligosaccharides.

Our results regarding Golgi targeting/retention signals demonstrate that a truncated C10T region of AtGMII is sufficient for Golgi localization without any contribution of lumenal sequences. We have shown that motifs enabling ER export are probably located within the 10 amino acids of the C region. Recently, a dibasic motif [RK](X)[RK] that efficiently promotes ER export has been characterized within the C region of animal glycosyltransferases (Giraudo and Maccioni, 2003). Indeed, two such motifs have been found in the C region of AtGMII. Moreover, our results suggest that the [RK] motif proximal to the T region is sufficient to promote ER export, as the truncated C10T construct, which lacks the first [RK] motif, is efficiently exported to and retained in the Golgi apparatus.

In animals, glycosylation enzymes have been used extensively to study Golgi targeting/retention sequences (for a review, see Colley, 1997). The results are mainly based on heterologously expressed sequences fused to reporter molecules; however, because of the different expression systems and methodologies used, the results obtained are difficult to compare. The distinct contributions of C, T and lumenal sequences to Golgi localization are different for individual enzymes and are not well understood. A recent study by Dirnberger et al. (2002) investigated plant Golgi targeting/retention signals in more detail. Their findings indicated the importance of CT sequences for complete Golgi targeting/retention and are in good agreement with the results obtained in the present study with AtGMII. Comparing our results with those obtained for mammalian glycosyltransferases, we conclude that the properties of AtGMII resemble those of β1,4-galactosyltransferase I, with the stem domain playing no major role in defining their subcellular distribution. This provides further evidence for similar Golgi localization mechanisms for glycosylation enzymes in different phyla.

Although there is convincing evidence that the mechanism of N-glycan processing is highly conserved in the early steps between animals and plants, it is not clear at which stage the process starts to differ. One of the major differences between animal and plant complex N-glycans is the presence of β1,2-xylose and core α1,3-fucose residues in plants, and there is an ongoing debate concerning the stage at which these sugar residues are attached. Work by Johnson and Chrispeels (1987) has indicated that α1,3-fucose and β1,2-xylose transfer occurs at the same time as, or some time after, the action of GMII. Lerouge et al. (1998) and Tezuka et al. (1992) propose the action of XylT and FucT in the final stages of the pathway, after GnTII activity. However, Bencur et al. (2005) showed the ability of XylT to act at multiple stages in the pathway. The data presented here indicate that the pathway may begin to diverge at an early stage, immediately after GnTI action, as GMII-deficient plants predominantly synthesize GlcNAcMan5XylFucGlcNAc2 structures. Furthermore, efficient in vitro conversion of GlcNAcMan5XylGlcNAc2 to GlcNAcMan3XylGlcNAc2 by AtGMII and the ability of GnTII to further process GlcNAcMan3XylGlcNAc2 (Bencur et al., 2005) strongly suggest that at least two routes are possible in the biosynthesis of complex N-glycans in plants in vivo (Figure 9). However, it remains to be elucidated whether these alternative routes are actually used in vivo or if spatial separation of the enzymes within the Golgi apparatus regulates these processing steps. In this respect, the elucidation of the subcellular localization of the respective processing enzymes within the Golgi compartments will be of great importance.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Cloning of Arabidopsis thaliana Golgi α-mannosidase II

Total RNA was isolated from seedlings using the TRIzol reagent (Invitrogen, Lofer, Austria) according to the supplier's instructions. We used 0.5 μg of total RNA to prepare cDNA using a SMART RACE cDNA Amplification kit (BD Biosciences, Erembodegem, Belgium). PCR was carried out with Turbo-pfu polymerase (Stratagene, La Jolla, CA, USA) using primers MII-5 (5′-AAAAAGGAAATTCCGAGGAGGAG-3′) and MII-9 (5′-GTGTGTGTATTTCACTTGTGAGG-3′). 5′-RACE was performed according to the supplier's recommendations using reverse primers MII-27 (5′-GCCGGAGAAGATAAGCCATGGTTGATG-3′) and MII-16 (5′-CCTGGTACGCCGAAGTGGA-3′) for nested PCR amplification. The PCR products were subcloned using a Zero Blunt TOPO PCR Cloning kit (Invitrogen) and subsequently sequenced using the PRISM BigDye Terminator Cycle Sequencing kit and the ABI 3100 genetic analyser (Applied Biosystems, Foster City, CA, USA).

Construction of baculovirus transfer vectors containing a truncated form of Arabidopsis thaliana GMII and its heterologous expression in insect cells

The N-terminal deletion construct containing amino acids 93–1173 was generated by PCR using primers MII-18 (5′-TATACTGCAGATCGTCAAGCCACGGAAGAATA-3′) and MII-15 (5′- TTTAGGTACCTCACTTGTGAGGTCGCAGTTC-3′). The PCR product was cleaved with PstI and KpnI restriction enzymes at the underlined sites and ligated into pVTBacHis1 baculovirus transfer vector (Sarkar et al., 1998), digested with the same enzymes. In this construct, the truncated GMII protein is placed downstream of the melittin signal peptide, a 6xHis tag and an enterokinase cleavage site. Expression in Spodoptera frugiperda Sf21 cells was performed exactly as described previously (Bencur et al., 2005). Briefly, the recombinant transfer vector (1 μg) was co-transfected with 200 ng of BaculoGold viral DNA (BD Biosciences) into Sf9 cells using Lipofectin (Invitrogen) as recommended by the manufacturer. After 5 days at 27°C, supernatants containing recombinant virus were used for infection of Sf21 cells. Cells and culture media were harvested after 4 days at 27°C and subjected to enzymatic analysis and immunoblotting.

Western blotting analysis of recombinant Arabidopsis thaliana GMII

Baculovirus-infected and non-infected Sf21 cells were lysed in 80 mm 2-(N-morpholino) ethanesulfonic acid (MES) (pH 5.8), 1% Triton X-100, 1 mm dithiothreitol (DTT), 1 mm phenylmethylsulphonyl fluoride (PMSF), 5 μg ml−1 E-64 and 5 μg ml−1 leupeptin. Cell lysates and culture supernatants of infected Sf21 cells and non-infected control cultures were subjected to 10% SDS-PAGE under reducing conditions. Fractionated proteins were electrophoretically transferred onto Hybond-C membranes. The blots were incubated with a mouse monoclonal antibody to the enterokinase recognition sequence (Invitrogen). Detection of bound antibodies was achieved with goat anti-mouse immunoglobulin G antibodies conjugated to horseradish peroxidase (Jackson ImmunoResearch, Soham, Cambridgeshire, UK) using the Supersignal West Pico Chemiluminescent substrate (Pierce, Rockford, IL, USA).

Purification of recombinant Arabidopsis thaliana GMII

Culture supernatants (200 ml) of AtGMII-producing Sf21 cells were cleared by low-speed centrifugation and then dialysed against 2 × 2 l of 10 mm sodium phosphate buffer, pH 7.0, 40 mm NaCl and 0.02% NaN3. After addition of 20 mm imidazole, the retentate was centrifuged for 30 min at 20 000 g. The resulting supernatant was loaded onto a 5-ml column of Chelating Sepharose (Amersham Biosciences, Freiburg, Germany) charged with Ni2+ ions, equilibrated in the same buffer. The matrix was washed successively with 40 and 80 mm imidazole prior to elution of the bound enzyme with 250 mm imidazole in dialysis buffer. Protein-containing eluate fractions were pooled and dialysed two times against 1 l of 50 mm MES buffer, pH 6.5, containing 150 mm NaCl and 0.02% (w/v) NaN3. After concentration by ultrafiltration and addition of proteinase inhibitors (1 mm PMSF, 5 μg ml−1 E-64 and 5 μg ml−1 leupeptin), the purified enzyme was analysed by SDS-PAGE and Coomassie Blue staining. Enzymatic deglycosylation of recombinant AtGMII using peptide N-glycosidase F (PNGase F; Roche, Mannheim, Germany) was performed as described previously (Bencur et al., 2005).

Assays of GMII activity

Standard GMII activity assays of cell lysates (obtained as described above) and culture supernatants were performed in a total volume of 20 μl of buffer containing 5 mm p-nitrophenyl-α-D-mannoside as substrate and 40 mm MES (pH 5.8), 0.5% Triton X-100, 0.5 mm DTT, 0.5 mm PMSF, 2.5 μg ml−1 E-64, and 2.5 μg ml−1 leupeptin. Triton X-100 and the proteinase inhibitors PMSF, E-64 and leupeptin were omitted for assays using purified enzyme. After incubation at 37°C for an appropriate time, the reactions were stopped by addition of 100 μl of 0.4 m glycine buffer (pH 10.4) prior to spectrophotometric analysis at 405 nm. One unit (U) of GMII activity corresponds to the formation of 1 nmol reaction product per minute. The preparation of Man5Gn [Manα1-6(Manα1-3)Manα1-6(GlcNAcβ1-2Manα1-3)Manβ1-4GlcNAcβ1-4GlcNAc]-GP (GP denotes glycopeptide) and Man5 [Manα1-6(Manα1-3)Manα1-6(Manα1-3)Manβ1-4GlcNAcβ1-4GlcNAc]-GP substrates was described previously (Bencur et al., 2005). Synthesis of β1,2-xylosylated Man5Gn-GP was performed in a final volume of 200 μl of buffer (50 mm MES, pH 7.0, and 1 mg ml−1 BSA), containing 2 mm Man5Gn-GP and 10 mm UDP-xylose, and 3.2 U of purified recombinant A. thalianaβ1,2-xylosyltransferase (Bencur et al., 2005) for 16 h at 23°C. Product formation was monitored by MALDI-TOF mass spectrometry. The reaction product was purified by sequential ion-exchange chromatography with Dowex-1X8 and Dowex-50X8 (Sigma-Aldrich, Vienna, Austria) eluted in water. As a final preparation step, the sample was fractionated on a Superdex Peptide PE (7.5 × 300 mm) gel filtration column (Amersham Biosciences) eluted with 0.1 M NH4HCO3, using an Äkta Purifier chromatography system (Amersham Biosciences). To detect β1,2-xylosylated Man5Gn-GP in the fractions, aliquots were spotted onto thin layer chromatography (TLC) plates and stained with orcinol/H2SO4. Positive fractions were combined and lyophilized. The yield of purified product was determined by amino sugar analysis as described previously (Bencur et al., 2005).

Man5Gn-PA was produced as described by Altmann and März (1995). Its β1,2-xylosylated derivative was prepared essentially as described above for β1,2-xylosylated Man5Gn-GP. The gel filtration step was substituted by HPLC purification of the samples.

GMII activity assays with glycopeptide substrates were performed with 100 ng of purified GMII protein [diluted in 50 mm MES (pH 6.5), 150 mm NaCl and 1 mg ml−1 BSA] in a total volume of 10 μl of a buffer containing 0.1 mm glycopeptide as a substrate, 40 mm MES (pH 5.8) and 10 μm 2-acetamido-1,2-deoxynojirimycin. After incubation at 37°C for 30 min the reaction was stopped by the addition of 1 μl of swainsonine (100 μg ml−1) and heating to 95°C for 3 min. The assays were either immediately analysed or stored at −20°C. For MALDI-TOF mass spectrometry analysis, all samples were diluted 1:4 in deionized water and 1 μl of the diluted sample was mixed with 1 μl of matrix [2% (w/v) 2,5-dihydroxybenzoic acid in 70% acetonitrile], dried under mild vacuum and analysed using a linear time-of-flight mass spectrometer (Thermo Bioanalysis, Hemel Hempstead, UK) as described previously (Wilson et al., 2001). Specific activities were calculated based on peak areas from MALDI-TOF spectra. All samples were analysed at least twice. The peaks of the potassium and sodium adducts of each component were both included in the quantification of the results.

GMII activity assays with 0.03 mm pyridylaminated oligosaccharide substrates were performed as above, followed by HPLC analysis as described previously (Altmann et al., 1993).

Identification of T-DNA insertion lines

Arabidopsis thaliana T-DNA lines SALK_052443 and SALK_141821 from the SALK Institute (San Diego, CA, USA) were obtained via the Nottingham Arabidopsis Stock Centre (Nottingham, UK). Homozygous hgl1-1 (derived from SALK_052443) and hgl1-2 (SALK_141821) seedlings were selected from the progeny by PCR using the forward primers MII-6 (5′-AGAGACGCTTCACCTAATAAACAA-3′) and MII-1 (5′-GGTTGTTGGTTTCCCTTCTCTG-3′) and the reverse primers MII-2 (5′-CGATGTCTGCCGGTGAATAGTTTG-3′) and MII-11 (5′-GAAGCTTTAATGCCCTCTCC-3′) as well as a T-DNA specific primer LBa1 (5′- TGGTTCACGTAGTGGGCCATCG-3′). For RT-PCR, RNA was extracted from hgl1-1 and hgl1-2 seedlings using the SV Total RNA System (Promega, Mannheim, Germany) and reverse transcription was performed from 500 ng of total RNA using oligo(dT) primers and AMV-RT (Promega). A control PCR reaction was performed using primers FTB-9 (5′-GGTGCTCGGAATTTTCGTCTACA-3′) and FTB-10 (5′-TCATTCTCTTTGCAACTGGCTCTA-3′), which amplify a part of the core α1,3-fucosyltransferase B cDNA.

Preparation of N-linked glycans from wild type and hgl1 mutants and MALDI-TOF mass spectrometry

For total N-glycan analysis, 500 mg of fresh rosette leaves was ground and suspended in 2.5 ml of 5% (v/v) formic acid and 0.1 mg ml−1 pepsin. The slurry was incubated at 37°C for 20 h with occasional stirring. Insoluble material was then removed by centrifugation. Glycopeptides were enriched from the supernatant by cation exchange chromatography and gel filtration as described previously (Wilson et al., 2001). N-glycans were subsequently released from glycopeptides with peptide N-glycosidase A (Roche), purified by cation exchange chromatography, gel filtration and passage through a reversed phase matrix, and subsequently analysed by MALDI-TOF mass spectrometry.

Complementation of GMII-deficient hgl1-1 plants

A full-length cDNA fragment of A. thaliana GMII was amplified by PCR using the Turbo-pfu DNA polymerase (Stratagene) and primers MII-21 (5′-TATATCTAGATGCCGTTCTCCTCGTATATCG-3′) and MII-22 (5′-TATATCTAGACTTGTGAGGTCGCAGTTCAAG-3′). The resulting PCR product was extracted from an agarose gel using a Wizard gel purification kit (Promega), XbaI digested and cloned into the binary plant expression vector p17 which had been linearized by XbaI digestion and treated with shrimp alkaline phosphatase (Roche). For the construction of vector p17, a myc tag linker fragment containing a stop codon was ligated into the BamHI site of vector pPT2 (Strasser et al., 2005) to create vector pPT8. The whole CaMV 35S promoter expression cassette was excised from pPT8 and cloned into HindIII/EcoRI sites of pPZP221 (Hajdukiewicz et al., 1994) to create p17. This results in the expression of the AtGMII fragment with a C-terminal myc tag. Correct orientation of the AtGMII insert was confirmed by DNA sequencing. Agrobacterium tumefaciens strain UIA143 containing the pMP90 plasmid (Hamilton, 1997) was used to transform hgl1-1 knockout plants. Transformation was performed according to standard protocols for bacterial transformation (Clough and Bent, 1998) and primary transformants were selected for gentamycin resistance. Expression of AtGMII was confirmed by Western blot analysis as described above using a polyclonal anti-myc antibody (A-14; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and T2 plants were subjected to total N-glycan analysis.

Cloning of targeting sequences of AtGMII into TMV expression vector

The AtGMII CTS region and fragments thereof were amplified by PCR, SalI digested and ligated into a SalI-restricted p4GD-GFPmod tobacco mosaic virus expression vector (Dirnberger et al., 2002), which is derived from vector p4GD-pl (Casper and Holt, 1996). Correct orientation of the inserted fragments was confirmed by DNA sequencing.

A sequence of 276 nucleotides corresponding to the N-terminal 92 amino acids (the putative CTS region) of A. thaliana GMII was amplified by PCR with the forward primer MII-25 (5′-TATAGTCGACATGCCGTTCTCCTCGTATAT-3′) and the reverse primer MII-26 (5′-TATAGTCGACCCGATTGGATCTGGAGGTAAG-3′) to construct CTS-GFP (Figure 5). Similarly, a fragment of 216 nucleotides corresponding to the N-terminal 72 amino acids (putative CT domain) was amplified with the primer combination MII-25 and MII-27 (5′-TATAGTCGACGAGAGTGAGGAGGAAGAAGA-3′) to construct CT-GFP. A sequence of 150 nucleotides encoding the 50 N-terminal amino acids (putative C domain) was amplified using primers MII-25 and MII-28 (5′-TATAGTCGACATTGACTACGAGAGTTCGTTT-3′), resulting in the construct C-GFP. The complementary oligonucleotides MII-32 (5′-TCGACATGTTCATCTTCGCCAACTTCTTCGTCATCGCACTCACCGTCTCACTCCTCTTCTTCCTCCTCACTCTCG-3′) and MII-33 (5′-TCGACGAGAGTGAGGAGGAAGAAGAGGAzGTGAGACGGTGAGTGCGATGACGAAGAAGTTGGCGAAGATGAACATG-3′) encoding the 22 amino acids of the putative transmembrane (T) region were phosphorylated using T4 DNA kinase (Promega), annealed and ligated to the SalI-digested construct p4GD-GFPmod, resulting in the construct T-GFP. For the construction of TS-GFP, a fragment of 129 nucleotides corresponding to the 43 amino acids of the putative transmembrane domain and the stem region (TS) was amplified using the forward primer MII-30 (5′-TATAGTCGACATGTTCATCTTCGCCAACTTCTTCG-3′) and the reverse primer MII-26. A fragment of 96 nucleotides corresponding to 32 amino acids of the truncated N-terminal cytoplasmic tail (deletion of amino acids 2–41) and the putative transmembrane domain (C10T) was amplified using the primer combination MII-31 (5′-TATAGTCGACATGCCACGAAAACGAACTCTCGTAG-3′) and MII-27, resulting in the construct C10T-GFP (Figure 6).

In vitro transcription and inoculation of Nicotiana benthamiana plants

Nicotiana benthamiana plants were grown in a controlled growth chamber with a day and night temperature of 22°C, 50% humidity and a 16-h light period. The recombinant viral vectors were linearized by KpnI digestion and in vitro transcripts thereof were made using a RiboMax kit (Promega). RNA was used to inoculate the plants mechanically at the six-leaf stage. Symptoms of infections were visible 8–10 days after inoculation as leaf deformations, with some variable leaf mottling and growth retardation.

Subcellular localization of GMII-GFP fusion proteins by fluorescence microscopy

The tissue sections from TMV-infected plants were processed for in situ localization of GFP and indirect immunofluorescence analysis exactly as described previously (Essl et al., 1999). For detection of the Lewisa epitope, the rat monoclonal IgM antibody JIM84 was used in a 1:50 dilution and the secondary antibody, Cy3 conjugated goat anti-rat IgM (Jackson ImmunoResearch), was diluted 1:200. The polyclonal rabbit anti-BiP antiserum was used in a 1:500 dilution and detected with a 1:200 diluted Cy3 conjugated anti-rabbit F(ab)2–IgG specific antibody (Sigma-Aldrich).

Imaging was conducted on a Leica TCS SP II confocal laser scanning microscope (Leica, Wetzlar, Germany) equipped with an Ar/Kr mixed gas laser and a Gre/Ne laser. For screening purposes, a ×40 oil immersion objective was used, but, for the double labelling imaging, a ×60 or ×100 oil immersion objective was chosen. Images from the confocal system were imported into Adobe Photoshop 7.0 for coloration.

Other methods

The metal content of purified recombinant AtGMII was determined by means of inductively coupled plasma mass spectrometry as reported previously (Bencur et al., 2005). The protein content of cell lysates was determined by the Bradford method with the Bio-Rad Protein Assay kit (Bio-Rad, Glattbrugg, Switzerland), using bovine serum albumin as a standard. The protein content of purified protein samples was determined with the BCA Protein Assay Kit (Pierce). Densitometric analysis of Coomassie Blue-stained SDS-PAGE gels was performed using ImageQuaNT v4.2 software (Molecular Dynamics, Palo Alto, CA, USA).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank Barbara Svoboda for producing the insect cell culture and for excellent technical assistance; Serena Curiale for her involvement in the initial stages of GMII cloning; Friedrich Altmann and Daniel Kolarich (Department of Chemistry) for MALDI-TOF mass spectrometry analyses; Peter Bencur and Thomas Dalik (Department of Chemistry) for providing glycopeptide substrates; Stephan Hann and Gunda Köllensperger (Department of Chemistry) for ICP-MS analyses; Ulrike Vavra for RT-PCR analyses; Doug Kuntz (Ontario Cancer Institute, Toronto, Canada) for providing purified recombinant D. melanogaster GMII; Jürgen Denecke and Paul Knox (University of Leeds, Leeds, UK) for providing BiP and Lea antibodies, respectively. We also thank the Salk Institute Genomic Analysis Laboratory for providing the sequence-indexed Arabidopsis T-DNA insertion mutants. This work was financed in part by the EC Pharma-Planta Project (LSHB-CT-2003-503565).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Altmann, F. and März, L. (1995) Processing of asparagine-linked oligosaccharides in insect cells: evidence for α-mannosidase II. Glycoconj. J. 12, 150155.
  • Altmann, F., Kornfeld, G., Dalik, T., Staudacher, E. and Glössl, J. (1993) Processing of asparagine-linked oligosaccharides in insect cells. N-acetylglucosaminyltransferase I and II activities in cultured lepidopteran cells. Glycobiology, 3, 619625.
  • Bakker, H., Bardor, M., Molthoff, J.W. et al. (2001) Galactose-extended glycans of antibodies produced by transgenic plants. Proc. Natl Acad. Sci. USA, 98, 28992904.
  • Bencur, P., Steinkellner, H., Svoboda, B. et al. (2005) Arabidopsis thalianaβ1,2-xylosyltransferase: an unusual glycosyltransferase with the potential to act at multiple stages of the plant N-glycosylation pathway. Biochem. J. 388, 515525.
  • Boisson, M., Gomord, V., Audran, C., Berger, N., Dubreucq, B., Granier, F., Lerouge, P., Faye, L., Caboche, M. and Lepiniec, L. (2001) Arabidopsis glucosidase I mutants reveal a critical role of N-glycan trimming in seed development. EMBO J. 20, 10101019.
  • Casper, S.J. and Holt, C.A. (1996) Expression of the green fluorescent protein-encoding gene from a tobacco mosaic virus-based vector. Gene, 173, 6973.
  • Chui, D., Oh-Eda, M., Liao, Y.F., Panneerselvam, K., Lal, A., Marek, K.W., Freeze, H.H., Moremen, K.W., Fukuda, M.N. and Marth, J.D. (1997) Alpha-mannosidase-II deficiency results in dyserythropoiesis and unveils an alternate pathway in oligosaccharide biosynthesis. Cell, 90, 157167.
  • Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735743.
  • Colley, K.J. (1997) Golgi localization of glycosyltransferases: more questions than answers. Glycobiology, 7, 113.
  • Denecke, J., Goldman, M.H., Demolder, J., Seurinck, J. and Botterman, J. (1991) The tobacco luminal binding protein is encoded by a multigene family. Plant Cell, 3, 10251035.
  • Dirnberger, D., Bencur, P., Mach, L. and Steinkellner, H. (2002) The Golgi localization of Arabidopsis thalianaβ1,2-xylosyltransferase in plant cells is dependent on its cytoplasmic and transmembrane sequences. Plant Mol. Biol. 50, 273281.
  • van den Elsen, J.M., Kuntz, D.A. and Rose, D.R. (2001) Structure of Golgi alpha-mannosidase II: a target for inhibition of growth and metastasis of cancer cells. EMBO J. 20, 30083017.
  • Essl, D., Dirnberger, D., Gomord, V., Strasser, R., Faye, L., Glössl, J. and Steinkellner, H. (1999) The N-terminal 77 amino acids from tobacco N-acetylglucosaminyltransferase I are sufficient to retain a reporter protein in the Golgi apparatus of Nicotiana benthamiana cells. FEBS Lett. 453, 169173.
  • Giraudo, C.G. and Maccioni, H.J. (2003) Endoplasmic reticulum export of glycosyltransferases depends on interaction of a cytoplasmic dibasic motif with Sar1. Mol. Biol. Cell, 14, 37533766.
  • Gomez, L. and Chrispeels, M.J. (1994) Complementation of an Arabidopsis thaliana mutant that lacks complex asparagine-linked glycans with the human cDNA encoding N-acetylglucosaminyltransferase I. Proc. Natl Acad. Sci. USA, 91, 18291833.
  • Hajdukiewicz, P., Svab, Z. and Maliga, P. (1994) The small, versatile pPZP family of Agrobacterium binary vectors for plant transformation. Plant Mol. Biol. 25, 989994.
  • Hamilton, C.M. (1997) A binary-BAC system for plant transformation with high-molecular-weight DNA. Gene, 200, 107116.
  • Horsley, D., Coleman, J., Evans, D., Satiat-Jeunemaitre, B. and Hawes, C. (1993) A monoclonal antibody, JIM 84, recognizes the Golgi apparatus and plasma membrane in plant cells. J. Exp. Bot. 44, 223229.
  • Johnson, K.D. and Chrispeels, M.J. (1987) Substrate specificities of N-acetylglucosaminyl-, fucosyl-, and xylosyltransferases that modify glycoproteins in the Golgi apparatus of bean cotyledons. Plant Physiol. 84, 13011308.
  • Katsuko, Y. (2002) Biosynthetic pathway of N-glycans. In Handbook of Glycosyltransferases and Related Genes (Taniguchi, N., Honke, K. and Fukuda, M., eds). Tokyo: Springer Verlag, pp. 625630.
  • Kaushal, G.P., Szumilo, T., Pastuszak, I. and Elbein, A.D. (1990) Purification to homogeneity and properties of mannosidase II from mung bean seedlings. Biochemistry, 29, 21682176.
  • Kawar, Z., Karaveg, K., Moremen, K.W. and Jarvis, D.L. (2001) Insect cells encode a class II alpha-mannosidase with unique properties. J. Biol. Chem. 276, 1633516340.
  • Kornfeld, R. and Kornfeld, S. (1985) Assembly of asparagine-linked oligosaccharides. Annu. Rev. Biochem. 54, 631664.
  • Leiter, H., Mucha, J., Staudacher, E., Grimm, R., Glössl, J. and Altmann, F. (1999) Purification, cDNA cloning, and expression of GDP-L-Fuc:Asn-linked GlcNAc alpha1,3-fucosyltransferase from mung beans. J. Biol. Chem. 274, 2183021839.
  • Lerouge, P., Cabanes-Macheteau, M., Rayon, C., Fischette-Laine, A.C., Gomord, V. and Faye, L. (1998) N-glycoprotein biosynthesis in plants: recent developments and future trends. Plant Mol. Biol. 38, 3148.
  • Lerouxel, O., Mouille, G., Andeme-Onzighi, C., Bruyant, M.P., Seveno, M., Loutelier-Bourhis, C., Driouich, A., Höfte, H. and Lerouge, P. (2005) Mutants in DEFECTIVE GLYCOSYLATION, an Arabidopsis homolog of an oligosaccharyltransferase complex subunit, show protein underglycosylation and defects in cell differentiation and growth. Plant J. 42, 455468.
  • Lowe, J.B. and Marth, J.D. (2003) A genetic approach to mammalian glycan function. Annu. Rev. Biochem. 72, 643691.
  • Misago, M., Liao, Y.F., Kudo, S., Eto, S., Mattei, M.G., Moremen, K.W. and Fukuda, M.N. (1995) Molecular cloning and expression of cDNAs encoding human alpha-mannosidase II and a previously unrecognized alpha-mannosidase IIx isozyme. Proc. Natl Acad. Sci. USA, 92, 1176611770.
  • Moremen, K.W. (2002) Alpha mannosidase II. In Handbook of Glycosyltransferases and Related Genes (Taniguchi, N., Honke, K. and Fukuda, M., eds). Tokyo: Springer Verlag, pp. 600606.
  • Nilsson, T. and Warren, G. (1994) Retention and retrieval in the endoplasmic reticulum and the Golgi apparatus. Curr. Opin. Cell Biol. 6, 517521.
  • Nilsson, T., Rabouille, C., Hui, N., Watson, R. and Warren, G. (1996) The role of the membrane-spanning domain and stalk region of N-acetylglucosaminyltransferase I in retention, kin recognition and structural maintenance of the Golgi apparatus in HeLa cells. J. Cell Sci. 109, 19751989.
  • Oh-eda, M., Nakagawa, H., Akama, T.O., Lowitz, K., Misago, M., Moremen, K.W. and Fukuda, M.N. (2001) Overexpression of the Golgi-localized enzyme α-mannosidase IIx in Chinese hamster ovary cells results in the conversion of hexamannosyl-N-acetylchitobiose to tetramannosyl-N-acetylchitobiose in the N-glycan-processing pathway. Eur. J. Biochem. 268, 12801288.
  • Palacpac, N.Q., Yoshida, S., Sakai, H., Kimura, Y., Fujiyama, K., Yoshida, T. and Seki, T. (1999) Stable expression of human β1,4-galactosyltransferase in plant cells modifies N-linked glycosylation patterns. Proc. Natl Acad. Sci. USA, 96, 46924697.
  • Rabouille, C., Kuntz, D.A., Lockyer, A., Watson, R., Signorelli, T., Rose, D.R., van den Heuvel, M. and Roberts, D.B. (1999) The Drosophila GMII gene encodes a Golgi alpha-mannosidase II. J. Cell Sci. 112, 33193330.
  • Sarkar, M., Pagny, S., Ünligil, U., Joziasse, D., Mucha, J., Glössl, J. and Schachter, H. (1998) Removal of 106 amino acids from the N-terminus of UDP-GlcNAc: alpha-3-D-mannoside β1,2-N-acetylglucosaminyltransferase I does not inactivate the enzyme. Glycoconj. J. 15, 193197.
  • von Schaewen, A., Sturm, A., O'Neill, J. and Chrispeels, M.J. (1993) Isolation of a mutant Arabidopsis plant that lacks N-acetyl glucosaminyl transferase I and is unable to synthesize Golgi-modified complex N-linked glycans. Plant Physiol. 102, 11091118.
  • Stanley, P. and Ioffe, E. (1995) Glycosyltransferase mutants: key to new insights in glycobiology. FASEB J. 9, 14361444.
  • Strasser, R., Altmann, F., Mach, L., Glössl, J. and Steinkellner, H. (2004) Generation of Arabidopsis thaliana plants with complex N-glycans lacking β1,2-linked xylose and core α1,3-linked fucose. FEBS Lett. 561, 132136.
  • Strasser, R., Stadlmann, J., Svoboda, B., Altmann, F., Glössl, J. and Mach, L. (2005) Molecular basis of N-acetylglucosaminyltransferase I deficiency in Arabidopsis thaliana plants lacking complex N-glycans. Biochem. J. 387, 385391.
  • Tezuka, K., Hayashi, M., Ishihara, H., Akazawa, T. and Takahashi, N. (1992) Studies on synthetic pathway of xylose-containing N-linked oligosaccharides deduced from substrate specificities of the processing enzymes in sycamore cells (Acer pseudoplatanus L.). Eur. J. Biochem. 203, 401413.
  • Wee, E.G., Sherrier, D.J., Prime, T.A. and Dupree, P. (1998) Targeting of active sialyltransferase to the plant Golgi apparatus. Plant Cell, 10, 17591768.
  • Wilson, I.B. (2002) Glycosylation of proteins in plants and invertebrates. Curr. Opin. Struct. Biol. 12, 569577.
  • Wilson, I.B., Zeleny, R., Kolarich, D., Staudacher, E., Stroop, C.J., Kamerling, J.P. and Altmann, F. (2001) Analysis of Asn-linked glycans from vegetable foodstuffs: widespread occurrence of Lewis a, core alpha1,3-linked fucose and xylose substitutions. Glycobiology, 11, 261274.

GenBank accession number of AtGMII cDNA DQ029214; Protein id: AAY90120

Arabidopsis seed stock lines Salk_052443, SALK_141821