Deficiency in human lysosomal α-mannosidase (MAN2B1) results in α-mannosidosis, a lysosomal storage disorder; patients present a wide range of neurological, immunological, and skeletal symptoms caused by a multisystemic accumulation of mannose-containing oligosaccharides. Here, we describe the expression of recombinant MAN2B1 both transiently in Nicotiana benthamiana leaves and in the leaves and seeds of stably transformed N. tabacum plants. After purification from tobacco leaves, the recombinant enzyme was found to be N-glycosylated and localized in vacuolar compartments. In the fresh leaves of tobacco transformants, MAN2B1 was measured at 10 200 units/kg, and the purified enzyme from these leaves had a specific activity of 32–45 U/mg. Furthermore, tobacco-produced MAN2B1 was biochemically similar to the enzyme purified from human tissues, and it was internalized and processed by α-mannosidosis fibroblast cells. These results strongly indicate that plants can be considered a promising expression system for the production of recombinant MAN2B1 for use in enzyme replacement therapy.
Alpha-mannosidases are ubiquitous enzymes involved in the biosynthesis and turnover of N-linked glycoproteins. These enzymes are classified, according to their amino acid sequences and biochemical properties, into two glycosylhydrolase families: 47 (EC 18.104.22.168) and 38 (EC 22.214.171.124/EC 126.96.36.199) (Henrissat and Davis, 1997). Acidic α-mannosidases (EC 188.8.131.52) are hydrolases found in plants (Hossain et al., 2009), animals (Herscovics, 1999), and micro-organisms (Santacruz-Tinoco et al., 2010). They can be localized to the cytoplasm, vacuoles, or lysosomes using various cell transport pathways. In animal cells, these hydrolases are transported to lysosomes by the mannose 6-phosphate pathway (Thomas, 2001). In plants, vacuolar α-mannosidase is targeted to its final destination via the classic secretory pathway involving the ER–Golgi system (Faye et al., 1998), whereas in yeast, vacuolar delivery of α-mannosidase can be mediated by both cytoplasm to vacuole targeting (Cvt) and autophagy pathways (Hutchins and Klionsky, 2001).
Lysosomal α-mannosidase is an exoglycosidase that cleaves α-linked mannose residues during the degradation of the N-linked glycans of glycoproteins. The predominant activity in most tissues has an acidic pH optimum of 3.5–4.5, is relatively stable, is activated by Zn2+, and is located in the lysosomal fraction of cells (Daniel et al., 1994). The substrate specificity of lysosomal α-mannosidase from human, rat, bovine, and feline liver has been studied in vitro. Each of these enzymes can hydrolyse α-1,2, α-1,3, and α-1,6 mannosidic linkages in oligosaccharides derived from N-linked glycans of the glycoproteins, albeit at different rates. In all tissues investigated, the enzyme exists as two major forms, A and B, which can be separated by ion-exchange chromatography despite being immunologically identical. Deficiency in human lysosomal α-mannosidase (MAN2B1) results in α-mannosidosis (MIM#248500), a lysosomal storage disorder characterized by a wide range of neurological, immunological, and skeletal symptoms caused by a multisystemic accumulation of mannose-containing oligosaccharides (Thomas, 2001). α-Mannosidosis occurs in approximately 1 of 500 000 live births and is expected to be found in any ethnic group anywhere in the world (Malm and Nilssen, 2008). The disease is known to occur also in guinea pigs (Crawley et al., 1999), cattle (Hocking et al., 1972), and cats (Vandevelde et al., 1982). In addition to these naturally occurring animal models for the human disorder, a mouse model was generated by targeted disruption of the α-mannosidase gene (Stinchi et al., 1999). The gene encoding human lysosomal α-mannosidase (MAN2B1; GenBank accession number U60266.1) has been assigned to the centromeric region of chromosome 19, and α-mannosidosis is caused by mutations in MAN2B1. The human enzyme is synthesized as a polypeptide of 1011 amino acids that is post-translationally modified by N-glycosylation, disulphide bridge formation, proteolysis, zinc binding, and homodimer formation (Nilssen et al., 1997; Tollersrud et al., 1997). As a result, intracellularly, the enzyme is composed of several proteolytic fragments with molecular weights ranging from 15 to 70 kDa. The precursor of MAN2B1 is partly secreted and taken up by neighbouring and distant cells via mannose 6-phosphate receptors on the cell surface (Pohlmann et al., 1983). The fact that cells can acquire lysosomal α-mannosidase from extracellular sources indicates that enzyme replacement may be an effective therapy for α-mannosidosis. Enzyme replacement therapy (ERT) and bone marrow transplantation (BMT) are the major therapeutic options in lysosomal storage disorders (Neufeld, 2004; Dobrenis, 2004). The few attempts at BMT in human α-mannosidosis have produced variable outcomes but have indicated that successful engraftment can lead to improved language, social and motor skills (Will et al., 1987; Wall et al., 1998; Malm, 2004). ERT is an effective means to improve the clinical manifestations in type I Gaucher’s disease (Barton et al., 1991) and has been approved for several lysosomal storage disorders, including some disorder affecting the brain (Neufeld, 2004). To evaluate the efficacy of ERT in human α-mannosidosis, Roces et al. (2004) showed the correction of oligosaccharide storage in a mouse model of α-mannosidosis after intravenous administration of mouse recombinant MAN2B1. To date, there is not a system for the production of recombinant MAN2B1 available on a large scale.
Plants are considered to be promising bioreactor platforms for the production of recombinant proteins (Ma et al., 2003; Hellwig et al., 2004). The plant-based recombinant protein production system has significant advantages as a high-level expression system having rapid scalability, ease and speed of genetic manipulation, lower endotoxin levels than microbial systems, and lack of risk of horizontal transmission and contamination by mammalian pathogens. In addition, this system retains the eukaryotic protein modification machinery (e.g. proteolytic and glycosylation systems) (Bardor et al., 2009), and it has the potential for economical production, even if very few examples of commercialization of plant-made recombinant proteins are known (see for example Nandi et al., 2005; Witcher et al., 1998). An interesting example of a plant-based platform for the large-scale production of biopharmaceutical proteins is that of glucocerebrosidase. This human lysosomal enzyme was first produced and stored in transgenic tobacco seeds (Reggi et al., 2005). Recently, glucocerebrosidase was also produced in a carrot cell suspension culture (Shaaltiel et al., 2007; Aviezer et al., 2009). These latter results will probably lead to the commercialization of taliglucerase alfa, a plant-cell-expressed form of glucocerebrosidase for the potential treatment of Gaucher’s disease. Reading these reports, we decided to investigate the possibility of MAN2B1 production in plants for a therapeutic approach. In this study, the recombinant enzyme was shown to be N-glycosylated and localized in the vacuolar compartment. The purified enzyme was biochemically characterized and evaluated for uptake in human α-mannosidosis fibroblast cells.
A plant secretory signal peptide is required to obtain MAN2B1 expression
The tobacco nuclear genome was transformed by A. tumefaciens using the vector pROK8.(SP)MAN2B1 (Figure S1). The full-length MAN2B1 cDNA, including 144 nucleotides coding for its native N-terminal signal peptide, was cloned into the pROK8 vector. The 48-amino acid stretch of signal peptide in mammal cells can translocate the MAN2B1 polypeptide to the ER which is the starting point of MAN2B1 traffic to the lysosomes. Unfortunately, even though the MAN2B1 cDNA harboured by the transgenic pROK8.(SP)MAN2B1 tobacco plants was integrated into the genome (data not shown) and showed a good transcription efficiency with an expected mRNA length around 3.0 kb (Figure S1), no detectable levels of the corresponding enzyme were obtained by Western blot or enzymatic assays. We hypothesized that the native signal peptide of MAN2B1 was responsible for the lack of recombinant protein accumulation in plants. Indeed, the expression of human glucocerebrosidase (GCD), another lysosomal enzyme, was obtained in plant by the replacement of the GCD native signal peptide with that of a secretory plant protein (Shaaltiel et al., 2007). To test our hypothesis, two other vectors were assembled that differ only in the type of MAN2B1 signal peptide encoded (Figure 1). In pDHA.(sp1)MAN2B1, the original MAN2B1 signal peptide was replaced by 30 amino acids coding for the tobacco pathogenesis–related protein 1 b (PRl) signal peptide, a protein secreted in tobacco mosaic virus-infected tobacco leaves. In contrast, plasmid pDHA.(SP)MAN2B1 was identical to pDHA.(sp1)MAN2B1 except for the presence of the native MAN2B1 signal peptide. Both vectors encoded a FLAG epitope at the C-terminus of MAN2B1. We knew from previous results that the PR1 signal peptide can direct a heterologous protein to the ER of transgenic plants (De Virgilio et al., 2008). The importance of the signal peptide in MAN2B1 expression was studied in transiently transfected tobacco leaf protoplasts. Homogenates from these protoplasts were subjected to immunoblot analysis using an antibody specific for the MAN2B1 enzyme (Figure 2). Western blot analysis of the protoplasts transformed with the plasmid pDHA.(sp1)MAN2B1 revealed a band corresponding to the entire protein precursor of about 110 kDa, indicating that MAN2B1 was correctly synthesized (Figure 2, lane 2, peptide a), and a peptide likely deriving from its fragmentation with a molecular mass around 20 kDa (Figure 2, lane 2, asterisk). The recombinant form of MAN2B1 with the native signal peptide did not accumulate in the protoplasts transformed with the vector pDHA.(SP)MAN2B1 because the results of this transformation were identical to that of the wild-type protoplasts (Figure 2, lanes 4 and 1, respectively, where only contaminant peptides were detected). The MAN2B1 expression cassette was then transferred from plasmid pDHA.(sp1)MAN2B1 to the pGreenII binary vector, which can also be used for Agrobacterium transformation. As expected, protoplast transformed with this last plasmid too expressed MAN2B1 (Figure 2, lane 3). Therefore, signal peptide replacement proved to be the right strategy for MAN2B1 expression, and all of the following experiments reported herein refer to tobacco pGreen.(sp1)MAN2B1 transformants.
Expression of MAN2B1 in transformed N. tabacum and N. benthamiana plants
To verify the possibility of MAN2B1 accumulation in plants, tobacco plants were stably transformed with the pGreen.(sp1)MAN2B1 vector. Tobacco transformants could express the MAN2B1 cDNA at both the transcriptional level (Figure 3a) and the translational level (Figure 3b–c). Western blot analysis in reducing conditions of leaf total proteins with the anti-FLAG antibody revealed a band of about 110 kDa corresponding to the entire protein precursor, indicating that MAN2B1 was correctly synthesized (Figure 3b). Immunoblot analysis using an antibody specific for the MAN2B1 enzyme revealed several bands corresponding to the single chain precursor and four peptides deriving from its fragmentation (Figure 3c). As a positive control for antibody specificity, a purified recombinant mouse-produced MAN2B1 was used; this protein was largely isolated in its precursor form together with two fragmentation products (Roces et al., 2004). In tobacco leaves, the molecular mass of the precursor polypeptide is around 110 kDa (Figure 2C, peptide a), slightly lower than the 130-kDa mouse or human precursor (Figure 3c, peptide P). The four proteolytically processed forms of MAN2B1 showed apparent molecular masses of approximately 70, 40, 32, and 18 kDa (Figure 3c, peptides b, c, d, and e, respectively). The 70-kDa peptide typical of human MAN2B1 processing likely corresponded to the 70-kDa peptide detected in plant tissues. Moreover, tobacco seeds accumulated only the precursor enzyme and the 70- and 32-kDa fragments. The additional MAN2B1 peptides extracted from leaves with respect to seed MAN2B1 forms are probably the result of a tissue-specific proteolytic process. To evaluate MAN2B1 accumulation in the N. benthaminana transient expression system, preliminary experiments were conducted on N. benthamiana agroinfiltrated leaves. As for stable tobacco transformants, when the anti-FLAG antibody was used in Western blot experiments, the strongest staining-detected peptide had a molecular mass of about 110 kDa, indicating the correct synthesis of the full-length protein (Figure 4a). Moreover, we also tested whether the addition of the reducing agent β-mercaptoethanol (2-ME) to the protein extraction buffer increased the amount of extractable MAN2B1 enzyme, and this was the case (Figure 4a, compare lane 3 with lane 4). Thus, 2-ME was always used for MAN2B1 extraction from the plant tissues. MAN2B1 precursor polypeptide was processed in the same manner in the two Nicotiana expression systems because all four b–e peptides were detected in both N. benthamiana and tobacco leaves (Figure 4b, peptides c and e become visible after prolonged exposure of the filter).
MAN2B1 is N-glycosylated
To verify MAN2B1 import into the ER, its glycosylation status was analysed. MAN2B1 is a glycoprotein that contains eleven potential N-glycosylation sites. Most N-glycosylation sites are occupied, and both complex-type and high-mannose-type N-linked oligosaccharides are present (Nilssen et al., 1997). Addition of high-mannose-type N-glycans usually takes place in the ER, and we examined whether plant recombinant MAN2B1 had this type of sugar chain. Total protein extracts from transgenic protoplasts expressing MAN2B1 were subjected to digestion with endoglycosidase H (endo H) because this enzyme cleaves N-linked glycans of the high-mannose type. The MAN2B1 precursor was clearly digested (Figure 5a, lanes 3–4). This result confirms that recombinant MAN2B1 is imported into the ER where it is glycosylated. To understand whether the MAN2B1 precursor also had complex-type N-glycans, leaf protoplasts isolated from the same plants were treated for 16 h in vivo with or without tunicamycin. Tunicamycin is a potent inhibitor of N-glycosylation; therefore, if the MAN2B1 precursor also has complex-type N-linked oligosaccharides, we would expect to observe a MAN2B1 precursor upon treatment with tunicamycin with a molecular mass lower than that observed after endo H digestion. On the contrary, each polypeptide detected after two treatments showed the same SDS-PAGE mobility (Figure 5a, compare lane 2 with lane 4), suggesting that only high-mannose-type N-linked oligosaccharides are present on the MAN2B1 precursor. To investigate which processed fragments are glycosylated, total leaf extracts from transgenic plants expressing MAN2B1 were subjected to digestion with endo H and immunoblotted with the anti-MAN2B1 antiserum (Figure 5b). The 70- and 40-kDa peptides were glycosylated because they underwent a shift in SDS-PAGE mobility upon treatment with endo H (Figure 5b, peptides b and c, respectively), whereas the 32-kDa peptide was not digested, indicating that it is likely not glycosylated (Figure 5b, peptide d).
MAN2B1 is localized in the vacuole
Once it enters the secretory pathway, MAN2B1 will be secreted by default because it does not contain known plant sorting signals that would direct it to cell organelles. To investigate the possibility of MAN2B1 secretion from tobacco transformed cells, protoplasts were isolated from transgenic plants. After isolation and incubation overnight at 23 °C in the dark, protoplasts were pelleted and lysed with protein extraction buffer, and the supernatant derived from centrifugation was TCA precipitated (Figure 6). Aliquots of decreasing quantities of protein extracts (lanes 5–8) together with proportional amounts of the medium (lanes 1–4) were analysed by Western blot using anti-FLAG (Figure 6a) or anti-MAN2B1 antiserum (Figure 6b) to identify recombinant MAN2B1. No signal could be detected in the medium, indicating an absence of MAN2B1 secretion. The protein profile of both membranes corresponded to that obtained in Figure 3b-c. If MAN2B1 was not secreted outside the cell, it should be localized somewhere inside the cell. To determine MAN2B1 localization, an isopycnic sucrose gradient centrifugation was performed with a leaf homogenate to separate the different organelles according to their densities. Proteins from each fraction of this gradient were separated by SDS-PAGE and detected with anti-MAN2B1 antiserum. Figure 7 shows that the four peptides derived from MAN2B1 processing migrated at the top of the gradient, at a density of less than 1.10 g/mL. Cytosol and soluble vacuolar proteins, as a result of vacuoles breaking during homogenization, remain at the top of the gradient (Pedrazzini et al., 1997). Therefore, MAN2B1 fragments should be localized in the cytosol or in the vacuole. A small portion of the MAN2B1 precursor was also detected at the top of the gradient (Figure 7, lane 2), but the majority migrated in the pellet, together with a fraction of the 70-kDa peptide (Figure 7, lane 20). The polypeptides detected in the pellet are likely to be insoluble because they are associated with organellar membranes. The MAN2B1 precursor was expected to localize mainly in the ER fraction, corresponding to a density of 1.17 g/mL, but this was not the case, even after long membrane exposure. To better understand the cell compartment in which MAN2B1 peptides accumulated, leaf cells of MAN2B1 plants were analysed by immunoelectron microscopy with anti-MAN2B1 or anti-FLAG antiserum. Although the anti-FLAG antibody poorly labelled in the same manner the leaf sections of both transformed and untransformed plants (data not shown), several electron-opaque aggregates were detected inside the vacuole of the transformed cells by the anti-MAN2B1 antiserum (Figure 8c–d). Small structures (in the size range 0.10–0.25 μm), localized at the boundary between the cytoplasm and the vacuole, were also labelled by the antiserum (Figure 8e–f). This result, together with those of the isopycnic sucrose gradient, indicates that the MAN2B1 recombinant protein accumulates in the vacuole in processed forms. No labelled structures were visualized in the wild-type cells (Figure 8a–b) or when preimmune serum was used (data not shown). Therefore, it is reasonable to hypothesize that recombinant MAN2B1 is cotranslationally imported into the ER, and then the precursor protein rapidly traffics along the secretory system to the vacuole, where it is proteolysed.
Purification and biochemical characterization of MAN2B1 from tobacco leaves
In wild-type tobacco, 0.2 units of endogenous α-mannosidase/g fresh leaves was detected, with a specific activity of 0.03 units/mg, whereas in transformed tobacco, 10.2 units of α-mannosidase/g fresh leaves was measured, with a specific activity of 1.15 units/mg. Heat treatment and a combination of anion-exchange and affinity chromatography were used to purify recombinant MAN2B1 from transformed tobacco leaves (Table 1). DEAE elution profiles of recombinant MAN2B1 from tobacco revealed the separation of the total activity into two major peaks: A and B (Figure 9), like in human tissues (Daniel et al., 1994). Both peaks were retained by a column of concanavalin-A-Sepharose (ConA), and all of the activity was recovered as a single peak after elution with α-methylmannoside. Thus, both isoenzymes are glycoproteins. Table 1 clearly shows that there is a progressive increase in MAN2B1 specific activity from the heat treatment step (7.51 units/mg) to the anion-exchange chromatography step (30 and 11 units/mg for peak A and peak B, respectively) and then to the ConA column (45 and 32 units/mg for peak A and peak B, respectively). As shown in Figure 10a (lane 3), two main polypeptides with molecular masses of around 70 and 32 kDa were detected when the purified MAN2B1 isoenzyme A was analysed by SDS-PAGE. A third very faint polypeptide of 110 kDa was also detected in the purified fraction. Western blot analysis of the same purified samples confirmed that the recombinant MAN2B1 obtained after purification was represented largely by the two peptides of 70 and 32 kDa derived from MAN2B1 processing (peptides b and d, respectively). Furthermore, Western blot analysis shows the presence of the MAN2B1 precursor (peptide a) and the 40-kDa fragment (peptide c) (Figure 10b, lane 4). The overall yield from this purification was approximately 34% (24% of isoenzyme A plus 10% of isoenzyme B), and the specific activity of the final preparation was 45 units/mg for the isoenzyme A and 32 units/mg for the isoenzyme B (Table 1). Isoenzymes A and B showed an acidic pH optimum near 4.5, similar to their human orthologs. The recombinant MAN2B1 isoenzymes showed a high thermostability at 70 °C for 1 h (Figure 11) compared with that of the endogenous α-mannosidase, which lost all activity after 15 min at 70 °C (data not shown). The two recombinant isoenzymes had Km values (isoenzyme A, 0.88 mm and isoenzyme B, 0.65 mm) similar to that of MAN2B1 purified from CHO cells (0.75 mm) (Berg et al., 2001).
Table 1. Purification of human recombinant MAN2B1 from tobacco leaf
Total protein (mg)
Total activity (units)
Specific activity (units/mg)
Crude extract (2 g of leaf)
Heat treatment (60 °C, 20 min)
DEAE anion exchange, isoenzyme A
DEAE anion exchange, isoenzyme B
Concanavalin-A-Sepharose, Isoenzyme A
Concanavalin-A-Sepharose, Isoenzyme B
Uptake of recombinant α-mannosidase in α-mannosidosis fibroblasts
We then investigated whether the plant recombinant MAN2B1 enzyme could be taken up by human cells. The isoenzyme A purified fraction shown in the SDS-PAGE of Figure 10a, lane 3 (the corresponding Western blot is shown in Figure 10b, lane 4) was used for this assay. The addition of 0.48 units of purified recombinant MAN2B1 to α-mannosidosis Gm00654 fibroblasts resulted in an eightfold increase in enzyme activity compared with that of the controls (Table 2). Proteins extracted from these fibroblasts were analysed by Western blot using anti-MAN2B1 antiserum (Figure 12). The results indicated that the anti-MAN2B1 antiserum detected the 130-kDa native MAN2B1 precursor in the control fibroblast cell lines (Figure 12, arrow, lanes 1 and 2), whereas in the α-mannosidosis Gm00654 cell line, the native precursor was absent (Figure 11, arrow, lane 3). Moreover, the antiserum detected the plant-derived MAN2B1 precursor in the α-mannosidosis cell line supplemented with increasing amounts of purified recombinant enzyme (Figure 12, arrow, lanes 4–6). The amount of recombinant MAN2B1 precursor was reduced with the increased amount of 70-kDa peptide (Figure 12, arrowhead, lanes 4–6). This 70-kDa peptide could be also clearly detected in the control fibroblast cell lines (Figure 12, arrowhead, lanes 1 and 2), indicating that it is the most abundant fragment derived from native MAN2B1 precursor processing in these cells. Therefore, the plant-made MAN2B1 precursor was correctly processed in the α-mannosidosis Gm00654 cell line in the same manner as the native precursor in the control fibroblast cell lines. This result implies that the recombinant MAN2B1 precursor, clearly detectable in MAN2B1 purified samples (Figure 10b, lane 4), was taken up by the α-mannosidosis cell line. Conversely, the other purified MAN2B1 fragments derived from the processing of the enzyme in the plant tissues (Figure 10b, lane 4, peptides b, c, d) did not appear to be imported into the human cells.
Table 2. α-Mannosidase activity in wild-type or α-mannosidosis fibroblasts
α-Mannosidase activity (mU/mL per min)†
*The α-mannosidosis fibroblasts were incubated with different units of recombinant ΜΑΝ2Β1 for 48 h.
†Each value represents the mean ± SD of three independent experiments.
0.65 ± 0.02
α-Mannosidosis (Gm 00654)
0.064 ± 0.004
α-Mannosidosis (Gm 00654) + 0.016 U
0.07 ± 0.003
α-Mannosidosis (Gm 00654) + 0.16 U
0.17 ± 0.05
α-Mannosidosis (Gm 00654) + 0.48 U
0.51 ± 0.04
In this study, we demonstrated that tobacco-produced MAN2B1 can be taken up by enzyme-deficient fibroblasts. MAN2B1 expression was obtained both transiently in N. benthamiana leaves and in the leaves and seeds of stable transformed N. tabacum plants. In tobacco, MAN2B1 was imported into the ER of plant cells because it is N-glycosylated, a typical ER post-translational modification. The importance of changing the original MAN2B1 signal peptide for a plant signal peptide to have MAN2B1 expression was clearly demonstrated in our results. In tobacco leaves, the 110-kDa MAN2B1 precursor reached the vacuole, where it was processed mainly into two polypeptides of 70 and 32 kDa, although minor fragments of 40 and 18 kDa were also detected. The human enzyme in the placenta is synthesized as a precursor of 130 kDa and processed into three peptides of 70, 42, and 15 kDa. The 70-kDa peptide is partially proteolysed into three more peptides of 39, 10, and 22 kDa, which are joined together by disulphide bridges (Nilssen et al., 1997). The small difference in molecular weight between the recombinant plant MAN2B1 precursor and the placenta enzyme precursor could be ascribed to a different glycosylation pattern in plants with respect to the human tissues. In fact, in transfected COS cells (African-green-monkey kidney cells), recombinant MAN2B1 was synthesized as a 120-kDa precursor because it contained only high-mannose-type N-glycans (Hansen et al., 2004). In the same way, MAN2B1 proteolysis yields different fragments according to the type of cells in which the recombinant enzyme is accumulated. For example, when MAN2B1 is expressed in Chinese hamster ovary (CHO) cells, it is secreted as a 130-kDa precursor that is then proteolysed into only two polypeptides of 72 and 55 kDa (Berg et al., 2001), whereas in COS cells, it is cleaved into three peptides of 70, 40, and 15 kDa (Hansen et al., 2004). However, in CHO cells, there are no significant differences between the biochemical properties of the processed and precursor forms, suggesting that MAN2B1 processing may not have a functional role (Berg et al., 2001). MAN2B1 purified from tobacco leaves was largely isolated in its processed forms of 70 and 32 kDa, but this enzyme preparation also contained the 110-kDa precursor form. The proportions of the precursor and processed forms recovered after MAN2B1 purification vary according to the method used and the organism or tissue in which they are expressed (Berg et al., 2001; Roces et al., 2004). The enzyme-specific activity reported for CHO cells is 38–43 U/mg, similar to the 30 U/mg described for the enzyme purified from placenta (Liao et al., 1996; Nilssen et al., 1997). We have shown in this study that the recombinant enzyme purified from tobacco leaves has a specific activity of 32–45 U/mg, which is comparable to the values reported for CHO cells and placenta, and that it has the same biochemical properties of the enzyme from human tissues. These results indicate that the structural and functional characteristics of the recombinant tobacco-produced MAN2B1 are similar to those of the human enzyme. This is also supported by the fact that the 110-kDa precursor of the plant-derived recombinant MAN2B1 can be internalized and processed to a 70-kDa peptide by α-mannosidosis fibroblast cells, strongly indicating that plants can be considered a promising expression system for the production of recombinant MAN2B1 to be used in ERT. Furthermore, the amount of MAN2B1 purified by chromatography from tobacco transformants was determined to be 3500 units/kg of fresh leaves. Following purification from CHO culture medium, 145 units/L of recombinant MAN2B1 was recovered (Berg et al., 2001), and recombinant human acid β-glucosidase content in tobacco seed was reported to be 720 units/kg after immunochromatography (Reggi et al., 2005). According to these results, we believe that the tobacco leaf system can be considered a highly efficient expression system for the production of recombinant MAN2B1.
Contrary to what has been reported for GCD, no plant storage vacuole targeting signal was added to the recombinant MAN2B1 in this study because this enzyme was efficiently targeted to the vacuole thanks to its own targeting signals. Alternatively, there is the remote possibility that the FLAG epitope we added to the C-terminal of recombinant MAN2B1 acted as a vacuolar sorting signal. Soluble lysosomal enzymes are routed to the lysosomes of mammalian cells by mannose 6-phosphate receptors (MPRs), which bind to the exposed mannose 6-phosphate moiety on glycosylated enzymes in the trans Golgi network (Storch and Braulke, 2005). The phosphorylated oligosaccharide side chains act as targeting signals for the lysosomal compartment. N-glycosylated plant proteins are not phosphorylated in the Golgi apparatus, and MPRs have not been reported from plants (Gaudrault and Beevers, 1984). Moreover, it is generally believed that N-linked glycans have no specific role in the targeting of plant glycoproteins to the vacuole (Lerouge et al., 1998). Therefore, the targeting signal utilized by the N-glycoprotein MAN2B1 to reach the plant vacuole could not have been its phosphorylated mannose residues (even if we did not experimentally verify the presence of 6-phosphate linked to recombinant MAN2B1). The import of externally added plant recombinant MAN2B1 into fibroblast cells raises the important question of how MAN2B1 can be transported through the cell membrane, but it is known that terminal mannose residues are sufficient for internalization as demonstrated for the recombinant human GCD expressed in the carrot cells (Shaaltiel et al., 2007).
In summary, it is possible to purify from plant cells a recombinant lysosomal α-mannosidase with biochemical properties that are similar to those of the native human enzyme and that can be taken up by α-mannosidosis fibroblasts. These results pave the way for further studies aimed at improving the production and purification of plant MAN2B1 for the development of effective and safe ERT.
Materials and methods
A schematic representation of the DNA vectors used in this study is provided in Figures 1 and S1. The entire cDNA sequence coding the human lysosomal alpha-mannosidase (MAN2B1) was inserted into the binary vector pROK8 which was derived from plasmid pBI131.1 (Jefferson et al., 1987). In detail, the MAN2B1 cDNA was PCR-amplified from the expression vector pcDNA3.1-MAN2B1 (Berg et al., 2001) using the forward primer P1 (5′-GGCTCGTCGACATGGGCGCCTACGCGCGGGCTTCGGGG-3′) and the reverse primer P2 (5′- GGCTCGGTACCCTAACCATCCACCTCCTTCCATTGAAC-3′), to insert a SalI and a KpnI restriction site (underlined) at the 5′ and 3′ end of the MAN2B1 gene, respectively. The resulting PCR product was cleaved with SalI and KpnI and cloned into the SalI/KpnI sites of pROK8 under the control of the promoter of the small subunit of Rubisco (rbcS) and the nopaline synthase (NOS) terminator, thus obtaining pROK8.(SP)MAN2B1.
In the pDHA.(sp1)MAN2B1 vector, the native endoplasmic reticulum (ER) signal peptide of the MAN2B1 gene was substituted with the ER signal peptide of the tobacco PR1 (pathogenesis-related) protein (Marusic et al., 2007). The MAN2B1 cDNA was amplified from pROK8.(SP)MAN2B1 using the forward primer P3 (5′-CTAGTCTAGAGGGGGATACGAGACATGCCCCACAGTG-3′) and the reverse primer P4 (5′-CTAGATGCATCTACTTATCGTCGTCATCCTTGTAATCACCATCCACCTCC-TTCCATTGAACTGAGGCCAG-3′), thus inserting XbaI and NsiI restriction sites (underlined) at the 5′ and 3′ end of the PCR product, respectively. Moreover, the P4 primer carries the sequence of the FLAG epitope (in bold) that, in the resulting MAN2B1 protein, was inserted at the C-terminus just before the stop codon. The PCR product, which was cut with XbaI/NsiI (NsiI has compatible ends with PstI), was inserted into the XbaI/PstI sites of the pDHA.nef–zein vector (De Virgilio et al., 2008). In pDHA.nef–zein, the NEF–zein fusion has the PR1 signal peptide. In this way, the entire cassette coding for the NEF–zein fusion protein, except for the PR1 signal peptide, was replaced by MAN2B1. The resulting intermediate plasmid was named pDHA.(sp1)MAN2B1. For the production of transgenic plants, the fragment excised by EcoRI digestion from pDHA.(sp1)MAN2B1, including the 35S promoter, the sequence coding for the MAN2B1 protein and the 35S terminator, was introduced into the EcoRI site of the pGreenII binary vector (Hellens Roger et al., 2000), thus obtaining the pGreen.(sp1)MAN2B1.
To obtain plasmid pDHA.(SP)MAN2B1, the first pDHA vector assembled, pDHA.(sp1)MAN2B1, was digested with the restriction enzymes BamHI and XbaI to excise the PR1 signal peptide. The native signal peptide (SP) was amplified from the vector pcDNA3.1-MAN2B1 (Berg et al., 2001) using the forward primer BamHI-MAN (5′-GACTTCGGATCCATGGGCGCCTACGCGCGGGCTTC-3′) and the reverse primer XbaI-MAN (5′-CATGACTCTAGAGGCCCGAGCACCGGCAGCCGCCAG-3′), thus inserting BamHI and XbaI restriction sites (underlined) at the 5′ and 3′ end of the PCR product, respectively. The PCR product, cut with XbaI/BamHI, was inserted into the XbaI/BamHI sites of the previously digested pDHA.(sp1)MAN2B1 vector, thus replacing the PR1 signal peptide.
Plasmid pROK8.(SP)MAN2B1, pDHA.(sp1)MAN2B1, and pDHA.(SP)MAN2B1 were sequenced at both strands to confirm correct amplification and ligation.
Agrobacterium tumefaciens strains LBA4404 or GV3101 were transformed by electroporation with pROK8.(SP)MAN2B1 or pGreen.(sp1)MAN2B1, respectively, and used to produce transgenic tobacco (Nicotiana tabacum) cv. Petit Havana SR1 as described (Bellucci et al., 2000). Briefly, leaf discs from N. tabacum were transformed by cocultivation with A. tumefaciens strains harbouring pROK8.(SP)MAN2B1 or pGreen.(sp1)MAN2B1. After cocultivation, the leaf discs were grown on MS regeneration medium (Murashige and Skoog, 1962) containing 1 mg/L N6-benzyladenine (BA), 0.1 mg/L α-naphthaleneacetic acid (NAA), 500 mg/L cefotaxime, and 100 mg/L kanamycin. Regenerated shoots of each explant were kept separated to guarantee the regeneration of independent transformants. Transformed plants were grown at 25 °C and 16-h light/8-h dark photoperiod in axenic conditions without antibiotics and propagated every 5–6 weeks. Ten pROK8.(SP)MAN2B1 and sixteen pGreen.(sp1)MAN2B1 plants were regenerated, and their transgenic status was confirmed by PCR. Three pGreen.(sp1)MAN2B1 plants (number 4, 6, and 11) were selected by Western blot as high-expressing MAN2B1 transformants and analysed by Southern blot to investigate the MAN2B1 gene copy number (data not shown). T0 seeds of these three plants were sown on agar-solidified MS medium with kanamycin (100 mg/L), and T1 transformed plants were obtained.
Agrobacterium tumefaciens strain GV3101, harbouring the binary vector pGreen.(sp1)MAN2B1, was cultured in 5 mL of liquid YEB with appropriate antibiotics. Agroinfiltration was performed on Nicotiana benthamiana leaves according to Sparkes et al. (2006). Four days after infiltration, agroinfiltrated leaves were collected and protein extraction was performed as described in the following sections.
Isolation and analysis of nucleic acids
DNA extraction was performed using the GenElute™ Plant Genomic DNA Kit (Sigma-Aldrich, St. Louis, MO). For Southern blot analysis, total DNA (10 μg) was digested overnight with HindIII, electrophoresed on 0.8% agarose gel, and transferred to a Hybond-XL nylon membrane (GE Healthcare, Chalfont St. Giles, UK), according to the manufacturer’s instructions. Total RNA was extracted with the RNAeasy Plant Mini Kit (QIAGEN N.V., Venlo, The Netherlands). RNA (10 μg) was electrophoretically separated on 1.4% formaldehyde agarose gels and transferred to Hybond-N nylon membranes (GE Healthcare, Chalfont St. Giles, UK), according to the manufacturer’s instructions. MAN2B1 cDNA was used as probe for both Southern and Northern blots. Hybridization was performed for both DNA and RNA filters as indicated by the membrane supplier with 32P-labelled probes and the Ready-To-GoTM kit (GE Healthcare, Chalfont St. Giles, UK).
Leaves of transgenic and wild-type tobacco or Nicotiana benthamiana plants were homogenized in an ice-cold mortar with homogenation buffer (200 mm NaCl, 1 mm EDTA, 0.2% Triton X-100, 100 mm Tris-HCl, pH 7.8) supplemented with ‘Complete’ protease inhibitor cocktail (Roche, Basel) with or without 4%β-mercaptoethanol (2-ME). The homogenate was centrifuged at 14 000 g for 10 min at 4 °C. Total protein contained in the supernatant was measured by Bradford assay (Bradford, 1976). When indicated, endoglycosidase H (endo H) digestion of total proteins was performed as described previously (Frigerio et al., 2001). Extracted proteins were analysed by SDS-PAGE followed by Western blot. Proteins were electrotransferred to a Hybond-P membrane (Amersham 13 Biosciences, Little Chalfont, UK) and detected using antinative MAN2B1 antibody (1:2000 dilution) or anti-FLAG (1:1000 dilution) antiserum. Protein bands were visualized with peroxidase-linked goat anti-rabbit secondary antibody diluted 1:20 000 (Pierce Chemical, Rockford, IL) using SuperSignal West Pico chemiluminescent substrates (Pierce Chemical). Prestained protein molecular weight markers (Fermentas, Vilnius, Lithuania and EuroClone, Milano, Italy) were used as molecular mass markers.
Protoplast isolation, pulse-chase labelling, and immunoprecipitation
Protoplasts, prepared from the young leaves of wild-type or transgenic plants as described by Pedrazzini et al. (1997), were resuspended at a concentration of 1 × 106 cells/mL and incubated overnight at 23 °C in the dark. To investigate MAN2B1 secretion in the medium, they were diluted with four volumes of W5 buffer (154 mm NaCl, 5 mm KCl, 125 mm CaCl2·2H2O, and 5 mm Glc) and then centrifuged at 500 g for 20 min at 4 °C. The resulting pellet containing intact protoplasts was nitrogen frozen and then resuspended in 0.8 mL of homogenation buffer. Unless otherwise stated, protoplast homogenation was performed by adding to frozen samples ice-cold homogenation buffer (150 mm Tris-HCl, 150 mm NaCl, 1.5 mm EDTA, and 1.5% Triton X-100, pH 7.5) and 4% of 2-ME, supplemented with ‘Complete’ (Roche, Basel) protease inhibitor cocktail. The supernatant, derived from centrifugation, containing secreted proteins was precipitated with trichloroacetic acid (TCA), and the resulting pellet was dissolved in 0.8 mL of the homogenetion buffer. After SDS-PAGE, proteins were analysed by immunoblot as described elsewhere. For transient protein expression, protoplasts were isolated from small leaves of wild-type tobacco SR1 plants and subjected to polyethylene glycol–mediated transfection using 40 μg of plasmid DNA as described by Pompa and Vitale (2006). After overnight recovery, protoplasts were homogenated and subjected to Western blot analysis. For treatment with tunicamycin, 50 μg/mL of tunicamycin (from a 5 mg/mL of 10 mm NaOH stock solution; Boehringer Ingelheim, http://www.boehringer-ingelheim.com) or equivalent quantities of solvent for the controls were added to the incubation medium and the protoplasts incubated for 16 h at 23 °C in the dark. When indicated, endoglycosidase H (endo H) digestion of total proteins from protoplasts was performed as described previously (Frigerio et al., 2001).
For isopycnic gradient, young leaves of transgenic tobacco encoding MAN2B1 were homogenized with homogenization buffer (12% of sucrose, 10 mm KCl, 100 mm tris-Cl pH 7.8, and 2 mm MgCl2) with 4% 2-ME. A continuous sucrose gradient, between 16% and 55%, was prepared using the same buffer, and 600 μL of the MAN2B1 homogenate was loaded on the top of the gradient. After centrifugation at 28 000 rpm for 4 h at 4 °C in a Beckman SW28 rotor, fractions of 750 μL were collected. An equal aliquot of each fraction (40 μL) was analysed by SDS-PAGE and Western blot as described elsewhere with anti-MAN2B1 antiserum.
Electron and fluorescence microscopy
For electron microscopy, immunolocalization was performed as described by Mainieri et al. (2004). Ultrathin sections were mounted on 300-mesh nickel grids and immunogold-labelled. Grids were then incubated with anti-MAN2B1 (1:5000 dilution) or anti-FLAG (1:1000 dilution) antisera. Controls were incubated in preimmune rabbit serum. Grids were examined under an electron microscope (EM 400 T; Philips, Eindhoven, The Netherlands).
Recombinant MAN2B1 was extracted from 2 g of tobacco leaves as described for the protein extraction method but without 2-ME in the homogenization buffer. MAN2B1 enzyme activity was determined in the supernatant obtained after centrifugation using the fluorogenic substrate 4-methylumbelliferyl-α-D-mannopyranoside (4MUαMann) as described by Beccari et al. (1997). For the assay, 50 μL of sample was mixed with 100 μL of a 3 mm solution of substrate in 0.2 m sodium acetate buffer, pH 4.5. Then, after a period of incubation at 37 °C lasting between 30 min and 2 h, the reaction was stopped by adding 2.85 mL of an ice-cold 0.2 m glycine-NaOH buffer, pH 10.4. Fluorescence of the liberated 4-methylumbelliferone was measured on a Perkin-Elmer LS-3 fluorimeter (excitation, 360 nm; emission, 446 nm). One unit is the amount of enzyme that hydrolyses 1 μmol of substrate/min at 37 °C.
Protein extracts were kept at 60 °C for 20 min. The precipitate was removed by centrifugation at 10 000g for 10 min. The supernatant was dialysed overnight against 10 mm sodium phosphate buffer, pH 6.0, and then loaded onto a DEAE-cellulose column (2 mL volume) and equilibrated with 10 mm sodium phosphate buffer, pH 6.0. Unretained protein was eluted with the column buffer, and then a linear gradient of NaCl (0–0.6 m in 40 mL of column buffer) was applied. Finally, the column was washed with 1.0 m NaCl in the same buffer. Each fraction was tested for MAN2B1 activity. Following DEAE-cellulose chromatography, pooled fractions were dialysed against the 20 mm Tris/HCl buffer, pH 7.4, containing MnCl2 1 mm, MgCl2 1 mm, CaCl2 1 mm, and NaCl2 0.5 m and then applied to a Concanavalin-A-Sepharose affinity column (4 cm length × 0.5 cm diameter) equilibrated with the dialysis buffer. A linear gradient containing α-methyl-D-mannoside (0–0.5 m) in 20 mm Tris/HCl buffer, pH 7.4, and NaCl2 0.5 m was applied. Finally, the column was eluted with 1.0 m NaCl in the same buffer.
Km values were determined by Lineweaver and Burk (1934) plots. Samples of purified isoenzymes (50 μL) were incubated with 4MUαMann in the range 0.066–0.528 mm in 0.2 m sodium acetate buffer, pH 4.5. Optimal pH conditions were determined using 3 mm 4MUαMann in 0.1 m/0.2 m citric acid/sodium phosphate buffers and 0.2 m sodium acetate in the pH range 3.0–6.5. Thermal stability was determined by incubating samples of purified recombinant MAN2B1 isoenzyme A and B (50 μL) for various periods of time (5′, 15′, 30′, and 60′) at 50, 60, and 70 °C. Samples were cooled on ice for 2 h and then assayed for MAN2B1 activity as previously described.
Uptake of recombinant α-mannosidase in α-mannosidosis fibroblasts
α-Mannosidosis fibroblasts (ATCC Gm00654) were cultured to confluency in DMEM (Gibco) supplemented with GlutaMAX (Gibco), 10% FBS (Gibco), and 60 U/mL penicillin-G and 60 μg/mL streptomycin (Gibco) at 37 °C in 5% CO2, split 1:2 and grown overnight before being supplemented with increasing amounts of total recombinant α-mannosidase (0.016, 0.16, and 0.48 U). Cells were harvested after 48 h using M-PER (Mammalian Protein Extraction Reagent, Pierce) supplemented with Complete Mini EDTA-free protease inhibitor cocktail (Roche). The lysed cells were centrifuged, and the lysate was assayed for α-mannosidase activity as described earlier. Next, 10 μg of protein was denatured with NuPAGE Sample Reducing agent (Invitrogen, Life Technologies Corp., Carlsbad, CA) and LDS loading buffer (Invitrogen), analysed by SDS-PAGE (NuPage Novex Bis-Tris 4–12% polyacrylamide gel; Invitrogen) under reducing conditions, and electroblotted onto a PVDF membrane (Invitrolon™; Invitrogen) using the Novex mini cell system (Invitrogen). MAN2B1 was detected using a 1:7500 dilution of antinative human recombinant MAN2B1 antibody as primary antibody and alkaline phosphatase–conjugated chicken anti-rabbit antibody (sc-2967, Santa Cruz Biotechnology Inc., Santa Cruz, CA) diluted 1:2000 as secondary antibody. The membrane was developed using Tropix CDP-star (Applied Biosystems, Carlsbad, CA), and immunoreactive bands were visualized using the FUJIFILM Luminescent Image Analyzer LAS-3000 (Fuji photo film, Co., Ltd., Fujifilm Holdings Corp., Tokyo, Japan).
We thank Giancarlo Carpinelli, Rita Scimmi, and Silvia Ticconi for technical support and Dr. Andrea Porceddu and Dr. Sergio Arcioni for their initial encouragement to carry on this study. The antinative MAN2B1 antibody was generously provided by Zymenex. This work was supported by the COST Action FA0804, Molecular Farming: Plants as a Production Platform for High Value Proteins. We also thank Fondazione Cassa di Risparmio di Città di Castello, Perugia, Italy, for financial support. Contribution no. 362 from the Institute of Plant Genetics, Research Division of Perugia, CNR.