• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntphysLink to Publisher's site
Plant Physiol. Nov 2004; 136(3): 3457–3466.
PMCID: PMC527145
Focus Issue on ER-Derived Compartments in Plants

Unexpected Deposition Patterns of Recombinant Proteins in Post-Endoplasmic Reticulum Compartments of Wheat Endosperm1


Protein transport within cereal endosperm cells is complicated by the abundance of endoplasmic reticulum (ER)-derived and vacuolar protein bodies. For wheat storage proteins, two major transport routes run from the ER to the vacuole, one bypassing and one passing through the Golgi. Proteins traveling along each route converge at the vacuole and form aggregates. To determine the impact of this trafficking system on the fate of recombinant proteins expressed in wheat endosperm, we used confocal and electron microscopy to investigate the fate of three recombinant proteins containing different targeting information. KDEL-tagged recombinant human serum albumin, which is retrieved to the ER lumen in leaf cells, was deposited in prolamin aggregates within the vacuole of endosperm cells, most likely following the bulk of endogenous glutenins. Recombinant fungal phytase, a glycoprotein designed for secretion, was delivered to the same compartment, with no trace of the molecule in the apoplast. Glycan analysis revealed that this protein had passed through the Golgi. The localization of human serum albumin and phytase was compared to that of recombinant legumin, which contains structural targeting information directing it to the vacuole. Uniquely, legumin accumulated in the globulin inclusion bodies at the periphery of the prolamin bodies, suggesting a different mode of transport and/or aggregation. Our results demonstrate that recombinant proteins are deposited in an unexpected pattern within wheat endosperm cells, probably because of the unique storage properties of this tissue. Our data also confirm that recombinant proteins are invaluable tools for the analysis of protein trafficking in cereals.

Storage proteins accumulate during seed development and are then broken down and mobilized during germination in order to provide the embryo with carbon, nitrogen, and sulfur (Müntz, 1998; Herman and Larkins, 1999). There are two major groups of seed storage proteins: the alcohol-soluble prolamins and the salt-soluble globulins (Shewry and Casey, 1999). Prolamins are the predominant form of storage protein in most cereals, and they are deposited within the seed endosperm cells in patterns characteristic for each species. For example, maize and rice prolamins aggregate into dense protein bodies within the rough endoplasmic reticulum (ER) and remain attached to this organelle (Shotwell and Larkins, 1989; Coleman and Larkins, 1999; Muench et al. 1999). In contrast, wheat and oat prolamins are stored in protein bodies that are sequestered into large vacuoles (Shotwell and Larkins, 1989; Galili et al., 1993).

The major classes of wheat prolamins are the gliadins and glutenins, the latter responsible for the elasticity of wheat dough (Shewry, 1995; Shewry and Casey, 1999). Gliadins accumulate in protein bodies as monomers and assemble via noncovalent interactions, whereas glutenins assemble via noncovalent interactions and intermolecular disulfide bonds. There are two distinct classes of glutenins, the high-molecular weight (HMW) class and the low-molecular weight (LMW) class (Shewry, 1995). Wheat also contains small amounts of an 11S globulin homolog, triticin, which represents approximately 5% of the total seed protein (Singh et al., 1991). This is stored in dense inclusions found at the surface of prolamin protein bodies (Bechtel et al., 1991).

Although numerous studies have considered the transport of wheat storage proteins to the vacuole, the precise route remains elusive. Levanony et al. (1992) demonstrated that significant amounts of the wheat prolamins are transported to the vacuole via a route that bypasses the Golgi complex. These prolamins assemble in protein bodies within the ER (Parker, 1982; Parker and Hawes, 1982; Levanony et al., 1992; Rubin et al., 1992; Altschuler et al., 1993), which then bud off and translocate to the vacuole membrane, allowing the protein bodies to be internalized by a process analogous to autophagy (Levanony et al., 1992; Galili et al., 1993). However, immunocytochemical studies of developing seeds have also shown prolamins in Golgi-associated vesicles, suggesting a second transport route that does involve the Golgi complex (Parker and Hawes, 1982). Since the number of Golgi organelles and transcripts for Golgi-associated proteins declines with seed maturation, it has been suggested that the role of Golgi vesicles in the trafficking of wheat prolamins is mainly restricted to the early stages of development (Parker, 1982; Galili et al., 1993; Shy et al., 2001).

Wheat prolamins are generally nonglycosylated, and any low level of glycosylation that may occur is very difficult to detect (Shewry, 1996). This hinders the use of Golgi-specific glycan modifications to ascertain the extent to which the Golgi complex is involved in their transport. The general consensus is that some prolamins, principally gliadins, are transported to the vacuole via the Golgi complex, whereas others, principally glutenins, aggregate within the lumen of the ER and are subsequently incorporated into the same vacuole without passing through the Golgi. There they merge to form a homogeneous protein matrix as found in mature grains (Shewry and Halford, 2002). The origin of the autophagic vacuole is not entirely clear. However, the tonoplast integral proteins pyrophosphatase (V-PPase) and γ-TIP have been detected in the membranes surrounding the protein bodies inside this vacuole (Galili et al., 1996; Herman and Larkins, 1999). These membrane proteins are known as markers of the tonoplast of vegetative tissues in higher plants (Höfte et al., 1992; Rea and Poole, 1993).

To gain further insight into the various protein-trafficking pathways that exist in the wheat endosperm cell, and to assess how recombinant proteins respond to the unique intracellular sorting machinery therein, we investigated the fates of three recombinant proteins containing different targeting information. We show that an artificial reticuloplasmin (recombinant human serum albumin [HSA] containing an N-terminal signal sequence and a C-terminal KDEL tag for retrieval to the ER lumen) is delivered to prolamin bodies within the vacuole. Immunolocalization and glycoanalysis experiments demonstrate that a glycoprotein designed for secretion (phytase, with signal sequence present but no KDEL tag) is delivered to the same destination, mainly via the Golgi complex. We compare the localization of these two proteins with that of legumin, which contains structural information targeting it to the protein storage vacuole. This protein accumulated within specific regions of the inclusion bodies at the periphery of the prolamin bodies (Stoger et al., 2001). Our results show that all three recombinant proteins eventually accumulate in the same compartment, albeit in specific subareas. These data allude to the existence of different routes for protein trafficking within the endosperm, some of which may be unique to endosperm cells. What is also clear from these results is that the deposition patterns observed for each recombinant protein are unexpected given the nature of the targeting information each carries.


Identification and Characterization of Protein Bodies

Protein bodies in the wheat endosperm are sequestered within a large, central vacuole, predominantly in the cells of the subaleurone layer (Fig. 1, A and B). Since the aleurone cells divide periclinally during endosperm development, the cells in the subaleurone layer are the youngest and contain more protein in relative terms than the older cells, which occupy the central part of the endosperm. Here the cells do not have a vacuole, as it has been reabsorbed, and they contain high levels of starch (Fig. 1A). For the purpose of this study, we focused on the cells within and immediately below the subaleurone layer.

Figure 1.
Wheat protein bodies. A, Light microscopy. Cross-section of a wild-type seed. Spurr semithin section, stained with methylene blue. Wheat protein bodies (arrow heads) can be seen in the large, central vacuole (v) of cells in the first layers of the endosperm ...

Different microscopy techniques were used in this study, making it necessary to develop procedures for the identification of protein bodies in each case. The immunolocalization of HMW glutenins using the IFRN 1602 monoclonal antibody (kindly provided by C. Mills, Institute of Food Research, Norwich, UK) allowed us to visualize and identify protein bodies in confocal microscopy experiments. Protein bodies were formed after the aggregation of individual prolamin bodies, together with some triticin inclusion bodies (Fig. 1C). The signal obtained after labeling corresponds to the x-type HMW subunits of glutenin (Mills et al., 2000), which generally appeared to be distributed uniformly within the protein body, with the exception of a few dark, well-defined round areas that corresponded to inclusion bodies. No clusters or special distribution patterns were observed (Fig. 1C).

Electron microscopy showed that protein bodies were formed from prolamin bodies, round structures with medium electron density, together with some smaller inclusion bodies with higher electron density (Fig. 1D) previously identified as triticin-containing entities (Bechtel et al., 1991). The whole protein body was shown to be surrounded by a membrane or a system of membranes (Fig. 1D). To reveal the membranes around and perhaps within the protein body, we used the thiocarbohydrazide (TCH)-silver proteinate method for the detection of unsaturated lipids on wild-type wheat seed sections. After treatment with TCH followed by silver proteinate, a fine, electrodense precipitate was formed on the membranes, enhancing their contrast. It can be seen clearly that the protein body is surrounded by a membrane or system of membranes that also contains the inclusion bodies (Fig. 1, E and F). As expected, there were no membranes or any trace of lipids within the prolamin matrix or the inclusion bodies (Fig. 1, E and F). Nevertheless, there were membranes between the aggregating prolamin bodies, which were also observed using general, nonspecific contrasting (Fig. 1D). Surprisingly, no membranes were observed between the inclusion bodies and the prolamin bodies, since their interface appears precipitate-free (Fig. 1, E and F). The TCH-silver proteinate technique also revealed the presence of some myelin-like structures associated with the protein bodies (Fig. 1E). As protein bodies mature, the individual prolamin bodies coalesce to form a large, single structure so that the mature protein body comprises a homogeneous proteinaceous matrix containing scattered inclusion bodies. Those inclusion bodies can be found either on the surface of the prolamin body or embedded within it (Fig. 2, D, G–J).

Figure 2.
Localization of recombinant HSA. A, Fluorescence microscopy. Spurr semithin section. Strong labeling can be seen in the protein bodies (arrows). Note the labeling in several small, individual prolamin bodies. B and C, Confocal microscopy. Tangential (B) ...

A Recombinant Protein Containing a Leader Sequence and ER Retrieval Signal Accumulates in Prolamin Bodies

In order to explore the fate of an artificial reticuloplasmin in wheat endosperm, we expressed a recombinant protein, HSA, containing an N-terminal signal sequence and a C-terminal ER retrieval signal. Seeds were collected from each transgenic line and tested by immunoblot analysis using an anti-HSA antibody. This revealed a 66-kD band, the expected molecular weight of HSA, as well as two minor degradation fragments (Fig. 5A, lanes 1–5). The 66-kD band was also detected using an antibody against the KDEL tag (Fig. 5A, lane 6). One representative plant with a seed HSA content of approximately 0.5% total soluble protein, as estimated by semiquantitative immunoblot analysis, was selected for further immunolocalization studies.

Figure 5.
Immunoblot analysis. A, Detection of HSA in endosperm extracts from single seeds. Lanes 1 and 7, Nontransformed seeds; lanes 2 to 6, transgenic seeds containing HSA. Detection with anti-HSA antibody (lanes 1–5) or with anti-KDEL antibody (lanes ...

The distribution of recombinant HSA in seeds was unexpected: The protein was found to accumulate in the vacuole along with the storage protein aggregates (Fig. 2A). More significantly, HSA was also detected within individual prolamin bodies still in the cytoplasm and not yet incorporated in the vacuole (Fig. 2F). Confocal microscopy clearly demonstrated the deposition of HSA in the protein bodies. More precisely, the confocal images showed a protein body clearly labeled and with four defined dark areas, indicating the presence of four different nonlabeled inclusion bodies, three on the surface (Fig. 2, D, G–I) and one embedded within the protein body (Fig. 2J). Electron microscopy confirmed this deposition pattern since the gold particles were only found in the large prolamin bodies, leaving the inclusion bodies label-free (Fig. 2E). In order to discard the possibility of a nonfunctional KDEL tag, we also investigated HSA localization in vegetative leaf cells (the transgene was expressed using a constitutive promoter). HSA was effectively retained within the cells, which showed dense labeling of ER-like structures, and no signal was present in the apoplast or the vacuole (Fig. 2, B and C).

As a comparison, Figure 3 shows the deposition of recombinant pea legumin in the inclusion bodies. In a previous investigation, we established by immunogold labeling that recombinant pea legumin expressed in wheat seeds accumulated within the inclusion bodies (Stoger et al., 2001). In this study, we compared the distribution of recombinant HSA and recombinant legumin by immunofluorescence and confocal microscopy, expecting a complementary pattern of labeling. Our expectations were confirmed, as shown in Figure 3, A and B.

Figure 3.
Localization of recombinant pea legumin. A, Confocal microscopy. 3-D reconstruction from optical confocal sections. Pea legumin is deposited within the inclusion bodies (arow heads). B, Optical section from the protein body in (A). Pea legumin is located ...

A Recombinant Protein with a Leader Sequence but No Retrieval Signal Is Not Secreted and Accumulates in Protein Bodies

In order to define the default secretory pathway for phytase in wheat endosperm cells, we carried out further immunolocalization studies in transgenic seeds expressing fungal phytase carrying an N-terminal signal sequence but lacking a KDEL tag. Transgenic wheat plants expressing the recombinant phytase were screened by immunoblot, and the highest expressing line was selected for immunolocalization analysis. The recombinant phytase should enter the endomembrane system and follow the secretory pathway, finally arriving in the apoplast. However, a general view of the transgenic endosperm tissue, labeled with antiphytase antibodies, showed no trace of labeling in the apoplast, suggesting that the phytase enzyme is not secreted. Instead, and contrary to our expectations, dense labeling was found once again in the vacuolar and cytosolic protein bodies showing that at least some of the phytase was retained within the endosperm cell and had accumulated with the storage proteins (Fig. 4A). Electron microscopy provided a more detailed characterization of the phytase deposition pattern in the protein bodies. The recombinant protein was found to accumulate preferentially in the prolamin bodies, since the inclusion bodies showed no labeling (Fig. 4, B and C). Gold particles could also be seen defining the border between a prolamin and an inclusion body, indicating that phytase is completely excluded from the inclusion bodies (Fig. 4C).

Figure 4.
Localization of recombinant phytase. A, Fluorescence microscopy. LR White semithin section. Phytase accumulates in the protein bodies (pb). Note the absence of a signal in the apoplast (arrows). B and C, Electron microscopy. LR White thin sections. Significant ...

A Recombinant Protein Following the Secretory Pathway Is Transported through the Golgi Apparatus in Wheat Endosperm Cells

Given the unexpected localization of the recombinant phytase, we next wanted to address the issue of how this protein was routed through the cell, and in particular whether it passed through the Golgi complex. Since phytase is a glycoprotein, the obvious way to find the answer was to scrutinize its glycan structure for Golgi-specific modifications. Since the phytase gene was expressed using an endosperm-specific promoter, we were able to extract recombinant phytase originating from endosperm cells only, without contamination from other tissues.

After partial purification and concentration, the protein sample was separated by SDS PAGE, and immunoblot analysis was carried out using antiphytase antiserum and a lectin from Aleuria aurantia that binds specifically to Fuc linked (α-1,6) or (α-1,3) to N-acetylglucosamine. Figure 5B clearly shows that the band corresponding to the recombinant phytase was detected by the lectin, indicating that at least a proportion of the phytase is modified in the trans-Golgi. The phytase band was excised from the gel and its identity was confirmed by peptide mass fingerprinting of the tryptic peptides (Fig. 6). In this fingerprint, one glycopeptide (121–145) was shown to carry MMXF and MUXF glycans, confirming Asn 123 as an N-glycosylation site. The other potential glycopeptides were not detected either as modified or unmodified forms. The detailed analysis of the N-linked glycans revealed the presence of xylosylated and fucosylated paucimannosidic structures (Fig. 6). The predominant glycan structure was MMXF (Man α-1-6(Man α 1-3)(Xyl β 1-2)Man β 1-4GlcNAc β 1-4(Fuc α 1-3)GlcNAc). Paucimannosidic-type N-glycans such as MMXF, MUXF (Man α 1-6(Xyl β 1-2)Man β 1-4GlcNAc β 1-4(Fuc α 1-3)GlcNAc), and MMX (Man α-1-6(Man α 1-3)(Xyl β 1-2)Man β 1-4GlcNAc β 1-4 GlcNAc) are typical for vacuolar glycoproteins (Lerouge et al., 1998). Approximately 30% of the N-glycans we detected were of the oligomannose type, which are not further modified in the Golgi. This minority of glycans could either correspond to sites that are not accessible to the Golgi enzymes or they could reflect the presence of phytase that has not passed through the trans-Golgi. We did not detect significant amounts of complex-type glycans with terminal GlcNAc residues, as typically found on secreted glycoproteins.

Figure 6.
Glycoproteomic analysis of endosperm-derived phytase. The upper section presents the sequence of phytase with identified tryptic peptides shown in bold and underlined; the signal peptide is shown in italic; and potential glycopeptides after tryptic digestion ...


There are several routes by which proteins can reach the lumen of the protein storage vacuole in plant cells (Vitale and Galili, 2001; Brandizzi and Hawes, 2004). They may follow the classical route, passing through the Golgi complex and exiting at the level of the trans-Golgi compartments, or they may segregate as early as the cis-Golgi (Hillmer et al., 2001). Alternative routes, bypassing the Golgi apparatus, involve vesicles that are released from the ER and become internalized into vacuoles (Hara-Nishimura et al., 1998) or that fuse directly with the storage vacuoles (Toyooka et al., 2000).

A similar Golgi-independent pathway for the delivery of storage proteins has long been known to exist in wheat endosperm cells (Levanony et al., 1992). ER-derived prolamin bodies up to 10 μm in diameter are sequestered into autophagic provacuoles of unclear origin. This additional route found in wheat endosperm contrasts with trafficking pathways in other species, where proteins are delivered into the vacuole mainly by vesicles that fuse with the protein body membrane (Müntz, 1998; Herman and Larkins, 1999).

The complexity of the protein-trafficking machinery in wheat endosperm cells has been widely acknowledged (Galili et al., 1993, 1996; Shewry and Halford 2002). Many studies have dealt with wheat protein bodies, but even with these extensive investigations the formation of the protein bodies is still not fully understood. The past decade represents a gap in the morphological analysis of wheat storage protein accumulation. To our knowledge, this study is the first to show confocal images of developing wheat protein bodies, providing a unique insight into the structural features of these organelles. As a consequence of the autophagic process in wheat, the entire protein body is surrounded by a system of membranes. These are derived from individual limiting membranes of the prolamin bodies and the tonoplast (Levanony et al., 1992). Our experiments show that aggregating prolamin bodies are initially separated by a membrane, supporting the notion that the enlargement of wheat endosperm protein bodies occurs by coalescence or fusion (Levanony et al., 1992). Consequently, the membranes between merged prolamin bodies may be degraded and expelled from the protein body, where they form ordered myelin-like structures that remain associated with the protein body (Parker, 1980).

Inclusion bodies, containing triticin, are usually enclosed within the membrane delimiting the protein bodies. However, despite the use of a specific lipid-staining method, we failed to detect any membrane between the inclusion and prolamin bodies. This finding might indicate that the triticin inclusion bodies merge with the prolamin bodies in the vacuole by membrane fusion. Alternatively, triticin may be incorporated into the prolamin bodies before they are internalized into the autophagic vacuole. In any case, triticin occupies a separate domain within the protein body resulting in a biphasic structure, consistent with the findings previously described for wheat and oat (Shewry and Halford, 2002).

The atypical intracellular sorting machinery in wheat endosperm could be reflected by the trafficking of recombinant proteins. For example, it is possible that ER resident proteins might be incorporated into the prolamin aggregates passively and then transported toward the vacuole rather than remaining in the ER lumen. To test this hypothesis, we determined the localization of a recombinant protein (HSA) with a C-terminal KDEL sequence. It has been shown that the addition of a KDEL signal is usually sufficient to retrieve recombinant proteins to the ER lumen (Herman et al., 1990; Denecke et al., 1992; Wandelt et al., 1992). Crofts et al. (1999) showed that some leakage of KDEL-tagged proteins to the apoplast could occur due to saturation of the retrieval machinery. Frigerio et al. (2001) reported that a small proportion of KDEL-tagged phaseolin reached the vacuole of tobacco (Nicotiana tabacum) leaf cells via a Golgi-independent route, but most of the protein was retrieved to the ER. Therefore, it is reasonable to assume that recombinant HSA, an artificial reticuloplasmin, should be localized predominantly in the ER cisternae and ER bodies and should not extend beyond the Golgi. We confirmed this distribution in transgenic wheat leaves, confirming the functionality of the retrieval signal within the HSA sequence. In wheat endosperm, however, the immunofluorescence signal corresponding to HSA was localized in protein bodies within the vacuole. It has previously been shown that the ER resident chaperone BiP also accumulates in the vacuolar protein bodies of wheat (Levanony et al., 1992). A second chaperone, calreticulin, and a KDEL-tagged scFv antibody were shown to be localized to a minor extent within the protein storage vacuoles in rice, while most of the scFv was detected in prolamin bodies surrounded by an ER membrane (Torres et al., 2001). In wheat endosperm cells, the same KDEL-tagged antibody was detected mainly in the prolamin bodies within the vacuole (E. Arcalis, unpublished data), in agreement with our localization data for HSA. This distribution of KDEL-tagged recombinant proteins suggests that a substantial portion of those proteins is passively trapped in the prolamin aggregates and, following the bulk movement of glutenins, is eventually internalized into the vacuole.

It has been suggested that the Golgi-independent pathway may be more prevalent for glutenins accumulating later in seed development, whereas the Golgi-dependent pathway may be predominant in early development (Parker, 1982). This may account for the decreasing number of Golgi organelles in the maturing endosperm (Parker, 1982; Galili et al., 1993) and questions the importance of the Golgi apparatus in the production of endosperm proteins. To address this question, we expressed a recombinant glycoprotein (fungal phytase) containing a leader peptide for entry into the secretory pathway, but no retrieval signal. Phytase is a stable protein with 10 potential N-glycosylation sites (Wyss et al., 1999) and is therefore suitable as a model for protein trafficking. We used an endosperm-specific promoter active during seed development, allowing the purification of phytase exclusively from endosperm tissue, without contamination from the embryo or outer layers of the seed.

Immunolocalization experiments revealed that recombinant phytase did not accumulate in the apoplast, but was instead sequestered in the protein bodies, mostly within the vacuole. Two possibilities were considered to explain this observation: The protein might follow the bulk of the prolamins bypassing the Golgi, or it might be diverted to the vacuole after passing through the Golgi complex. To distinguish between these possibilities, we investigated the glycan profile of phytase, looking specifically for core α(1,3)-Fuc and β(1,2)-Xyl residues that represent late Golgi modifications. The prevalence of these residues suggested that the major part of the phytase had traveled to the vacuole via the Golgi complex. Approximately 30% of the N-glycans were of the oligomannose type. This minority of glycans may correspond to sites that are not accessible to the Golgi enzymes. Alternatively, they could reflect a proportion of phytase that has not passed through the trans-Golgi. While secreted glycoproteins contain terminal GlcNAc residues in addition to the core Fuc and Xyl, enzymatic cleavage of terminal GlcNAc occurs in the vacuole (Lerouge et al., 1998). Since the glycans we detected on the recombinant phytase were trimmed, this provided supporting evidence for the vacuolar localization of the protein and clearly showed that phytase had reached the trans-Golgi and was subsequently diverted to the vacuolar compartment. This was unexpected, since the phytase contained only a leader peptide and should be secreted. Indeed, we, as well as others, have observed efficient secretion in tobacco leaves (Verwoerd et al., 1995). We therefore postulate that this may be a feature specific to cereal endosperm. Given the specialized architecture of endosperm cells, which are designed for storage, it is conceivable that this will influence the intracellular route of recombinant proteins. Further experiments in our laboratory are in progress to characterize the fate of recombinant proteins in the cereal endosperm as compared to vegetative organs.

Thus far, only a few recombinant proteins have been localized in cereal endosperm. In rice, the list comprises a scFv antibody (Torres et al., 2001), lysozyme (Yang et al., 2003), and phaseolin (Zheng et al., 1995). None of these proteins was secreted, but the secretion of phaseolin was not expected because it is a protein storage vacuole resident protein, and the scFv contained a KDEL signal facilitating retrieval to the ER lumen. Among these three proteins, only phaseolin is glycosylated, and no precise tracking was performed. In maize, recombinant Escherichia coli toxin B was found in starch granules, contrary to the predicted extracellular secretion (Chikwamba et al., 2003). In wheat, only three recombinant proteins have been localized: HSA-KDEL, phytase (this study), and pea legumin (Stoger et al., 2001). Notably, pea legumin was the only one of the three recombinant proteins localized to the inclusion bodies. This is interesting, since neither an artificial reticuloplasmin (HSA-KDEL) nor a protein that passed through the Golgi (phytase) was found in this region. Hillmer et al. (2001) reported that, in pea seeds, legumin is sorted into distinct dense vesicles that segregate as early as the cis-Golgi. It is conceivable that such a mechanism may also be used to sort legumin in wheat. Alternatively, the segregation of legumin together with wheat globulin (triticin) may result from the phase separation between globulins and prolamins as described for cereals generally (Shewry and Halford, 2002).

In this investigation, we used transgenic wheat plants to study the trafficking of recombinant proteins in cereal seeds, which appear to possess a more complex and diverse protein-sorting machinery than other plant cells. Our data suggest that the unique features of different plant species and specialized tissues may significantly affect the localization and consequent modification of recombinant proteins. This is of practical importance in molecular farming applications, where the destination of the recombinant protein and its state of modification may influence protein recovery and activity. Our experiments also confirm that recombinant proteins are useful reporter molecules for the elucidation of trafficking pathways, since they can be utilized for in situ immunolocalization analysis and simultaneous glycan profiling to establish their intracellular trafficking route. Therefore, the analysis of recombinant proteins may help to clarify some of the remaining unsolved questions about protein trafficking in different plant tissues.


Vectors for Plant Transformation

The phytase expression construct (pLPL-phyA) was assembled by linking the Aspergillus niger phyA gene to a murine immunoglobulin leader peptide sequence and inserting the cassette into vector pTO126, which contained the rice glutelin-1 seed-specific promoter and the rice ADPGPP (ADP-Glc pyrophosphate) gene terminator (a kind gift from Dr. T. Okita; Washigton State University, Pullman, Washington). The leader sequence was amplified using forward primer 5′-GGATCCACTAGTACACAATCAGA-3′ and reverse primer 5′-GCGGCCGCTCTAGAGATGATAACTG-3′, and the PCR product was digested with BamHI and XbaI. The phyA gene (accession no. M94550; kindly provided by Dr. E Mullaney; USDA, New Orleans, Louisiana) was amplified using forward primer 5′-TGTAGAGTCACCTCCGGACTGGCAGTC-3′and reverse primer 5′-CCGCGGCTAAGCAAAACACTCCG-3′, and the PCR product was digested with XbaI. The two sequences were ligated together and amplified using forward primer 5′-GCAGCGGCCGCACACAATCAGA-3′ and reverse primer 5′-CATGCGGCCGCCTAAGCAAAACACTCC-3′. This product was digested with NotI and ligated into the vector.

The human HSA cDNA (accession no. A15293) was joined at the 5′ end to the murine immunoglobulin leader peptide sequence and at the 3′ end to a sequence encoding the tetrapeptide tag KDEL. The cassette was inserted into the EcoRI site of a pUC-based vector carrying the maize ubiquitin-1 promoter and first intron and the nos terminator. In all transformation experiments, plasmid pAHC20 (Christensen and Quail, 1996) was cointroduced for phosphinothricin selection.

Wheat Transformation

Immature wheat (Triticum aestivum L. cv Bobwhite) embryos were aseptically removed and cultured according to Drake et al. (2000). After 6 d on callus induction medium, embryos were bombarded using a Bio-Rad 2000 particle gun (1100 psi rupture disc; Bio-Rad Laboratories, Hercules, CA). Plasmids containing the phytase or HSA cDNAs, respectively, and the bar gene (pAHC20) were mixed for cotransformation at a 3:1 molar ratio and precipitated onto gold particles. Selection was carried out on medium containing phosphinothricin (Drake et al., 2000). Regenerated plantlets were screened for transgene expression by immunoblot analysis, and the best expressing lines were grown in the greenhouse and growth rooms at 15°C day/12°C night temperature with a 10-h photoperiod during the first 40 d, followed by maintenance at 21°C day/18°C night temperature with a 16-h photoperiod thereafter. Transgenic wheat plants expressing legumin were described previously (Stoger et al., 2001).

Immunoblot Analysis

Western-blot analysis was carried out according to standard protocols using commercially available antisera for HSA (Sigma, St. Louis). For the detection of phytase, a 1:2,000 dilution of rabbit antiserum was used, which was kindly provided by Dr. A. Ohmann (Novozyme, Kalundborg, Denmark). Rabbit antiserum against pea legumin was a kind gift from Dr. R. Casey (John Innes Centre, Norwich, UK). Biotinylated Aleuria aurantia lectin was obtained from Vector Laboratories (Burlingame, CA) and detected using alkaline phosphatase-labeled streptavidin.

Extraction and Glycan Analysis

Tissue samples were homogenized in extraction buffer (phosphate-buffered saline containing 10 mm ascorbic acid, 500 mm NaCl, and 5% β-mercaptoethanol) and centrifuged. Phytase in the supernatant was enriched by ammonium sulfate fractionation, dialyzed in phosphate-buffered saline, and concentrated by ultrafiltration before loading on a 10% (w/v) SDS-PAGE gel. The band corresponding to phytase was excised from the Coomassie-stained gel after electrophoresis under reducing conditions and subjected to tryptic digestion as described elsewhere (Katayama et al., 2001). The extracted and dried peptides were dissolved in water/acetonitrile/trifluoroacetic acid (95:5:0.1; v/v/v) and analyzed by MALDI-TOF mass spectrometry. Further preparation and mass spectrometry analysis of N-glycans were performed according to Kolarich and Altmann (2000).


Developing grains were bisected transversely and the embryo was removed. The half-grain formerly containing the embryo was processed for recombinant protein analysis (either HSA, phytase, or legumin) by western blot. The remaining half was fixed and processed for microscopy as described below.

Light and Electron Microscopy

Wild-type wheat seeds and seeds expressing recombinant HSA were fixed in 2% (w/v) paraformaldehyde and 2.5% (v/v) glutaraldehyde in phosphate buffer (0.1 m, pH 7.4) overnight at 4°C. After several washing steps with phosphate buffer (0.1 m, pH 7.4), samples were postfixed in 1% (w/v) osmium tetroxide with 0.8% (w/v) KFeCN in phosphate buffer (0.1 m, pH 7.4) for 3 h at 4°C. The tissue was then dehydrated through an acetone series and infiltrated and polymerized in Spurr epoxy resin. This method preserves the tissue well, but reduces the sensitivity for the immunodetection of some proteins, including recombinant phytase. For this reason, wheat seeds expressing recombinant phytase were fixed in 4% (w/v) paraformaldehyde and 0.5% (v/v) glutaraldehyde in phosphate buffer (0.1 m, pH 7.4) overnight at 4°C. Samples were dehydrated through an ethanol series and then infiltrated and polymerized in LR White resin. The weak fixation and low temperature (−20°C) helps to increase the sensitivity of detection, but the quality of the tissue sections can be rather poor.

For light microscopy, 1-μm sections were stained in methanol blue. For electron microscopy, sections showing silver interference colors were stained in 2% (w/v) aqueous uranyl acetate. The sections were observed using a Philips EM-400 transmission electron microscope (Philips, Eindhoven, The Netherlands).


Sections mounted either on glass slides for fluorescence microscopy or on gold grids for electron microscopy were preincubated in 5% (w/v) bovine serum albumin (BSA fraction V) in phosphate buffer (0.1 m, pH 7.4) and then incubated with the appropriate antibodies: polyclonal rabbit anti-HSA or polyclonal rabbit antiphytase. Sections were then treated with the secondary antibody diluted in phosphate buffer (0.1 m, pH 7.4), goat anti-rabbit Alexa Fluor 594 for fluorescence microscopy, and goat anti-rabbit labeled with 10-nm gold particles for electron microscopy.

Lipid Detection

Lipids were detected on thin sections from wild-type wheat seeds following a protocol described by Rowley and Dahl (1977) for developing Artemisia vulgaris pollen grains. The technique used TCH as a reagent to bind macromolecules containing aldehyde groups in osmified samples. Reactions were visualized using silver proteinate, which forms an electron-dense precipitate. Thin sections were treated with a 0.2% (w/v) solution of TCH in 20% acetic acid for 24 h, followed by washes with an acetic acid series (10%, 5%, 1%). Sections were then treated with 1% (w/v) aqueous silver proteinate for 30 min in the dark and then washed thoroughly with water.

Confocal Microscopy

Sections of wild-type wheat seeds, wheat seeds expressing recombinant HSA, and wheat seeds expressing recombinant pea legumin were fixed in 4% (w/v) paraformaldehyde overnight at 4°C. After several washes, 40-μm vibratome sections were placed on glass slides coated with 0.1% (w/v) poly-l-lysine (Sigma). Sections were dehydrated and rehydrated through an ethanol series and then treated with 2% (w/v) cellulase (Onozuka R-10 from Trichoderma viride) in phosphate buffer (0.1 m, pH 7.4) for 1 h at room temperature. Samples were then treated with 0.5% (v/v) Triton X-100 in phosphate buffer (0.1 m, pH 7.4) for 1 h and preincubated in 5% (w/v) BSA (fraction V) in phosphate buffer (0.1 m, pH 7.4). The appropriate primary antibodies (IFRN 1602 monoclonal anti-HMW glutenin, polyclonal rabbit anti-HSA, or polyclonal rabbit antilegumin) were applied and diluted in phosphate buffer (0.1 m, pH 7.4). The antigen-antibody binding reaction was revealed by applying goat anti-mouse Alexa Fluor 488 or goat anti-rabbit Alexa Fluor 594 diluted in phosphate buffer (0.1 m, pH 7.4). Samples were then observed using a Leica TCS SP confocal microscope (Wetzlar, Germany).

Sequence data from this article have been deposited with the EMBL/GenBank data libraries under accession numbers M94550 and A15293.


The authors thank Duncan Keen and Julian Rodriguez for excellent technical assistance, Dr. Günter Hollweg and the staff at the Pathology Department of the RWTH Aachen for allowing us to use their microscopy facilities, Dr. Mary Parker for helpful discussions, and Dr. Richard Twyman for critical reading of the manuscript and help with its preparation.


1This work was supported by the Sofia Kovalevskaja Prize awarded by the Alexander von Humboldt Foundation and the European Framework VI project PharmaPlanta.

Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.104.050153.


  • Altschuler Y, Rosenberg N, Harel R, Galili G (1993) The N- and C-terminal regions regulate the transport of wheat γ-gliadins through the endoplasmic reticulum in Xenopus oocytes. Plant Cell 5: 443–450 [PMC free article] [PubMed]
  • Bechtel DB, Wilson JD, Shewry PR (1991) Immunocytochemical localization of wheat storage protein triticin in developing endosperm tissue. Cereal Chem 68: 573–577
  • Brandizzi F, Hawes C (2004) A long and winding road: symposium on membrane trafficking in plants. EMBO Rep 5: 245–249 [PMC free article] [PubMed]
  • Chikwamba RK, Scott MP, Mejia LB, Mason HS, Wang K (2003) Localization of a bacterial protein in starch granules of transgenic maize kernels. Proc Natl Acad Sci USA 100: 11127–11132 [PMC free article] [PubMed]
  • Christensen AH, Quail PH (1996) Ubiquitin promoter-based vectors for high level expression of selectable and/or screenable marker genes in monocotyledonous plants. Transgenic Res 5: 213–218 [PubMed]
  • Coleman CE, Larkins BA (1999) The prolamins from maize. In PR Shewry, R Casey, eds, Seed Proteins. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 109–139
  • Crofts AJ, Leborgne-Castel N, Hillmer S, Robinson DG, Phillipson B, Carlsson LE, Ashford DA, Denecke J (1999) Saturation of the endoplasmic reticulum retention machinery reveals anterograde bulk flow. Plant Cell 11: 2233–2247 [PMC free article] [PubMed]
  • Denecke J, De Rycke R, Botterman J (1992) Plant and mammalian sorting signals for protein retention in the endoplasmic reticulum contain a conserved epitope. EMBO J 11: 2345–2355 [PMC free article] [PubMed]
  • Drake P, Stoger E, Nicholson L, Christou P Ma J KC (2000) Antibody production in plants. In P Shepherd, C Dean, eds, Monoclonal Antibodies, A Practical Approach. Oxford University Press, Oxford, pp 181–206
  • Frigerio L, Pastres A, Prada A, Vitale A (2001) Influence of KDEL on the fate of trimeric or assembly-defective phaseolin: selective use of an alternative route to vacuoles. Plant Cell 13: 1109–1126 [PMC free article] [PubMed]
  • Galili G, Altschuler Y, Levanony H (1993) Assembly and transport of seed storage proteins. Trends Cell Biol 3: 437–443 [PubMed]
  • Galili G, Shimoni Y, Giorini-Silfen S, Levanony H, Altschuler Y, Shani N (1996) Wheat storage proteins: assembly, transport and deposition in protein bodies. Plant Physiol Biochem 34: 245–252
  • Hara-Nishimura I, Shimada T, Hatano K, Takeuchi Y, Nishimura M (1998) Transport of storage proteins to protein storage vacuoles is mediated by large precursor-accumulating vesicles. Plant Cell 10: 825–836 [PMC free article] [PubMed]
  • Herman EM, Larkins BA (1999) Protein storage bodies and vacuoles. Plant Cell 11: 601–613 [PMC free article] [PubMed]
  • Herman EM, Tague BW, Hoffman LM, Kjemtrup SE, Chrispeels MJ (1990) Retention of phytohemagglutinin with carboxyterminal tetrapeptide KDEL in the nuclear envelope and the endoplasmic reticulum. Planta 182: 305–312 [PubMed]
  • Hillmer S, Movafeghi A, Robinson DG, Hinz G (2001) Vacuolar storage proteins are sorted in the cis-cisternae of the pea cotyledon Golgi apparatus. J Cell Biol 152: 41–50 [PMC free article] [PubMed]
  • Höfte H, Hubbard L, Reizer J, Ludevid D, Herman EM, Chrispeels MJ (1992) Vegetative and seed-specific forms of tonoplast intrinsic protein in the vacuolar membrane of Arabidopsis thaliana. Plant Physiol 99: 561–570 [PMC free article] [PubMed]
  • Katayama H, Nagasu T, Oda Y (2001) Improvement of in-gel digestion protocol for peptide mass fingerprinting by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Rapid Commun Mass Spectrom 15: 1416–1421 [PubMed]
  • Kolarich D, Altmann F (2000) N-Glycan analysis by matrix-assisted laser desorption/ionization mass spectrometry of electrophoretically separated nonmammalian proteins: application to peanut allergen ara h 1 and olive pollen allergen ole e 1. Anal Biochem 285: 64–75 [PubMed]
  • Lerouge P, Cabanes-Macheteau M, Rayon C, Fischette-Laine AC, Gomord V, Faye L (1998) N-glycoprotein biosynthesis in plants: recent developments and future trends. Plant Mol Biol 38: 31–48 [PubMed]
  • Levanony H, Rubin R, Altschuler Y, Galili G (1992) Evidence for a novel route of wheat storage proteins to vacuoles. J Cell Biol 119: 1117–1128 [PMC free article] [PubMed]
  • Mills EN, Field JM, Kauffman JA, Tatham AS, Shewry PR, Morgan MR (2000) Characterization of a monoclonal antibody specific for HMW subunits of glutenin and its use to investigate glutenin polymers. J Agric Food Chem 48: 611–617 [PubMed]
  • Muench DG, Ogawa M, Okita TW (1999) The prolamins of rice. In PR Shewry, R Casey, eds, Seed Proteins. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 93–108
  • Müntz K (1998) Deposition of storage proteins. Plant Mol Biol 38: 77–99 [PubMed]
  • Parker ML (1980) Protein body inclusions in developing wheat endosperm. Ann Bot 46: 29–36
  • Parker ML (1982) Protein accumulation in the developing endosperm of a high-protein line of Triticum dicoccoides. Plant Cell Environ 5: 37–43
  • Parker ML, Hawes CR (1982) The Golgi apparatus in developing endosperm of wheat (Triticum aestivum L.). Planta 154: 277–283 [PubMed]
  • Rea PA, Poole RJ (1993) Vacuolar H+-translocating pyrophosphatase. Annu Rev Pl Physiol Pl Mol Biol 44: 157–180
  • Rowley JR, Dahl AO (1977) Pollen development in Artemisia vulgaris with special reference to glycocalyx material. Pollen Spores 19: 169–284
  • Rubin R, Levanony H, Galili G (1992) Characterization of two types of protein bodies in developing wheat endosperm. Plant Physiol 99: 718–724 [PMC free article] [PubMed]
  • Shewry PR (1995) Cereal seed storage proteins. In H Kigel, G Galili, eds, Seed Development and Germination. Marcel Dekker, NY, pp 45–77
  • Shewry PR (1996) Assembly and post-translation modification of wheat gluten proteins. In CW Wrigley, ed, Gluten'96. RACI, North Melbourne, Australia
  • Shewry PR, Casey R (1999) Seed Proteins. Kluwer Academic Publishers, Dordrecht, The Netherlands
  • Shewry PR, Halford NG (2002) Cereal seed storage proteins: structures, properties and role in grain utilization. J Exp Bot 53: 947–958 [PubMed]
  • Shotwell MA, Larkins BA (1989) The biochemistry and molecular biology of seed storage proteins. In A Marcus, ed, The Biochemistry of Plants, Vol 15. Academic Press, San Diego, pp 297–345
  • Shy G, Ehler L, Herman E, Galili G (2001) Expression patterns of genes encoding endomembrane proteins support a reduced function of the Golgi in wheat endosperm during the onset of storage protein deposition. J Exp Bot 52: 2387–2388 [PubMed]
  • Singh NK, Shepherd KW, Langridge P, Gruen LC (1991) Biochemical characterization of triticin, a legumin-like protein in wheat endosperm. J Cereal Sci 13: 207–219
  • Stoger E, Parker M, Christou P, Casey R (2001) Pea legumin overexpressed in wheat endosperm assembles into an ordered paracrystalline matrix. Plant Physiol 125: 1732–1742 [PMC free article] [PubMed]
  • Torres E, Gonzalez-Melendi P, Stoger E, Shaw P, Twyman RM, Nicholson L, Vaquero C, Fischer R, Christou P, Perrin Y (2001) Native and artificial reticuloplasmins co-accumulate in distinct domains of the endoplasmic reticulum and in post-endoplasmic reticulum compartments. Plant Physiol 127: 1212–1223 [PMC free article] [PubMed]
  • Toyooka K, Okamoto T, Minamikawa T (2000) Mass transport of proform of a KDEL-tailed cysteine proteinase (SH-EP) to protein storage vacuoles by endoplasmic reticulum-derived vesicle is involved in protein mobilization in germinating seeds. J Cell Biol 148: 453–464 [PMC free article] [PubMed]
  • Verwoerd TC, van Paridon PA, van Ooyen AJ, van Lent JW, Hoekema A, Pen J (1995) Stable accumulation of Aspergillus niger phytase in transgenic tobacco leaves. Plant Physiol 109: 1199–1205 [PMC free article] [PubMed]
  • Vitale A, Galili G (2001) The endomembrane system and the problem of protein sorting. Plant Physiol 125: 115–118 [PMC free article] [PubMed]
  • Wandelt CI, Khan MR, Craig S, Schroeder HE, Spencer D, Higgins TJ (1992) Vicilin with carboxy-terminal KDEL is retained in the endoplasmic reticulum and accumulates to high levels in the leaves of transgenic plants. Plant J 2: 181–192 [PubMed]
  • Wyss M, Pasamontes L, Friedlein A, Remy R, Tessier M, Kronenberger A, Middendorf A, Lehmann M, Schnoebelen L, Rothlisberger U, Kusznir E, Wahl G, et al (1999) Biophysical characterization of fungal phytases (myo-inositol hexakisphosphate phosphohydrolases): molecular size, glycosylation pattern, and engineering of proteolytic resistance. Appl Environ Microbiol 65: 359–366 [PMC free article] [PubMed]
  • Yang D, Guo F, Liu B, Huang N, Watkins SC (2003) Expression and localization of human lysozyme in the endosperm of transgenic rice. Planta 216: 597–603 [PubMed]
  • Zheng Z, Sumi K, Tanaka K, Murai N (1995) The bean seed storage protein [beta]-phaseolin is synthesized, processed, and accumulated in the vacuolar type-II protein bodies of transgenic rice endosperm. Plant Physiol 109: 777–786 [PMC free article] [PubMed]

Articles from Plant Physiology are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...